2019-10-01 00:03:09 <cim209> what's the alpine equivalent of uuidgen from ubuntu?
2019-10-01 00:18:55 <iggy> cim209: that is part of util-linux
2019-10-01 00:19:44 <iggy> cim209: according to https://pkgs.alpinelinux.org/contents?file=uuidgen&path=&name=&branch=edge
2019-10-01 00:33:44 <cim209> ah damn i don't wanna install a bundle
2019-10-01 00:59:46 <tdtrask> qrvpzvb: you can report ACF bugs on gitlab.alpinelinux.org under aports
2019-10-01 01:41:58 <tdtrask> also, feel free to describe the bug here
2019-10-01 01:59:10 <tdtrask> qrvpzvb: I reproduced the slow load / timeout of openssh status page
2019-10-01 02:02:38 <tdtrask> seems to have changed between alpine 3.9 and 3.10
2019-10-01 02:40:50 <tdtrask> nah, scratch that, it's slow name lookup when listing the connected peers
2019-10-01 03:48:19 <maldridge> sounds like you should turn off DNS resolution
2019-10-01 03:48:26 <maldridge> unless that is a feature you actively use
2019-10-01 04:06:47 <priontology> Sup, maldridge?
2019-10-01 04:06:57 <maldridge> not much, how about you?
2019-10-01 04:19:05 <maldridge> were you asking a question priontology or just saying hi?
2019-10-01 04:21:58 <priontology> Just saying hi; I had thought of a couple of questions, but nothing serious.
2019-10-01 04:22:39 <priontology> Does udev see and create the state change of an interface gaining a link, or how would you detect that change for auto-upping of interfaces?
2019-10-01 04:23:06 <maldridge> I don't believe that udev sees that, but I may be wrong
2019-10-01 07:07:58 <CvH> Hi, is there anything known that at Allwinner H2/3 the USB is not working ?
2019-10-01 07:08:15 <CvH> at the generic arm image (armv7)
2019-10-01 07:08:38 <mps> CvH: edge or stable release
2019-10-01 07:08:47 <CvH> stable
2019-10-01 07:09:02 <mps> not enabled in kernel
2019-10-01 07:09:08 <CvH> O_o
2019-10-01 07:09:24 <CvH> on purpose?
2019-10-01 07:09:54 <mps> last night I built armv7 multi platform kernel, and it should work on Allwinner
2019-10-01 07:10:26 <mps> not on purpose, no one asked and no one tested
2019-10-01 07:11:48 <mps> few weeks ago I posted kernel for someone here to test linux-vanilla MP but didn't received any feedback
2019-10-01 07:12:40 <mps> are H2/H3 64bit or 32bit
2019-10-01 07:12:54 <CvH> do you have some link ? i have h2+ (aka h3) and h5 boards here to test if you like
2019-10-01 07:14:34 <CvH> I know that they should work, we (at LibreELEC) have already some H2/H5 images that work without any problems at those socs.
2019-10-01 07:14:41 <mps> I can put apk from my last night build, if you want to test it and you brave enough to install kernel from not official repo
2019-10-01 07:15:14 <CvH> yea :)  i am brave
2019-10-01 07:15:31 <mps> ok, give me some time
2019-10-01 07:15:37 <CvH> yea no problem
2019-10-01 07:15:56 <CvH> should I upgrade to edge ?
2019-10-01 07:16:52 <mps> not necessary, I test these kernels on 3.10
2019-10-01 07:19:33 <mps> http://arvanta.net/2x3d1kfg/linux-vanilla-4.19.75-r1.apk
2019-10-01 07:20:00 <mps> be patient, it is behind slow adsl link
2019-10-01 07:20:54 <mps> and please inform me when you finish download, I have to remove it to from there
2019-10-01 07:21:30 <mps> I will test this kernel later this day on allwinner A20
2019-10-01 07:22:40 <mps> ohm, and I put guide how to install alpine on arm's sdcard and forgot that I did it ;-)
2019-10-01 07:37:49 <CvH> mps i am dling :)  where did you put the guide ?
2019-10-01 07:39:28 <mps> http://arvanta.net/2x3d1kfg/
2019-10-01 07:39:51 <mps> index.hmtl there, for which I forgot
2019-10-01 07:40:05 <mps> and probably it is outdated
2019-10-01 07:40:30 <mps> but could give some ideas how to install
2019-10-01 07:43:09 <CvH> yea I did it by try and error, I have it now working that it behaves like at the RPi image, so one fat32 partition with all data
2019-10-01 07:47:51 <mps> CvH: I booted it successfully on my A20, 'uname -a' => Linux bpialp 4.19.75-1-vanilla #2-Alpine SMP Mon Sep 30 21:08:08 UTC 2019 armv7l GNU/Linux
2019-10-01 07:48:42 <mps> lsblk show usb attached flash as disk
2019-10-01 07:49:16 <mps> and, reboot works
2019-10-01 07:50:07 <CvH> mps dl finished
2019-10-01 07:50:14 <mps> now I have to go out for some time, please don't forget to inform me when you download kernel
2019-10-01 07:50:24 <mps> ahm, nice
2019-10-01 07:50:43 <mps> just on time
2019-10-01 07:50:57 <CvH> now i try (never need to manually install the kernel, never did before)
2019-10-01 07:52:35 <mps> apk add --allow-untrusted file.apk
2019-10-01 07:53:08 <mps> it should create initramfs if you have mkinitfs installed
2019-10-01 08:03:01 <CvH> mps (1/85) Installing linux-firmware-yamaha (20190322-r1)ERROR: Failed to create lib/firmware/yamaha/ds1_ctrl.fw: No such file or directory
2019-10-01 08:03:07 <CvH> i got several erros
2019-10-01 08:04:57 <mps> try 'apk add linux-firmware-none' and then 'apk add linux-firmware-which-you-need'
2019-10-01 08:05:43 <mps> see you later, will be on mobile device only for some time
2019-10-01 08:09:03 <CvH> ERROR: Failed to create lib/modules/4.19.75-1-vanilla/kernel/net/802/mrp.ko: No such file or directory
2019-10-01 08:10:07 <CvH> after "apk add linux-firmware-none"
2019-10-01 08:10:28 <CvH> sry never installed anything from local
2019-10-01 08:14:57 <CvH> or is it due the fat32 formatted partition ?
2019-10-01 08:25:42 <mps> root fs should/must be on some linux type partition, not fat
2019-10-01 08:27:20 <AinNero> 
ACTION scratches head

2019-10-01 08:27:34 <AinNero> its possible, but definitely not supported with alpine linux
2019-10-01 08:27:57 <AinNero> fat doesn't have ownership and chmod metadata (without some whacky extensions)
2019-10-01 08:28:23 <mps> AinNero: yes, but it calls for a lot of problems, especially for novices to alpine
2019-10-01 08:28:42 <AinNero> thats why im saying "not supported"
2019-10-01 08:28:54 <mps> right,
2019-10-01 08:29:33 <mps> someone who is brave enough to install alpine on fat probably will not ask for help
2019-10-01 08:34:13 <Gropi> Despite exim having https://www.exim.org/static/doc/security/CVE-2019-16928.txt in Alpine edge exim is still at 4.92.2  (https://pkgs.alpinelinux.org/packages?name=exim&branch=edge#) Do I need to be worried?
2019-10-01 08:58:00 <CvH> just saying, I run ~100 RPi 1-3 with Alpine + fat32 :)  works like a charm
2019-10-01 09:03:24 <AinNero> fat32 as root?
2019-10-01 09:04:13 <AinNero> i can imagine that thing like SSH fail because they expect their key material to have a certain chmod
2019-10-01 09:04:24 <AinNero> which you cant set on FAT file systems..
2019-10-01 09:07:17 <CvH> yes fat32 with just root, no problems so far, also worked well with users added
2019-10-01 09:13:42 <CvH> retried it with ext4 formatted partition and still got the same error, do I miss something obvious ? "mkdir -p /lib/modules/123" wont work either due RO
2019-10-01 09:14:49 <AinNero> uhm, are you asking why you can't install files onto a read-only file system?
2019-10-01 09:15:24 <CvH> not really, just wondering why it tires to do so
2019-10-01 09:15:37 <CvH> other apk just install without a problem
2019-10-01 09:15:53 <CvH> no idea if kernel.apk does something different
2019-10-01 09:16:05 <AinNero> it uses the /lib/modules folder
2019-10-01 09:16:16 <AinNero> could you paste the output from your mount command?
2019-10-01 09:17:03 <CvH> http://ix.io/1LAu
2019-10-01 09:17:28 <CvH> mounted sd rw, just in case (didn't worked before either)
2019-10-01 09:17:32 <AinNero> this is not FAT32 as root, this is diskless
2019-10-01 09:17:42 <mps> CvH: something is wrong with your system, it is installled and work very good on my box
2019-10-01 09:17:43 <AinNero> and you cant write to /lib/modules because the modloop is mounted there
2019-10-01 09:18:23 <AinNero> you dont have any issues with FAT becayse you only use it as boot partition
2019-10-01 09:18:24 <mps> oh, he use modloop, that explains
2019-10-01 09:18:54 <CvH> ah i guess I misunderstood your question, I had a fat32 partitions with all alpine files at it, not "installed" at fat32
2019-10-01 09:19:36 <mps> do you installed it in 'sys' mode or diskless
2019-10-01 09:20:49 <CvH> this should be diskless if I just copy the files from the alpine release into a partition and boot from it without any additional steps besides setup-alpine ?
2019-10-01 09:21:08 <AinNero> yeah
2019-10-01 09:21:29 <AinNero> you can tell if its diskless from that entry: "tmpfs on / type tmpfs (rw,relatime,mode=755)"
2019-10-01 09:21:52 <AinNero> tmpfs = in ram, /dev/... = sys mode
2019-10-01 09:22:45 <CvH> tbh I have not too much experience at Alpine, the RPi "just works" so it was not necessary to dig deeper
2019-10-01 09:23:12 <CvH> is there a way to install the kernel.apk at diskless too ?
2019-10-01 09:24:06 <mps> CvH: I never tried that, so don't know how to do that
2019-10-01 09:24:46 <AinNero> i /guess/ that it might help to /etc/init.d/modloop stop and than re-install the kernel module and firmware packages
2019-10-01 09:24:58 <AinNero> but this wont be permanent
2019-10-01 09:28:17 <CvH> yea won't work, as reboot is required to test reboot :)
2019-10-01 09:29:25 <AinNero> if you have everything (kernel, initramfs, modules) in ram, you can update the boot partition via the update-kernel command
2019-10-01 09:29:29 <AinNero> but i never did that before
2019-10-01 09:33:59 <CvH> i try to install it, maybe the faster way of testing it :D
2019-10-01 09:34:11 <Gropi> If I want to support the exim maintainer by providing a patch to the recent version where would I start to build my own exim package from sources etc?
2019-10-01 09:36:26 <AinNero> cloning aports
2019-10-01 09:36:39 <AinNero> there was a wiki guide onto how to get the alpine-sdk set up
2019-10-01 09:36:57 <_ikke_> ahttps://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package
2019-10-01 09:36:59 <_ikke_> ^
2019-10-01 09:43:46 <Gropi> AinNero, _ikke_: thanks
2019-10-01 09:48:34 <CvH> is there any guide how to do a sys install for alpine?
2019-10-01 09:59:03 <AinNero> CvH: its a bit more difficult for the raspi
2019-10-01 09:59:25 <AinNero> _ikke_: will probably have the wiki link faster :D
2019-10-01 10:31:31 <_ikke_> AinNero: too bad, I just had lunch
2019-10-01 10:32:25 <misterunknown> Hi everyone. I'm about to upgrade my LXC hostsystem with many containers from alpine 3.9 to alpine 3.10. Is there something I have to consider? I already read the release notes, but last time I upgraded (from 3.8 to 3.9) I got an important hint about installing lxc-templates-legacy, so I wonder if I have to consider something this time too.
2019-10-01 10:33:15 <_ikke_> clandmeter: did we run into issues when upgrading to alpine 3.10 on our infra hosts?
2019-10-01 10:33:49 <clandmeter> hmm, i dont remember
2019-10-01 10:33:53 <clandmeter> what is the version change?
2019-10-01 10:34:37 <clandmeter> i think the intrusive lxc change was prior to 3.9
2019-10-01 10:35:23 <clandmeter> yeah, its just a revision change
2019-10-01 10:35:27 <clandmeter> same lxc version
2019-10-01 10:42:06 <mps> misterunknown: yes, lxc-templates-legacy is still needed
2019-10-01 10:43:02 <clandmeter> that not 100% true
2019-10-01 10:43:12 <clandmeter> it depends how you build your containers
2019-10-01 10:45:47 <mps> true
2019-10-01 10:51:37 <qrvpzvb> I'm trying to packge prometheus, the metrics TSDB, and it seems to default to putting everything, both it's data and logs, in the current directory
2019-10-01 10:51:56 <qrvpzvb> Debian has a patch that changes those... would a conf.d be enough for us?
2019-10-01 11:23:00 <misterunknown> clandmeter, mps: Thank you. I'll give it a try.
2019-10-01 11:28:05 <CvH> mps do you have your sources somewhere public ? so i would build an "release" with your modifications and if something is missing I could easily add it
2019-10-01 11:34:26 <mps> no, it is branch in my local aports, but I thought to send it gitlab as MR WIP
2019-10-01 11:54:59 <mps> CvH: did you got it to work
2019-10-01 12:05:35 <shambat> I'm trying to get the elapsed time of a process using ps -o etime, how can I parse the output there into seconds? etimes option is not available.
2019-10-01 12:08:12 <AinNero> shambat: try ... | awk -F  ":" '! /ELAPSED/ {print($1*60+$2)}'
2019-10-01 12:08:26 <AinNero> but that only works if it stays in the nn:nn format
2019-10-01 12:08:46 <misterunknown> shambat: You might want to install procps.
2019-10-01 12:09:49 <shambat> it outputs in a variable format, depending on the duration, it could be 1d05, 5h50, 0:18
2019-10-01 12:10:27 <shambat> misterunknown: ok I'll give that a try, will that just extend the functionality of ps?
2019-10-01 12:10:34 <AinNero> shambat: procps has ps -o etimes
2019-10-01 12:10:40 <AinNero> which is the time in seconds
2019-10-01 12:10:46 <misterunknown> Correct.
2019-10-01 12:10:55 <shambat> right. sounds like what I want
2019-10-01 12:11:05 <misterunknown> procps has everything you would expect if you know debian/ubuntu systems
2019-10-01 12:11:38 <misterunknown> I mainly use it, because I often use "ps aufx" for my lxc hostsystem and want to see which process runs in the host, and which processes are in containers.
2019-10-01 12:11:57 <AinNero> that might be because it /is/ the same software that debian ships per default :P
2019-10-01 12:13:13 <shambat> I wouldn't mind using the ps that ships with alpine, but I would also like to avoid making buggy scripts to parse the time if the job is already done elsewhere
2019-10-01 12:13:18 <misterunknown> AinNero: Ok^^ I wasn't sure of that, but yeah, makes sense^^
2019-10-01 12:13:58 <misterunknown> shambat: IMHO the ps in alpine is provided by the busybox.
2019-10-01 12:14:43 <AinNero> its not IMHO, its AFAIK
2019-10-01 12:15:00 <AinNero> 
ACTION is nitpicky today

2019-10-01 12:16:30 <mps> yes, basic ps is busybox applet
2019-10-01 12:17:23 <misterunknown> 
ACTION acknowleges defeat... :D

2019-10-01 12:17:58 <shambat> thanks misterunknown and AinNero
2019-10-01 12:18:32 <AinNero> shambat: i dont see an easy way if you want to do it portably
2019-10-01 12:19:05 <AinNero> depending on your usecase there might be some other solutions
2019-10-01 12:20:10 <mps> traverse through /proc
2019-10-01 12:20:18 <qrvpzvb> I assume one could use /proc directly
2019-10-01 12:25:01 <CvH> mps no didn't get it working, found no proper faq how to install it in sys mode, thats why I asked for the sources because creating an "release" with your modifications is likely much easier then installing
2019-10-01 14:22:49 <fredrigu> hi, I'm on debian 9 linux and I've setup a chroot with alpine and followed the wiki guide to use abuild to build a package in aports, this has worked successfully. Now I want to cross-compile packages for arm. How can I do that? I've found bootstrap.sh but I fail to use it. Running ./bootstrap.sh arm gives me the error "ERROR: binutils-unknown: build failed
2019-10-01 14:22:54 <fredrigu> "
2019-10-01 14:23:12 <fredrigu> and a bit earlier in the log *** BFD does not support target arm-unknown-none.
2019-10-01 14:26:48 <artok> you need to give target arch as parameter to bootstrap
2019-10-01 14:26:48 <artok> like ./bootstrap.sh aarch64
2019-10-01 14:27:01 <fredrigu> I did, used ./bootsrap.sh arm
2019-10-01 14:27:13 <fredrigu> but I'll try with aarch64
2019-10-01 14:27:35 <AinNero> valid ARM values are aarch64 armel armhf armv7
2019-10-01 14:27:42 <AinNero> fredrigu: whats your target device?
2019-10-01 14:27:45 <fredrigu> AinNero: thanks, thata explains it
2019-10-01 14:28:03 <AinNero> i copied it from /usr/share/abuild/functions.sh
2019-10-01 14:28:37 <fredrigu> great. So when I've run bootstrap.sh with the correct arch, how do I run abuild? Because I guess I need to pass the arch to abuild as well?
2019-10-01 14:29:20 <artok> indeed bootstrap.sh has it's own bugs, wrong order of packages installing, might try to install packages that aren't needed or aren't available
2019-10-01 14:30:57 <artok> you set ARCH=armhf env variable before abuild
2019-10-01 14:30:57 <artok> for example armhf that is
2019-10-01 14:31:56 <fredrigu> thanks
2019-10-01 14:31:59 <AinNero> take note that you can't directly crosscompile all packages
2019-10-01 14:32:07 <fredrigu> I can't?
2019-10-01 14:32:20 <AinNero> bootstrap and cross compiling is just for base
2019-10-01 14:32:29 <fredrigu> oh that's fine
2019-10-01 14:32:32 <AinNero> and then you enter it with qemu-static and build natively
2019-10-01 14:32:46 <AinNero> "natively" in quotes
2019-10-01 14:33:17 <fredrigu> oh I see
2019-10-01 14:33:27 <AinNero> fredrigu: also, i asked you whats your target device
2019-10-01 14:33:28 <fredrigu> that's not very convenient..
2019-10-01 14:33:39 <AinNero> because not all arm processors can run all arm variants
2019-10-01 14:33:51 <fredrigu> AinNero: yeah, I'm not sure about that one yet, working on it
2019-10-01 14:34:07 <AinNero> then pick armhf
2019-10-01 14:34:18 <fredrigu> I've a yocto based system today that I'm investigating if I can move to alpine instead
2019-10-01 14:34:38 <AinNero> hm, i think armel might be the most portable
2019-10-01 14:34:44 <artok> oh yeah, lots of abuild stuff get the variables to the target based on which platform it is built on
2019-10-01 14:35:02 <AinNero> fredrigu: you can tell what device you have
2019-10-01 14:35:17 <AinNero> if you are lucky you aren't the first one to run alpine there
2019-10-01 14:36:42 <fredrigu> AinNero: I'm sure I am... It's an custom PCB with a custom CPU
2019-10-01 14:37:21 <AinNero> do you have a datasheet?
2019-10-01 14:37:43 <fredrigu> AinNero: no, I'm working on getting hold on one. Company politics...
2019-10-01 14:38:59 <AinNero> otherwise you could grab the compiler touple from yocto maybe?
2019-10-01 14:39:28 <fredrigu> yeah, that's my backup plan. I'm mostly interested in apk
2019-10-01 14:39:55 <AinNero> mapping compiler touple to alpine arch happens in /usr/share/abuild/functions.sh
2019-10-01 14:40:01 <AinNero> good luck!
2019-10-01 14:41:01 <fredrigu> thanks! Well I'm running bootstrap now and it will take a while. I'm sure I'll be back with more fun questions
2019-10-01 15:35:59 <fredrigu> so now I've run bootstrap.sh and of course it fails, https://paste.ubuntu.com/p/TvwYcmr83d/ I don't quite understand the error message, it seems as the installed version of libstdc++ interfere with the compiled version
2019-10-01 15:37:51 <fredrigu> or could it be so that I need to be on edge?
2019-10-01 18:42:34 <nepugia> are the noto emoji available in the package repos? i only found a wiki article that said something about a package in edge that i didn't find
2019-10-01 18:44:16 <_ikke_> https://pkgs.alpinelinux.org/packages?name=font-noto*&branch=edge
2019-10-01 18:50:12 <Cogitri> nepugia: https://pkgs.alpinelinux.org/packages?name=*noto*&branch=edge
2019-10-01 18:50:49 <_ikke_> ..
2019-10-01 19:01:49 <mo36> I am trying to make a full disk encrypted alpine install on my laptop, had used void linux before. I have one partition which is encrypted using luks and on top of this I have two lvm partitions: root and swap. I installed grub and set it up to decrypt the partition. I have activated the lvm, cryptsetup and cryptkey features in mkinitfs. When I
2019-10-01 19:01:49 <mo36> boot, grub loads fine, decrypts the drive and passes on to the initramfs but then initfs throws and error that my lvm group is not available. I then modified the init script and manually added "cryptsetup -d path/to/keyfile luksOpen /dev/sda1 alpinecrypt" and "lvm vgchange -a y alpinegroup" and tried to boot and it said "sda1 is not available". But
2019-10-01 19:01:50 <mo36> when I would enter the same commands manually in the rescue shell dropped after the error everything worked fine. I also couldn't find any documentation on the mkinitfs features
2019-10-01 19:39:39 <midasi_> fredrigu: this issue with bootstrap.sh was already reported and fixed on edge
2019-10-01 20:37:48 <kahiru> hey, I have a usb stick prepared with setup-bootable and keep state using lbu. What is the recommended way to update the kernel there?
2019-10-01 20:39:10 <kahiru> just rerun setup-bootable against it with a newer iso?
2019-10-01 20:50:12 <AinNero> the command you are looking for is update-kernel
2019-10-01 20:50:50 <kahiru> neat, will give that a shot
2019-10-01 20:51:43 <AinNero> it might be that it requires /lib/modules to be writable
2019-10-01 20:51:49 <AinNero> so you can install the newer kernel in the first place
2019-10-01 20:52:19 <AinNero> i haven't gone that path myself, but i think i can show you the direction
2019-10-01 20:52:48 <kahiru> it's ok, I think I should manage on my own from here, but thanks
2019-10-01 21:00:34 <midasi_> I'm trying to cross-compile a custom kernel to aarch64 which fails with aarch64-alpine-linux-musl-ld: unrecognized option '-Wl,--as-needed'
2019-10-01 21:01:03 <midasi_> The used command is CHOST=aarch64 BOOTSTRAP=bootimage abuild -r
2019-10-01 21:10:33 <qa3Txu0iak0F> midasi_: -Wl is a gcc passthrough to ld. if you run ld directly, you need to drop the '-Wl,' and pass --as-needed directly to ld
2019-10-01 21:16:51 <midasi_> hmm I'm using the same APKBUILD as for the linux-vanilla kernel which is doing unset LDFLAGS. so where are these flags coming from?
2019-10-01 22:21:39 <xrs> hey guys, I have sent a new patches to the aport mailinglist. What can I expect to happen next? =)
2019-10-02 02:41:47 <Geo> Hi. "Musl minimizes the performance hit inherent in dynamic linking by providing the entire standard library and the linker in a single shared object." I am running into issues when trying to compile a binary that includes its own compat code if b64_ntop is missing. Even though the configure script correctly id's that it does not exist, the compat code isn't called and the program errors with 'symbol
2019-10-02 02:41:48 <Geo> not found'. However, doing 'make static' works fine. How can I fix this issue?
2019-10-02 03:20:50 <dalias> geo, can you give more info on what you're doing? is this just running configure and make for the affected program on an alpine system?
2019-10-02 03:39:07 <Geo> dalias: hi!
2019-10-02 03:39:41 <Geo> running configure (correctly) detects that b64_ntop is not present on the system
2019-10-02 03:40:02 <Geo> the hack is to figure out what package include b64_ntop ;)
2019-10-02 03:40:18 <Geo> but since it is not, it *should* use the compat code included in the source, instead
2019-10-02 03:41:03 <Geo> so yes, just configure and make
2019-10-02 03:41:11 <Geo> interestingly, 'make static' fixes the problem
2019-10-02 03:41:53 <Geo> google hits suggest it is an issue of libc vs glibc
2019-10-02 03:42:08 <Geo> But I'm not smart enough to know how to solve that, I guess
2019-10-02 04:12:32 <dalias> what is the error for the dynamic linked one?
2019-10-02 04:12:35 <dalias> the exact message
2019-10-02 04:13:50 <Geo> sure
2019-10-02 04:13:52 <Geo> hang on
2019-10-02 04:14:08 <Geo> Can't load modules server: Error relocating /home/eggdrop/eggdrop/modules/server.so: b64_ntop: symbol not found
2019-10-02 04:14:56 <Geo> "Error relocating"*"symbol not found" is the common error message to search for, with musl
2019-10-02 04:15:59 <dalias> try rebuilding with LDFLAGS=-rdynamic
2019-10-02 04:16:33 <Geo> sure thing
2019-10-02 04:16:50 <dalias> i think the problem is that b64_ntop is in the main program and by default dynamic symbols are not exported from it unless a link-time .so (rather than a dlopened one, which linker can't see) uses the symbol
2019-10-02 04:18:49 <Geo> crap.. i didnt do a make clean, will that be an issue?
2019-10-02 04:19:47 <dalias> should suffice just to rm the executable file then rerun make with -rdynamic added to LDFLAGS
2019-10-02 04:19:56 <dalias> it's only the final link command that should change
2019-10-02 04:20:46 <Geo> nah, I'm getting some errors on that it looks like
2019-10-02 04:20:50 <dalias> nice to see eggdrop's still around :)
2019-10-02 04:20:50 <Geo> cant hurt
2019-10-02 04:20:52 <Geo> :)
2019-10-02 04:21:02 <dalias> it's been like 20 years since i used one :-P
2019-10-02 04:21:09 <Geo> yeah!
2019-10-02 04:21:15 <Geo> still kickin, and somewhat active
2019-10-02 04:21:26 <Geo> although the script side seems to have dwindled from what it once was
2019-10-02 04:22:49 <maldridge> well, its tcl, so I can't imagine there's as many people picking it up as a first scripting language anymore
2019-10-02 04:23:20 <Geo> I agree, but don't tell the other devs
2019-10-02 04:23:33 <Geo> dalias: Still no on that
2019-10-02 04:23:43 <maldridge> I think the last time I looked at TCL I was needing to break out of the crazy menu system of some HP gear
2019-10-02 04:26:50 <Geo> gcc -g -O2 -pipe -Wall -I../.. -I../.. -I../../src  -DHAVE_CONFIG_H -I/usr/include -g3 -DDEBUG -DDEBUG_ASSERT -DDEBUG_MEM -DDEBUG_DNS  -c base64.c
2019-10-02 04:28:06 <dalias> geo, what's the final link command?
2019-10-02 04:29:13 <Geo> Do you mean this?
2019-10-02 04:29:15 <Geo> gcc   -shared -g -O2 -pipe -Wall -I. -I../../.. -I../../.. -I../../../src/mod  -DHAVE_CONFIG_H -I/usr/include -g3 -DDEBUG -DDEBUG_ASSERT -DDEBUG_MEM -DDEBUG_DNS -o ../../../uptime.so ../uptime.o -L/usr/lib -ltcl8.6  -lz -lpthread -lssl -lcrypto   && touch ../../../uptime.so
2019-10-02 04:32:32 <Geo> I'm not a compile guy and I have thick skin, so tell me if I'm not giving you what you want
2019-10-02 04:32:59 <dalias> no, that's building a module
2019-10-02 04:33:12 <dalias> it should be the one with -o eggdrop or whatever
2019-10-02 04:36:49 <Geo> gcc -g -O2 -pipe -Wall -I.. -I..  -DHAVE_CONFIG_H -I/usr/include -g3 -DDEBUG -DDEBUG_ASSERT -DDEBUG_MEM -DDEBUG_DNS -o ../eggdrop bg.o botcmd.o botmsg.o botnet.o chanprog.o cmds.o dcc.o dccutil.o dns.o flags.o language.o match.o main.o mem.o misc.o misc_file.o modules.o net.o rfc1459.o tcl.o tcldcc.o tclhash.o tclmisc.o tcluser.o tls.o userent.o userrec.o users.o  -L/usr/lib -ltcl8.6  -lz -lpthread
2019-10-02 04:36:51 <Geo> -lssl -lcrypto  md5/md5c.o compat/*.o `cat mod/mod.xlibs`
2019-10-02 04:37:39 <dalias> so -rdynamic didn't get added.
2019-10-02 04:37:47 <dalias> you can try just running the command manually with it added
2019-10-02 04:37:49 <dalias> and see if that works
2019-10-02 04:37:56 <dalias> or figure out where you need to put it in the makefile
2019-10-02 04:39:18 <Geo> That's interesting
2019-10-02 04:39:24 <Geo> we even report out configure flags
2019-10-02 04:39:26 <Geo> Configure flags: 'LDFLAGS=-rdynamic'
2019-10-02 04:39:32 <Geo> that was from the final output
2019-10-02 04:39:51 <Geo> anyway
2019-10-02 04:39:54 <Geo> lets see
2019-10-02 04:43:15 <Geo> config.cache:ac_cv_env_LDFLAGS_value='-rdynamic'
2019-10-02 04:45:46 <Geo> ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
2019-10-02 04:45:52 <Geo> hrmph.
2019-10-02 04:57:16 <Geo> yeah, I guess I'm stuck on this
2019-10-02 07:32:09 <cim209> can anyone help me with this table bash script? it's not outputting correctly in alpine but it looks fine in ubuntu https://kopy.io/xCdDT
2019-10-02 07:32:58 <cim209> the dash hyphens are missing between the + characters
2019-10-02 07:40:24 <AinNero> how do you execute it? it seems not using too much external programs that could fail
2019-10-02 07:43:27 <cim209> AinNero: there's a usage comment at the top
2019-10-02 07:43:53 <AinNero> and you execute it explicitly with bash?
2019-10-02 07:44:08 <cim209> yeah
2019-10-02 07:44:40 <cim209> i had to install util-linux just to use columns
2019-10-02 07:45:05 <cim209> i think line 64 is the culprit but i don't know enough awk
2019-10-02 08:12:40 <dm1> zz
2019-10-02 11:15:28 <midasi_> I try to cross-compile a custom kernel. I copied the APKBUILD from linux-vanilla and adapted the kernel config. However, when I try to build, I always run into the error message: aarch64-alpine-linux-musl-ld: unrecognized option '-Wl,--as-needed'
2019-10-02 11:15:50 <midasi_> the command used to build the kernel is: CHOST=aarch64 BOOTSTRAP=bootimage abuild -r
2019-10-02 11:16:13 <midasi_> the cross compilation works perfectly fine with linux-vanilla, but not with my custom kernel
2019-10-02 11:20:44 <qa3Txu0iak0F> i have a dejavu
2019-10-02 11:21:15 <qa3Txu0iak0F> 23:10:34  qa3Txu0iak0F | midasi_: -Wl is a gcc passthrough to ld. if you run ld directly, you need to drop the '-Wl,' and pass --as-needed directly to ld
2019-10-02 11:22:08 <midasi_> I never set the LDFLAGS, in fact they are unset in APKBUILD
2019-10-02 11:22:18 <midasi_> so, where should I drop the '-Wl'?
2019-10-02 11:24:09 <qa3Txu0iak0F> wherever you invoke aarch64-alpine-linux-musl-ld
2019-10-02 11:25:29 <midasi_> That's done somewhere hidden behind the command abuild -r
2019-10-02 11:27:18 <qa3Txu0iak0F> maybe set LD=gaarch64-alpine-linux-musl-gcc?
2019-10-02 11:38:45 <midasi_> qa3Txu0iak0F: I assumed that should be set automatically, probably in /usr/share/abuild/functions.sh
2019-10-02 11:39:16 <midasi_> as mentoined, it works perfectly fine with the linux-vanilla kernel, but not with the custom one. I suspect the issue lies somewhere in the kernel config file
2019-10-02 12:48:03 <qa3Txu0iak0F> strange: fgrep as-needed /usr/share/abuild/* - returns nada.
2019-10-02 12:48:33 <qa3Txu0iak0F> also this --as-needed flag sounds totally unneccessary for the kernel:
2019-10-02 12:48:41 <qa3Txu0iak0F>  > This option affects ELF DT_NEEDED tags for dynamic libraries mentioned on the command line after the --as-needed option. Normally the linker will add a DT_NEEDED tag for each dynamic library mentioned on the command line, regardless of whether the library is actually needed or not. --as-needed causes a DT_NEEDED tag to only be emitted for a library that satisfies an undefined symbol reference from
2019-10-02 12:48:43 <qa3Txu0iak0F> a regular object file or, if the library is not found in the DT_NEEDED lists of other libraries linked up to that point, an undefined symbol reference from another dynamic library. --no-as-needed restores the default behaviour.
2019-10-02 12:49:29 <qa3Txu0iak0F> maybe you can strace this and look in the strace output where it comes from?
2019-10-02 14:47:30 <Gropi_> I now managed to update community/exim to 4.92.3 and have a aports patch ready. Only to discover there is already a PR for fixing the CVE for exim since 5 days: https://github.com/alpinelinux/aports/pull/11714/commits/0099f1b793935fd3c9df80652591ee1243ecc369
2019-10-02 14:48:17 <Gropi_> Will this PR merged soon or should I proceed with trying to get my patch through?
2019-10-02 15:26:47 <fredrigu> so I'm back! :), I've sucessfully used bootstrap.sh, then I'm trying to build zlib with arch=armv7 abuild -r and zlib builds successfully but for x86_64 and not armv7. Yesterday we talked about not all packages can be build cross compiled and that most compilation is done with qemu
2019-10-02 15:27:14 <fredrigu> how can I know which packages can be crosscompiled or not. I thought zlib was pretty basic/low down in the stack
2019-10-02 19:09:05 <midasi_> fredrigu: can you try with: CHOST=armv7 BOOTSTRAP=bootimage abuild -r
2019-10-03 00:49:34 <Geo> dalias: still around?
2019-10-03 00:49:41 <dalias> yes
2019-10-03 00:49:53 <Geo> cool
2019-10-03 00:50:07 <Geo> didn't think you would be, so I wasn't ready to talk yet, haha
2019-10-03 00:50:55 <Geo> and actually, I think the advice I got from someone else is wrong anyway so... nevermind!
2019-10-03 00:54:22 <dalias> :)
2019-10-03 06:45:31 <fredrigu> midasi_: thanks, it worked fine! At least for zlib for zip I get armv7-alpine-linux-musleabihf-strip:./usr/bin/zipcloak: file format not recognized
2019-10-03 06:46:11 <fredrigu> I guess it's because usually it's compiled in qemu
2019-10-03 06:46:32 <fredrigu> I wonder however if qemu isn't quite a bit slower than doing a native cross compile?
2019-10-03 08:24:59 <Xanza> Looking for a way to add a single custom sh script to the alpine iso and have it available in the setup environment. Can't seem to find much about it.
2019-10-03 08:26:32 <_ikke_> Xanza: maybe this can help? https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage
2019-10-03 08:30:32 <Xanza> Seems to be most of it, but as far as I can tell that uses a repository to build alpine, does alpine have a public repo for the filesystem?
2019-10-03 08:30:51 <Xanza> if it does, I could just clone the repo, add my file, and then build an image and be all set.
2019-10-03 08:36:50 <_ikke_> Xanza: the repo it's talking about is an alpine package repository, not a git repository
2019-10-03 08:43:19 <Xanza> Looks like `setup-bootable` is what I'm particularly looking for.
2019-10-03 08:44:03 <Xanza> extract the alpine iso to disk, add my files, and then `setup-bootable /source alpine.iso` (hopefully)
2019-10-03 08:47:42 <mps> Cogitri: I told yesterday that the firefox you just pushed doesn't build
2019-10-03 08:48:59 <mps> sorry, wrong channel
2019-10-03 08:49:24 <Cogitri> Forgot to update the checksum, dang
2019-10-03 08:49:36 <Cogitri> But how does it not work for you? It works for me on x86_64 and on aarch64 for the one who made the Pr
2019-10-03 08:50:03 <mps> it doesn't build on x86_64
2019-10-03 08:50:47 <mps> in my test, to be more precise
2019-10-03 08:50:56 <Cogitri> It just did for me
2019-10-03 08:51:00 <mps> if it builds on your then ok
2019-10-03 08:51:12 <Cogitri> Sure you didn't run out of memory?
2019-10-03 08:51:22 <Cogitri> It needs some 2GB of RAM for every build job which is pretty massive
2019-10-03 08:51:55 <mps> no, not because of memory, something else but didn't saved log
2019-10-03 08:52:34 <mps> but ok, if it pass build when you update checksum then it is ok
2019-10-03 08:53:15 <mps> and thank you for doing it, so I can move to something else
2019-10-03 08:54:34 <Cogitri> Sure 👍
2019-10-03 08:54:48 <mps> :)
2019-10-03 09:51:14 <midasi_> fredrigu: please report the issue with the zip package as a bug
2019-10-03 10:12:33 <xrs> hey guys, quick question: I submitted a patch a few days ago. What happens next?
2019-10-03 10:16:17 <Cogitri> Where have you posted it?
2019-10-03 10:16:26 <Cogitri> And a dev should look at it in a bit
2019-10-03 13:03:37 <icyphox> do Alpine's Firefox builds have DRM disabled?
2019-10-03 15:04:19 <mps> anyone have issue with mutt and sasl auth to smtp server, error is 'No authenticators available'
2019-10-03 15:04:33 <mps> same config on debian works
2019-10-03 15:07:21 <mps> I rebuilt it with gnutls but again no luck
2019-10-03 15:38:39 <dalias> mps, you have to install all the sasl module packages
2019-10-03 15:38:59 <dalias> mps, they were separated out silently in an upgrade and broke everything
2019-10-03 15:39:39 <mps> dalias: cyrus-sasl?
2019-10-03 15:40:28 <mps> iirc, it worked also on alpine year ago
2019-10-03 15:40:31 <dalias> cyrus-sasl-* (whichever ones you need)
2019-10-03 15:40:32 <dalias> right
2019-10-03 15:40:45 <dalias> it broke for me around january and i was really frustrated
2019-10-03 15:40:52 <dalias> because the change was not documented at all
2019-10-03 15:41:14 <mps> quite understand your feelings now
2019-10-03 15:41:16 <dalias> the change lets you avoid hard deps on libs you don't need/use
2019-10-03 15:41:33 <dalias> but it should not have been silent
2019-10-03 15:41:41 <dalias> the old package should have been made a metapackage depending on them all
2019-10-03 15:41:54 <dalias> and new package should have been introduced without the dep to let you pick just what you want
2019-10-03 15:41:54 <MY-R> icyphox, looks like no, is built with drm support
2019-10-03 15:42:15 <mps> dalias: fully agree
2019-10-03 15:42:52 <mps> I really hate when I upgrade a package and it stops to work without notice
2019-10-03 15:43:05 <dalias> yes
2019-10-03 15:43:12 <MY-R> oh ye I know that pain too mps :\
2019-10-03 15:43:16 <dalias> that's actually been my experience over and over with alpine :(
2019-10-03 15:43:22 <dalias> if you do apk upgrade not so much
2019-10-03 15:43:36 <dalias> but if you upgrade individual packages, they often fail to account for their library version deps
2019-10-03 15:43:57 <dalias> and then you have to track down which library the missing symbol is supposed to be provided by and upgrade that one
2019-10-03 15:44:14 <dalias> and apk won't let you upgrade it if you also have the -dev package installed unless you explicitly ask to upgrade the -dev package
2019-10-03 15:44:38 <dalias> because the -dev package has a forced exact version dependency on the lib, i think
2019-10-03 15:44:51 <mps> yes, that is price for small and simple
2019-10-03 15:46:06 <mps> I must admit that also I added something to this, removed polkit from 'slim'
2019-10-03 15:48:14 <dalias> :)
2019-10-03 15:49:53 <mps> hehe, I like to remove things, perfection is not when you have nothing to add but perfection is when you have nothing to remove :)
2019-10-03 15:50:37 <dalias> rootkit is always good to remove
2019-10-03 15:50:44 <mps> like 'someone' did when created musl
2019-10-03 15:53:01 <nepugia> mps, perhaps the solution to these problems is to remove dynamic linking entirely ;)
2019-10-03 15:53:12 <nepugia> That would avoid any library mismatch problems
2019-10-03 15:53:38 <mps> no, I don't think so
2019-10-03 15:53:40 <nepugia> (although linking to gpl libs statically is .... eh.... so maybe link them on the target computer? which then has the same dependency problems lol)
2019-10-03 15:54:31 <mps> we have to live with imperfection in world
2019-10-03 15:54:52 <mps> remember saying, perfect is enemy of good
2019-10-03 15:56:42 <artok> searching perfection is not getting stuff done
2019-10-03 15:57:24 <mps> nepugia: although to some degree I agree with you, static link with musl is not bad as with 'other'
2019-10-03 15:58:23 <mps> my first real experience with musl was on debian some years ago when I build go programs with musl instead of glibc
2019-10-03 15:59:37 <mps> and at this time I started to search good distro with musl as base libc
2019-10-03 16:04:34 <dalias> i still like alpine, these are just things i've been repeatedly frustrated with that i wish would improve
2019-10-03 16:06:32 <mps> yes, that is the reason I devote some of my free time to improve it
2019-10-03 16:10:24 <icyphox> MY-R: that's weird. netflix/prime/spotify -- none of them seem to work for me.
2019-10-03 16:13:28 <mps> dalias: to confirm, 'sudo apk add cyrus-sasl-plain' solved the mutt problem I had
2019-10-03 16:18:43 <MY-R> icyphox, in firefox preferences did you enable DRM content?
2019-10-03 16:26:28 <dalias> mps, yeah that's the one i needed too
2019-10-03 16:27:01 <dalias> my-r, i was never able to get that stuff to work. i'm skeptical that it ever works with open source firefox rather than official builds
2019-10-03 16:27:24 <dalias> if anyone knows how to get it to work i'd be interested
2019-10-03 16:28:13 <MY-R> dalias, I'm not using such a services and drm in general but every few months can see some news that it should support it
2019-10-03 18:36:35 <mps> what ':=' does in Makefile, assign variable explicitly
2019-10-03 18:36:44 <mps> gnu make, I mean
2019-10-03 18:43:02 <Cogitri> https://ftp.gnu.org/old-gnu/Manuals/make-3.79.1/html_chapter/make_6.html explains it well
2019-10-03 18:45:07 <mps> I read it but don't understand this ':='
2019-10-03 18:46:38 <mps> if i have 'VAR := foo' and I create env VAR=bar what will be value when I invoke make
2019-10-03 18:47:34 <Cogitri> Basically if you have `CPU_COUNT := $(nproc) ` the variable will always stay what nproc returned when you defined the variable
2019-10-03 18:48:05 <Cogitri> Even if you add cpu cores later on during te run of the script (and as such nproc would return something different) the variable won't change
2019-10-03 18:49:47 <mps> so, iiuc, setting env VAR will not overide one in makefile?
2019-10-03 18:51:26 <mps> that is
2019-10-03 18:51:31 <mps> just tested
2019-10-03 18:51:38 <mps> uhm
2019-10-03 18:52:46 <Cogitri> I think it will
2019-10-03 18:53:24 <mps> my issue is here, https://github.com/crystal-lang/crystal/blob/master/Makefile line 32 and 36
2019-10-03 18:55:30 <mps> whatever I set it in env CRYSTAL_CONFIG_BUILD_COMMIT it always do 'git rev-parse --short HEAD 2> /dev/null' and puts in this var
2019-10-03 18:57:29 <mps> ufm
2019-10-03 18:58:22 <mps> if use 'VAR ?= foo' then it works
2019-10-03 21:50:09 <nepugia> Is there some easy package-manager way to get the noto (fallback) emoji font and just that?
2019-10-03 21:53:25 <MY-R> nepugia, install font-noto-emoji package only?
2019-10-03 21:54:09 <MY-R> nepugia, which is in edge only :\
2019-10-03 21:54:32 <MY-R> so just wait for next release :)
2019-10-03 21:55:10 <mps> nepugia: I put most of fonts I need ~/.fonts dir
2019-10-03 21:56:35 <MY-R> ahh not only in edge but in test repo so ye like mps says :P
2019-10-03 22:26:57 <nepugia> Well, i dont have a problem with writing to /usr/share/fonts/noto really
2019-10-03 22:27:14 <nepugia> Putting fonts there doesn't seam to do much though, even after runnign fc-cache -f -v
2019-10-03 22:30:10 <mps> fc-cache -f /usr/share/fonts/noto
2019-10-03 22:30:37 <nepugia> it does list the dir as known
2019-10-03 22:30:52 <mps> hmm, don't know then
2019-10-03 22:31:06 <nepugia> putting fonts in ~/.fonts does seam to work, (and by work meaning giving me some terrible fonts out of noto)
2019-10-03 22:31:16 <nepugia> but still no emoji from the noto emoji font :/
2019-10-03 22:33:31 <mps> still have no idea, don't use emoji fonts
2019-10-03 23:11:54 <midasi_> how can I load an apkovl file after reboot?
2019-10-03 23:16:36 <manuellll> Hello everyone
2019-10-03 23:16:44 <manuellll> Just a short question:
2019-10-03 23:17:42 <manuellll> I have a apkovl in the /boot directory of the sdcard which seems to get restored properly at boot. With exception: /etc/runlevels/default/sshd
2019-10-03 23:17:54 <manuellll> That is, sshd doesn't get started automatically on boot
2019-10-04 07:15:58 <CvH> Hi, can anyone outline to me how I build from scratch the "alpine-uboot-3.10.2-armv7" release ? I already got it working that github.com/alpinelinux/alpine-chroot-install works for me (script is broken for Ubuntu since 1+ year) but then I am a bit lost. There is also no documentation (that I found)  around that topic.
2019-10-04 07:49:24 <midasi_> CvH: take a look at https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage
2019-10-04 08:05:56 <tmhoang> mps: it was me yesterday asking you about wiregaurd on Alpine :) I tried again with 2 different Alpine vps, same performance 0.5-2 MB/s. Tested with a local ubuntu server, seems to push to my max bandwidth. Client is still debian.
2019-10-04 08:07:52 <tmhoang> this week is not so nice for me, for the first time, I was pushing some bandwidth via openvpn through my local router, other local machine got lagging :(
2019-10-04 08:09:58 <tmhoang> local roouter is Alpine
2019-10-04 08:17:42 <tmhoang> and I only installed wireguard-vanilla, wireguard-tools, not sure that's enough
2019-10-04 08:22:16 <CvH> @midasi_ just to wrap up, enter chroot from "alpinelinux/alpine-chroot-install" -> "mkimage.sh" to build an release from the current repo files, if I want to build custom packages images i need to build my own repo and point to it at "mkimage.sh" ?
2019-10-04 08:34:06 <midasi> CvH: you need to clone the aports git first and then call mkimage.sh with a custom profile. What is your base OS?
2019-10-04 08:34:39 <CvH> ubuntu
2019-10-04 08:35:22 <CvH> btw my build environment in a x86_64 vm
2019-10-04 08:37:15 <midasi> oh ok, that's the reason for the chroot... I never tried to use mkimage within a chroot. I always use it on top of an Alpine Linux installation
2019-10-04 08:39:18 <CvH> yea but doing an build at some low end arm platform doesn't sounds right ;)
2019-10-04 08:39:57 <midasi> I also do cross-builds for arm on my x86_64 Alpine build server ;-)
2019-10-04 08:40:22 <CvH> how do you do it ?
2019-10-04 08:40:30 <fredrigu> CvH: what I've understood (and trying to reproduce) is that you avoid cross-builds on alpine but instead uses qemu to boot a small system that can do all the compilations
2019-10-04 08:40:41 <fredrigu> I'm also trying to avoid that and doing native cross compiles
2019-10-04 08:40:57 <CvH> i basically just want to build an alpine release for arm :D
2019-10-04 08:41:15 <CvH> at an x86_64 host
2019-10-04 08:43:23 <fredrigu> CvH: I'm also curious about how midasi does it. I've only done it half the way. Using bootstrap.sh to get a cross-compile toolchain and then abuild to compile packages for ARM, then how to build an image is still unknown to me
2019-10-04 08:44:40 <CvH> tbh I am a bit spoiled from the buildscript that I usually use that builds an hole distribution in a single command for every arch, but that won't help if you want alpine
2019-10-04 08:45:13 <fredrigu> CvH: which build script is that? yocto or buildroot?
2019-10-04 08:46:00 <CvH> libreelec ;)
2019-10-04 08:46:28 <fredrigu> CvH: never heard of, thanks!
2019-10-04 08:47:19 <midasi> CvH: I'm using exactly the same method as mentioned by fredrigu running scripts/bootstrap.sh to create a chroot environment that can be used to cross-compile packages.
2019-10-04 08:49:26 <midasi> then you build all the packages you need for your image and use mkimage.sh with a custom profile to create your image
2019-10-04 08:51:52 <CvH> midasi tx, I give it a go - btw that kind of information would be nice at some wiki page ;)
2019-10-04 08:53:30 <fredrigu> CvH: I guess the answer to that is, "yes it would, could you please write one?" ;)
2019-10-04 08:53:42 <fredrigu> CvH: please let me know your progress, I'm currently working on the same thing
2019-10-04 08:55:19 <midasi> CvH: agreed, but I would rather not use this solution for a production environment. Some packages cannot be cross-built with this method and have to be fixed. A native build is always better and safer.
2019-10-04 08:58:07 <CvH> midasi just out of congruity, can you name one that is not working ?
2019-10-04 08:58:22 <midasi> CvH: ask fredrigu ;-)
2019-10-04 08:58:31 <CvH> curiosity ;D
2019-10-04 08:58:48 <CvH> ok :D  fredrigu
2019-10-04 09:01:08 <fredrigu> I've not come so long yet, I've just compiled packages that I know will work in cross compilation, such as zlib
2019-10-04 09:01:24 <fredrigu> but I guess there's a good reason for not using cross-compilation all the way
2019-10-04 09:01:43 <midasi> fredrigu: you mentioned you had some issues with the zip package, didn't you?
2019-10-04 09:02:08 <TBB> I don't know... both embedded build systems I use do cross compilation from start to finish
2019-10-04 09:05:33 <midasi> TBB: with what method? bootstrap.sh?
2019-10-04 09:08:28 <fredrigu> midasi: yes I did but, I'm not yet sure if it was a real error or just me doing things wrong
2019-10-04 09:09:42 <midasi> fredrigu: I just tried it on my environment, I have the same issue...
2019-10-04 09:09:54 <fredrigu> midasi: with zip?
2019-10-04 09:09:58 <midasi> yes
2019-10-04 09:10:04 <fredrigu> great then I'm not alone :)
2019-10-04 09:10:05 <TBB> nope, other build systems made for building things for embedded platforms with many different kinds of hardware
2019-10-04 09:10:44 <fredrigu> TBB: yes, so what alpine does is either really smart or just a quick hack
2019-10-04 09:11:20 <fredrigu> from what I can see it's just a quick hack. You save to of development time but you pay in increase compile times. That might be ok for alpine, but it's unfortunately not for me
2019-10-04 10:22:58 <CvH> so I try to follow the steps midasi has provided and try to install alpine x86_64 vm image into a vmware vm and I got an kernel panic ....
2019-10-04 10:24:26 <CvH> so is there some magic how this should work ?
2019-10-04 10:24:31 <CvH> or just broken
2019-10-04 10:27:03 <RRedcroft> hi guys, is it possible to install mu4e with emacs-x11?
2019-10-04 10:27:57 <RRedcroft> mu4e wants to install emacs-nox, which removes emacs-x11
2019-10-04 10:28:45 <TBB> fredrigu, it could actually make some sense in some twisted way to utilize something like buildroot for building initial parts of Alpine, but I don't dare suggest anything like that since I'm not deeply familiar with the Alpine side of things
2019-10-04 10:59:57 <fredrigu> TBB: I'm working with yocto today and I'm looking in replacing it with abuild. We choosed yocto since buildroot isn't "good enough"
2019-10-04 11:00:19 <fredrigu> I would rather see that abuild is improved than that we introduce a dependecy on buildroot
2019-10-04 11:00:54 <fredrigu> anyway. I've just got OK in investigating if apk is faster than dnf, so whish me luck :)
2019-10-04 11:01:23 <TBB> it is, and it isn't, depending on what factors you're looking at
2019-10-04 11:06:29 <fredrigu> TBB: yes, it's way better in some cases (we used it at my previous work). That's why I used " marks
2019-10-04 11:08:00 <TBB> I meant that in the context of dnf vs apk, but you're correct nevertheless :)
2019-10-04 11:21:15 <fredrigu> TBB: ah, I see :). For my workload speed is very important. So if I can get the same end result and apk is faster (and has a smaller database) than dnf, I'll call it a win
2019-10-04 11:23:52 <TBB> from that point of view I believe apk is faster. dnf becomes handy when we're talking things like using deltas to upgrade some or the larger packages
2019-10-04 11:24:46 <fredrigu> yeah, that's my hope as well
2019-10-04 11:29:04 <nepugia> I have seen claims that dnf was faster than yum, but honestly could never reproduce that :(, both were really slow to me
2019-10-04 11:33:53 <Manuel> Hello everyoneJust a short question:I have a apkovl in the /boot directory of the sdcard which seems to get restored properly at boot. With exception: /etc/runlevels/default/sshdThat is, sshd doesn't get started automatically on boot
2019-10-04 11:34:21 <Manuel> Oh some newlines got removed.. Hope you can decifer it
2019-10-04 11:34:49 <nepugia> decipher what?
2019-10-04 11:34:56 <_ikke_> The question
2019-10-04 11:35:06 <nepugia> Ah
2019-10-04 11:35:12 <_ikke_> Manuel: did you run rc-update add sshd?
2019-10-04 11:36:26 <Manuel> Yes, I ran `rc-update add sshd default'
2019-10-04 11:37:15 <_ikke_> and after that `lbu ci`?
2019-10-04 11:38:04 <Manuel> And when peek into the apkovl (`tar tf alpine.apkovl.tar.gz') I created using `apk package /dev/mmcblk0p1/boot', I can see it contains a `/etc/runlevels/default/sshd'.
2019-10-04 11:39:02 <_ikke_> apk package?
2019-10-04 11:39:09 <Manuel> Oh no it was `lbu package /dev/mmcblk0p1/boot`, not `apk ...` .... I'm new to alpine
2019-10-04 11:39:16 <_ikke_> ok
2019-10-04 11:39:44 <_ikke_> I think you should run lbu ci (lbu commit)
2019-10-04 11:41:13 <Manuel> If I understood the wiki correctly, `lbu commit` and `lbu package` should do more or less the same.. except that the latter allows you to choose a path to save to
2019-10-04 11:41:57 <_ikke_> Ok, I wasn't sure
2019-10-04 11:42:22 <_ikke_> And are you sure it is using that specific apkovl? afaik, it needs to exist in the root
2019-10-04 11:42:27 <_ikke_> the rood of the boot media
2019-10-04 11:42:29 <_ikke_> root*
2019-10-04 11:46:07 <Manuel> Yes, because after booting it's connecting to the wlan. It wouldn't know the credentials otherwise.
2019-10-04 11:46:54 <_ikke_> ok
2019-10-04 12:04:28 <CvH> has anyone an idea how alpine works at vmware vm ? currently it results at an kernel panic at early boot, https://wiki.alpinelinux.org/wiki/Install_Alpine_on_VMWare isn't working either
2019-10-04 12:04:45 <Manuel> It also doesn't install the packages in `/etc/apk/world`, as it should.. :(
2019-10-04 12:06:18 <_ikke_> Manuel: just to be sure, is there an apkovl in the root of the sdcard?
2019-10-04 12:06:51 <little-birds> hi everybody
2019-10-04 12:07:19 <little-birds> midori is not working properly
2019-10-04 12:09:34 <little-birds> https://iili.io/H9edej.png
2019-10-04 12:10:03 <little-birds> EGLDisplay Initialization failed: EGL_NOT_INITIALIZED
2019-10-04 12:10:06 <little-birds> Cannot create EGL sharing context: invalid display (last error: EGL_SUCCESS)
2019-10-04 12:10:11 <little-birds> ^
2019-10-04 12:10:18 <little-birds> that's the error that i get
2019-10-04 12:10:57 <little-birds> i'm running alpine inside a docker container, with xvfb and x11vnc
2019-10-04 12:12:33 <little-birds> ncopa
2019-10-04 12:12:55 <little-birds> i think that you are the maintainer
2019-10-04 12:13:06 <little-birds> of the midori package
2019-10-04 12:19:53 <CvH> lol ... Alpine Virtual images for VM just supports <=3000MB RAM, >=3096MB RAM is already segfaulting at start .... , not an huge problem but likely not intended
2019-10-04 12:23:54 <AinNero> thats a big surprise to me
2019-10-04 12:26:25 <Manuel> _ikke_: Hm. It doesn't make sense to me. Yes, now my apkovl is in the root directory of the sdcard. Previously, it was in the `boot` directory.
2019-10-04 12:27:17 <fredrigu> Manuel: I have a vague memory about all apkovl in the root directory of every device will be autoloaded at boot
2019-10-04 12:28:08 <_ikke_> yes, it looks for the apkovl in the root dir
2019-10-04 12:28:33 <_ikke_> My suspicion is that that apkovl is used, and the updated apkovl files in /boot not
2019-10-04 12:28:43 <Manuel> Current state of affairs is: The `/etc/wpa_supplicant/wpa_supplicant` is from the apkovl, `/etc/apk/world` is as well, but `/etc/runlevels/default/` is not
2019-10-04 12:28:59 <_ikke_> Manuel: the question is *which* apkovl
2019-10-04 12:29:25 <_ikke_> Manuel: lbu is used a lot, if it was not working in such a fundamental way, it would have been noticed already
2019-10-04 12:29:25 <AinNero> CvH: i can't reproduce, boots fine for me with 10GB or more
2019-10-04 12:29:42 <Manuel> The one which I had in the `boot` directory in the sdcard before, and which I now put in the root of the sdcard
2019-10-04 12:30:51 <Manuel> When I copy it to `boot` before, I get a fresh system after boot. But I swear it wasn't like this at my previous tests.
2019-10-04 12:31:46 <Manuel> Something is a bit fishy here. Doesn't make sense to me at all.
2019-10-04 12:33:09 <AinNero> CvH: i just tested with 64GB, no issue.
2019-10-04 14:27:25 <ncopa> little-birds: patches are welcome...
2019-10-04 14:35:43 <little-birds> ncopa: please make them come.
2019-10-04 14:42:38 <ncopa> little-birds: can you please report the problem on gitlab.a.o/aports
2019-10-04 14:58:51 <misterunknown> Hi. I have an alpine hostsystem for containers and just created a devuan container (ascii). I installed openssh-server and it works, but I can't log in because SSH says:
2019-10-04 14:59:03 <misterunknown> Oct  4 14:57:06 service sshd[700]: fatal: linux_audit_write_entry failed: Unknown error -1
2019-10-04 14:59:15 <misterunknown> Has anyone seen this error yet?
2019-10-04 15:00:59 <AinNero> have you tried searching on google?
2019-10-04 15:00:59 <AinNero> https://github.com/moby/moby/issues/6770
2019-10-04 15:02:11 <AinNero> the container needs a audit subsystem to write to
2019-10-04 15:02:42 <AinNero> or you manage to disable auditing
2019-10-04 15:07:19 <misterunknown> AinNero: Ah, ok, thanks. I googled a bit, but I also run several other devuan containers without any problems. My special problem was, that I had a container with alpine linux and glibc installed which I wanted to replace with a devuan container, and simply switched the volume – but didn't change
2019-10-04 15:07:24 <misterunknown> lxc.include = /usr/share/lxc/config/common.conf
2019-10-04 15:07:54 <misterunknown> (prior to my fix, there was alpine-common.conf referenced)
2019-10-04 15:08:04 <misterunknown> So, thanks for your hint!
2019-10-04 20:02:31 <strfry> hi, i'm having a question regarding my VM setup. basically what i want to achieve is booting from a minimal disk containing only the kernel/initrd and an apkovl
2019-10-04 20:03:15 <strfry> to get those boot files, i resorted to using the netboot release, but that poses some severe limitations, like not even providing a ext4 fs driver
2019-10-04 20:03:36 <strfry> what's the correct place to report to with that kind of issue?
2019-10-04 20:04:35 <_ikke_> https://gitlab.alpinelinux.org/alpine/aports Probably (not sure if there is a more specific project involved that's more suited)
2019-10-04 20:04:40 <mps> strfry: gitlab.a.o
2019-10-04 20:04:43 <strfry> i dont get the story behind this, the netboot initrd has virtio blk drivers, but no proper filesystems (except FAt)?
2019-10-04 20:05:20 <strfry> mps: yeah i find that, but i need to pick a specific project. so i'll go with aports, as _ikke_ suggested, thanks
2019-10-04 20:06:25 <mps> yes, best option is to look there and see which section is appropriate to fill a bug
2019-10-04 20:07:31 <mps> and _ikke_ 's advice are usually best
2019-10-04 20:08:21 <mps> advises*
2019-10-04 20:14:45 <_ikke_> not always :)
2019-10-04 20:17:47 <mps> but *usually*
2019-10-04 20:50:59 <elxa> can someone explain why I get "Backtrace stopped: previous frame identical to this frame (corrupt stack?)" in gdb backtraces only when using alpine linux as my docker container but not with debian ? ARM system raspberry pi 3
2019-10-04 21:55:24 <elxa> doesn't happen on my x86 pc :(
2019-10-04 22:22:30 <agb> I see that back in April the python bindings for newt were disabled in commit https://git.alpinelinux.org/aports/commit/main/newt?id=c3e801b36a90066f1e2c45c722f063fd83ed00ee with a commit message that they will be disabled until someone asks for support. I'd like to request python support be readded to newt. Let me know if some testing is required!
2019-10-04 22:40:10 <_ikke_> agb: Could you add an issue on gitlab.a.o?
2019-10-05 01:11:01 <agb> _ikke_: sure. can do! (Also I didn't realize Alpine was using Gitlab now, cool!)
2019-10-05 01:23:27 <iggy> agb: keep in mind that may have also been part of the "move to python3" prep work, so make sure whatever it is works with python3
2019-10-05 04:21:15 <shodan45> the speed of alpine on bare hardware still blows my mind a little
2019-10-05 04:26:31 <shrizza> shodan45: Are you referring to any specific hardware?
2019-10-05 04:26:56 <desultory> presumably their own. ;)
2019-10-05 04:26:57 <shodan45> shrizza: hah, no. my now kinda old home server.
2019-10-05 04:27:26 <shodan45> just ssh'ing to the thing is stupidly fast
2019-10-05 04:27:26 <shrizza> What hardware is it?
2019-10-05 04:27:46 <shodan45> Intel(R) Pentium(R) CPU G2120 @ 3.10GHz
2019-10-05 04:29:05 <shodan45> it's perfect for a little home server :)
2019-10-05 04:29:38 <shodan45> mostly just backups ATM; trying to get syncthing going now
2019-10-05 04:30:45 <shodan45> I use centos at work, and it's just plain poky compared to this
2019-10-05 11:00:30 <mps> anyone know why Alpine use DEVICETREEDIR instead of FDTDIR in extlinux.conf for arm?
2019-10-05 11:01:49 <mps> for u-boot extlinux.conf , I mean
2019-10-05 12:21:59 <telmich> good day
2019-10-05 12:22:11 <telmich> I am trying to understand how the boot process in uefi work w/ alpine
2019-10-05 12:22:28 <telmich> I mounted the iso, created a fat filesystem on the efi partition on a usb stick
2019-10-05 12:22:32 <telmich> tar'ed everything over there
2019-10-05 12:22:55 <telmich> The bios detects it as bootable, however when I try to boot, only a cursor on the top left is shown and the dell logo does not vanish
2019-10-05 12:23:36 <telmich> If I see it correctly, ./efi/boot/bootx64.efi should be loaded and afaics this is a grub and should try to load the grub.cfg - do I see it roughly correcty?
2019-10-05 12:24:13 <telmich> I modified the grub.cfg and additionalyl the syslinux.cfg to contain modprobe.blacklist=intel_lpss_pci, but even without that kernel command line, the behaviour is identical
2019-10-05 12:24:42 <telmich> ubuntu/arch/debian usb sticks work however, so it might be something I am doing wrong on the way of creating the bootable usb stick
2019-10-05 12:38:01 <CvH> hi, anyone an idea what might be missing ? I still try to build alpine for armv7 (./bootstrap.sh armv7) but currently it errors out with "ERROR: build-base-armv7: builddeps failed"
2019-10-05 12:38:12 <CvH> ERROR: unsatisfiable constraints:
2019-10-05 12:38:12 <CvH> satisfies: g++-8.3.0-r0[gcc=8.3.0-r0] build-base-0.5-r1[gcc] .makedepends-build-base-armv7-20191005.143417[gcc]
2019-10-05 12:38:56 <telmich> yeah, I just booted alpine via the ubuntu grub!
2019-10-05 12:39:04 <CvH> i running this at an alpine 3.10.2 x86_64 vm image with cloned aports
2019-10-05 12:39:12 <telmich> and a kernel panic...
2019-10-05 12:39:45 <CvH> try to lower ram sub 3gb :)  worked for me
2019-10-05 12:40:52 <telmich> trying w/ noapic now
2019-10-05 12:40:57 <telmich> that boots!
2019-10-05 12:41:38 <telmich> the kernel on the install media is quite "old", for that laptop
2019-10-05 12:42:42 <nrxr> CvH: seems like you need to update your aports repository? That's what tha error means
2019-10-05 12:43:56 <CvH> damn ... i forgot to switch 3.10 branch at aports ...
2019-10-05 12:44:03 <CvH> new try :)
2019-10-05 12:44:10 <CvH> tx for pointer
2019-10-05 12:45:31 <nrxr> np ;-)
2019-10-05 12:59:22 <telmich> when trying to update the package list on the install media, I get:
2019-10-05 12:59:23 <telmich> Updating repository indexes... ERROR: http://mirror.ungleich.ch/mirror/packages/alpine/v3.10/main: Bad file descriptor
2019-10-05 13:01:19 <telmich> ah, /var unmounting was the problem
2019-10-05 13:23:05 <telmich> do I remember correctly that there were two kernel packages, one more bleeding edge?
2019-10-05 13:23:41 <telmich> I ask, because I am looking for a 5.3 kernel
2019-10-05 13:24:39 <_ikke_> telmich: There ideas to implement that, but for now we only have an LTS kernel
2019-10-05 13:24:46 <telmich> oh, ok
2019-10-05 13:24:59 <telmich> So if I need 5.3, best way is to build it myself at the moment?
2019-10-05 13:25:04 <_ikke_> yes
2019-10-05 13:25:08 <telmich> ok
2019-10-05 13:31:50 <telmich> hmm, after installation, it is again hanging with the cursor in the left upper corner
2019-10-05 13:50:34 <misterunknown> Hi guys. Quick question: I use php-fpm with different users, and want to update nextcloud via cli, which should be done as this user. The problem is, that this users don't have a login shell (/bin/false). I could admittedly change the login shell, but as I just realized, I forgot to change the login shell back to /bin/false the last time. Is there a way to start a bash under these users, despite
2019-10-05 13:50:36 <misterunknown> /bin/false as login shell?
2019-10-05 13:53:49 <Cogitri> `sudo -u`
2019-10-05 13:55:15 <misterunknown> Cogitri: Indeed. I hassled a bit with "su", but it didn't work, so I guessed that sudo won't work either, but it does. Thanks.
2019-10-05 13:56:23 <Cogitri> 👍
2019-10-05 13:58:10 <CvH> the ncurses version used at alpine 3.10 can't be downloaded from source mirror anymore :D   ideas ?
2019-10-05 14:05:16 <CvH> isn't there any mirror build in ? no idea what actual licence aports have but if it is gpl you need one (just a hint)
2019-10-05 14:18:12 <AinNero> CvH: its right there http://dl-cdn.alpinelinux.org/alpine/v3.10/main/x86_64/ncurses-6.1_p20190518-r0.apk
2019-10-05 14:19:35 <AinNero> the sources are at https://invisible-mirror.net/archives/ncurses/current/ncurses-6.1-20190914.tgz
2019-10-05 14:20:40 <CvH> https://github.com/alpinelinux/aports/blob/3.10-stable/main/ncurses/APKBUILD#L3
2019-10-05 14:21:03 <CvH> is not available at http://invisible-mirror.net/archives/ncurses/current/
2019-10-05 14:21:19 <CvH> so you can't build 3.10 without manual fixing
2019-10-05 14:24:39 <AinNero> it seems that upstream removed it
2019-10-05 14:24:48 <AinNero> ideas: use code from master
2019-10-05 14:25:04 <mps> backport
2019-10-05 14:25:20 <CvH> yea i used master for now
2019-10-05 14:25:31 <mps> and build it locally
2019-10-05 14:27:21 <CvH> just saying that keeping release buildable from source is one of the key gpl requirements ;)  just saying before some tin foil hat comes around the corner
2019-10-05 14:27:31 <AinNero> CvH: file a bug at gnu.org then
2019-10-05 14:27:45 <AinNero> and tell then how to do GPL
2019-10-05 14:30:55 <AinNero> what is your expectation here? that alpine rehosts every tarball forever?
2019-10-05 14:37:59 <AinNero> (note: upstream is a GNU project)
2019-10-05 14:50:10 <CvH> if you release your distro under gpl (no idea actually about alpine) you need to store every needed file to rebuild every release you made, this also includes the sources of you packages
2019-10-05 14:50:36 <CvH> don't get me wrong, I see pretty clear that this is a pain
2019-10-05 14:51:46 <CvH> we faced the same problems (~50% of the packages of alpine) but in the end we just wrote some small script that puts every file to some server, no big deal in the end and keeps the foil hats away
2019-10-05 14:52:18 <CvH> these requirements are pretty clearly never made with a distro in mind
2019-10-05 14:58:55 <CvH> https://github.com/alpinelinux/aports/commit/e6b7d47e7a19d968e9fee0faa43a74fd2ceba7c7 needs also a backport to 3.10, script won't work at 3.10 without it
2019-10-05 15:08:33 <AinNero> CvH: GPL Section 6 d)
2019-10-05 15:09:29 <AinNero> i think this covers our usecase
2019-10-05 15:10:57 <Cogitri> Also, you could always just checkout the release tag in SCM
2019-10-05 15:11:04 <Cogitri> No need to mirror the tarball
2019-10-05 15:12:32 <Cogitri> Also, GPL only requires you to publish downstream patches in case you're making binaries, so I'm pretty confident were covered in that department too since you can just pull the patches from our git repo
2019-10-05 15:14:11 <CvH> AinNero depends if alpine is gpl 2 or 3 (there is no licence stated anywhere so no idea), gpl3 looks much better and more releaxed compared to gpl2 (we use)
2019-10-05 15:15:34 <CvH> afaik there is nowhere a licence stated for alpine anywhere so maybe its public domain and all problems are gone :D
2019-10-05 15:16:35 <AinNero> CvH: there is no public domain?
2019-10-05 15:16:42 <AinNero> either its licensed or not
2019-10-05 15:17:20 <CvH> yea currently alpine is not licenced, or very very well hidden :)
2019-10-05 15:19:59 <_ikke_> Alpine Linux is a distribution with many components, so you cannot really speak of the Alpine Linux license
2019-10-05 15:21:01 <AinNero> IMHO i prefer contribution and fixes over giving half-researched opinions to a group of volunteers
2019-10-05 15:22:00 <AinNero> CvH: have you provides some steps to reproduce to crash alpine-virt on >3GB ?
2019-10-05 15:22:09 <AinNero> i tried with 128GB and couldn't reproduce
2019-10-05 15:25:47 <CvH> _ikke_ ofc the alpine system to build the distro has some licence (or is not licenced), regardless of the licence that the source packages (kernel, gcc ...) are using
2019-10-05 15:29:07 <CvH> AinNero http://cvh.libreelec.tv/alpine/alpine-vm.png just with more ram
2019-10-05 15:29:59 <CvH> was vmware 15.0.3, i updated today to 15.5 - not tested if there is the same problem
2019-10-05 15:30:24 <AinNero> i was testing with qemu
2019-10-05 15:31:01 <CvH> ¯\_(ツ)_/¯
2019-10-05 15:32:11 <AinNero> can you check what is causing the panic?
2019-10-05 15:33:06 <CvH> i just get kernel panic at start so not much to show (i make a pic in a hour, need to go for now)
2019-10-05 15:33:43 <AinNero> oh
2019-10-05 15:33:48 <AinNero> VMWare is not a generic emulator
2019-10-05 15:34:03 <AinNero> CvH: Alpine is not of any of guests that VMWare supports
2019-10-05 15:34:47 <CvH> https://wiki.alpinelinux.org/wiki/Install_Alpine_on_VMware
2019-10-05 15:34:58 <AinNero> https://www.vmware.com/resources/compatibility/search.php?deviceCategory=software
2019-10-05 15:35:20 <CvH> yea but our distro is also not supported and works perfectly as well
2019-10-05 15:35:51 <AinNero> do you self-build your kernel?
2019-10-05 15:38:12 <AinNero> if you work out some fixes a kernel can be run even by VMWare, please contribute them
2019-10-05 15:40:15 <_ikke_> CvH: You have apk-tools which has a license, you have abuild, which can have a license, then you have aports, where each APKBUILD potentially can have its own license
2019-10-05 15:41:07 <Cogitri> Yet we neither have a license file nor a license header AFAICS
2019-10-05 15:41:13 <_ikke_> correct
2019-10-05 15:41:20 <_ikke_> So officially they are copyrighted to the author
2019-10-05 15:41:46 <Cogitri> Oh, that's implied then?
2019-10-05 15:42:00 <_ikke_> That's the default
2019-10-05 15:43:05 <_ikke_> The structure of an APKBUILD is pre-defined though, so I wonder how copyrightable they are
2019-10-05 15:45:38 <mps> you all owe to me some number of $money because you use letters which are invented by my ancestors and by default copyrighted by them
2019-10-05 15:46:08 <Cogitri> Hm, at least other distros license it under GPLv2 (e.g. Gentoo& Exherbo) or (some) BSD license (e.g. Void)
2019-10-05 15:47:09 <mps> ok, maybe also they are ancestors to some of you but this doesn't change fact that you owe $money to me
2019-10-05 15:47:25 <mps> sorry, couldn't resist :D
2019-10-05 15:50:09 <_ikke_> Cogitri: That does require some license transfer though
2019-10-05 15:50:37 <_ikke_> You cannot just declare that something has a certain license if you don't own the copyright
2019-10-05 15:51:01 <Cogitri> Yes, it's unrealistic to change that now
2019-10-05 15:51:06 <_ikke_> s/license transfer/copyright transfer/
2019-10-05 15:51:23 <Cogitri> Yup, would need the consent of all contributors
2019-10-05 15:51:33 <Cogitri> Which is one kind of a pain to do (see the LLVM project switching to Apache-2.0)
2019-10-05 15:51:52 <Cogitri> Maybe we could do something about it for future Aports, but I guess that'd turn into a massive mess
2019-10-05 15:52:07 <_ikke_> yup
2019-10-05 15:52:22 <_ikke_> And I think Alpines default license is MIT
2019-10-05 15:54:40 <Cogitri> MIT's nice if you don't mind the possibility of someone using your work in a proprietary project without upstreaming the changes
2019-10-05 15:55:48 <Cogitri> But I guess realistically people will upstream their changes anyway most of the time anyway since maintaining downstream patches can be kind of a pain
2019-10-05 15:58:45 <AinNero> the first bold sentence of the MIT license is my favourite
2019-10-05 16:05:11 <dalias> cogitri, my feeling is that, for many things, contributions from commercial users who don't want to work with upstream would probably be so bad you don't want them anyway
2019-10-05 16:06:11 <dalias> of course sometimes that doesn't apply, like when writing software where you'd expect commercial users to want to make interesting advancements
2019-10-05 16:56:04 <CvH> that the copyright stays at the actual copyright owner is a real danger in case the author goes south and want all his commits removed - thats ofc "possible" because there is nothing that stated any oss license at all, btw changing a version from 9.0.0 -> 9.0.1 is not copyright worthy, bigger changes are likely worth it ... its a mess :D
2019-10-05 16:56:22 <_ikke_> Yes, it is
2019-10-05 16:57:49 <CvH> AinNero http://cvh.libreelec.tv/alpine/alpine-vm2.png
2019-10-05 16:59:28 <CvH> its getting interesting, if I boot with more then 3000mb the alpine-virt-3.10.2-x86_64.iso it crashs, after installing
2019-10-05 16:59:30 <CvH> buildbot:~# free -m
2019-10-05 16:59:33 <CvH> is no problem
2019-10-05 17:00:00 <CvH> so if i boot the iso for installing first time
2019-10-05 17:01:17 <CvH> btw we at LE build crosscompile the whole distro from scratch and we carry no pathes for vm usage at all
2019-10-05 17:18:33 <AinNero> can you try with the vanilla iso?
2019-10-05 17:18:40 <AinNero> to cross-check if it makes any difference
2019-10-05 17:24:37 <AinNero> this is in code that seems independent whether we use the vanilla or the virt kconfig
2019-10-05 17:27:49 <AinNero> ah right, the Bad IP value message is unrelated
2019-10-05 17:27:56 <AinNero> init process failure
2019-10-05 17:28:11 <AinNero> CvH: this happens on the first boot after install, right?
2019-10-05 17:30:52 <AinNero> CvH: could you provide your /boot/initramfs-virt ?
2019-10-05 18:38:48 <CvH> AinNero sry, need to feed the kids, the error happens if I boot the very first time into a fresh vm - so nothing is installed etc, after i successful installed alpine with less then 3gb ram i can adjust the ram again to whatever I want
2019-10-05 18:39:48 <AinNero> tried in qemu, works fine for me
2019-10-05 18:39:55 <AinNero> both initial boot and later install
2019-10-05 18:40:08 <AinNero> i blame VMWare
2019-10-05 18:40:54 <mps> CvH: I can confirm what AinNero's says, it works quite fine with qemu
2019-10-05 18:41:25 <mps> uh, s/AinNero's/AinNero/
2019-10-05 18:42:15 <CvH> alpine-standard-3.10.2-x86_64.iso is not booting at vm due "mount boot media failed"
2019-10-05 18:42:31 <mps> even aarch64 can boot with 4GB in qemu
2019-10-05 18:43:05 <CvH> every other linux I ever tried works at vmware easily with 4gb so no idea tbh
2019-10-05 18:43:32 <mps> I don't have any experience with vmware
2019-10-05 18:44:40 <CvH> I never experienced anything "stupid" vmware doing, because this was very unexpected
2019-10-05 18:45:56 <CvH> btw can someone with wiki write access at least add some hint that you should not use more then 3000MB for vm at vmware ? this should workaround at least
2019-10-05 18:46:20 <AinNero> CvH: im really suspicious on this issue
2019-10-05 18:46:49 <CvH> does the initial boot at the installer use the same kernel then an installed system ?
2019-10-05 18:46:56 <AinNero> on the boot side, alpine doesn't use any magic that other distros dont have
2019-10-05 18:48:37 <mps> only change in virt kernel is that we increased number of CPU's for x86 and x86_64, iirc
2019-10-05 18:49:08 <AinNero> mps: the kernel config is different
2019-10-05 18:50:18 <mps> AinNero: this git commit a43744566e77df60516944e78c2953049d97a819
2019-10-05 18:50:25 <CvH> I do a compare maybe something is obviouse
2019-10-05 18:51:00 <AinNero> mps: thats like, minor? diff -u config-virt.x86_64 config-vanilla.x86_64 is 7k lines
2019-10-05 18:51:46 <mps> could be, but not much is changed
2019-10-05 18:52:46 <AinNero> half of all lines is "not much"?
2019-10-05 18:54:13 <CvH> wondering why there is an virt kernel at all  ?
2019-10-05 18:54:43 <CvH> i understand to not build dvb devices and such but overall that should be 1:1
2019-10-05 18:55:13 <AinNero> the virt doesn't have as much drivers
2019-10-05 18:55:36 <mps> AinNero: don't understand what about you talking. 'git log -p a43744566e77df60516944e78c2953049d97a819' shows not much is changed
2019-10-05 18:56:09 <AinNero> mps: diff -u config-virt.x86_64 config-vanilla.x86_64
2019-10-05 18:56:17 <AinNero> your commit is a change in virt
2019-10-05 18:56:26 <AinNero> my diff is the difference between virt and vanilla
2019-10-05 18:56:40 <CvH> btw 3.10.0 is also not working, just in case :D
2019-10-05 18:56:44 <mps> ah, no. they are really different
2019-10-05 18:57:05 <mps> changes are for virt kernels
2019-10-05 18:57:38 <mps> virt and 'native' are really different
2019-10-05 18:59:18 <CvH> a lot options that are different are looking like random y/n decisions O_o
2019-10-05 19:00:59 <CvH> also that some stuff is build as module and at vm build in doesn't sounds too reasonable <- without knowing the background and reasoning at alpine
2019-10-05 19:01:30 <mps> small, simple, secure ;)
2019-10-05 19:03:53 <CvH> yea but why is there so much difference between virt and native - i don't mean the additional support you won't need at virt
2019-10-05 19:04:24 <CvH> normally you would have the same config just with a lot less subsystems build
2019-10-05 19:04:52 <mps> if they are same what is reason to have virt at the end
2019-10-05 19:07:21 <CvH> just wondering bout those things like "CONFIG_SATA_AHCI=y" <=> "CONFIG_SATA_AHCI=m"
2019-10-05 19:07:48 <CvH> that a lot stuff is not build at virt is pretty reasonable :)
2019-10-05 19:31:21 <CvH> anyone ? : (
2019-10-05 19:31:22 <CvH> buildbot:~/aports/scripts$ ./mkimage.sh --arch armv7 --profile uboot --repository http://dl-cdn.alpinelinux.org/alpine/v3.10/main/
2019-10-05 19:31:23 <CvH> APKINDEX.6456cd1a.tar.gz: No such file or directory
2019-10-06 01:31:01 <Greyfn> I have a bug report for the fancontrol package, but I want to be sure before I submit it - has anybody else used this package sucessfully?
2019-10-06 01:31:36 <Greyfn> it seems that the package uses echo to write PWM values to the fans, and expects "echo -n" behaviour by default - which Alpine doesn't do by default
2019-10-06 01:32:06 <Greyfn> if you change all instances of echo in the fancontrol script to echo -n, then it works just fine
2019-10-06 01:54:23 <Greyfn> I'm fairly convinced. First ever bug report raised: https://gitlab.alpinelinux.org/alpine/aports/issues/10854
2019-10-06 04:17:35 <crockwork> Does the current nmap version not work with scripts? I swear I've tried multiple times and not had any output.
2019-10-06 05:12:29 <crockwork> I'm closer.
2019-10-06 05:12:30 <crockwork> Hmm.
2019-10-06 08:26:46 <CvH> mps I was now able to create ">>> mkimage-armv7: Creating alpine-uboot-191005-armv7.tar.gz" ... (i guess I need to write an howto about it or i forgot everything very soon) , do you have your changes for allwinner somewhere ? (reboot fix + usb enabled at kernel.conf)
2019-10-06 08:27:36 <CvH> or at least could you provide some *.patch/*.diff file ?
2019-10-06 08:29:04 <CvH> I have an H2+ and an H5 board here and a Alwinner kernel dev at hand if another board needs testing
2019-10-06 08:35:34 <mps> CvH: good news :), but I'm outside of home now and on moving computer
2019-10-06 08:36:23 <mps> in a few hours I will be back and I will post you patch for linux-vanilla aports
2019-10-06 09:52:56 <CvH> mps tx
2019-10-06 11:20:34 <mps> CvH: are you here?
2019-10-06 11:25:02 <CvH> mps es
2019-10-06 11:25:03 <CvH> yes
2019-10-06 11:30:42 <mps> is it ok to post patch over tpaste.us to you
2019-10-06 11:31:18 <mps> don't know will tpaste accept big file
2019-10-06 11:34:11 <mps> CvH: it is accepted, here it is http://tpaste.us/NKbP
2019-10-06 11:35:22 <CvH> "cat file.ext | curl -F 'f:1=<-' ix.io" works for big files too :)
2019-10-06 11:35:28 <CvH> tx, i give it a try
2019-10-06 11:36:18 <CvH> is it basically armhf kernel.conf with some changes ?
2019-10-06 11:36:37 <mps> eh, I donwloaded ix.io client (shell script) but never tried
2019-10-06 11:36:55 <mps> no, it is intended for armv7 SOC's
2019-10-06 11:37:13 <mps> I suspect it will work on armv6 (armhf)
2019-10-06 11:37:19 <CvH> yea, just did you do it from scratch or is it based ontop of armhf ?
2019-10-06 11:37:55 <CvH> btw for ix.io you just need curl nothing else :)
2019-10-06 11:38:18 <mps> no, I removed symlink for armv7 which links it to armhf config and created separate config for armv7
2019-10-06 11:38:41 <CvH> oki, i report back as soon i know something
2019-10-06 11:38:49 <CvH> (this evening i hope)
2019-10-06 11:39:05 <mps> yes, I looked at script which can be added to shell as function
2019-10-06 11:39:48 <mps> I tested it only in qemu and on one real board (two, but similar)
2019-10-06 11:40:25 <mps> most things requested over IRC or bug reports works
2019-10-06 11:41:31 <CvH> kids are calling, I report back hopefully with a success message
2019-10-06 11:41:58 <mps> care for kids, this is most important in our life :) see you
2019-10-06 11:43:40 <_ikke_> Just a little too late :)
2019-10-06 11:44:45 <mps> _ikke_: playing with words: grandkids ;)
2019-10-06 11:45:49 <mps> and who knows, nothing is too late in life :P
2019-10-06 11:46:11 <_ikke_> I mean, they just left the channel before you said that
2019-10-06 11:47:33 <mps> ah, why people don't keep their connections open to IRC, can't understand. they can miss some msgs which are useful or important to them
2019-10-06 12:57:10 <la_ensalada> hi I'm new to alpine and I'm having some trouble with configuring networking. my installation didn't come with an /etc/network/interfaces file and after making one it doesn't seem to get used
2019-10-06 12:57:35 <la_ensalada> ultimately I want to use interfaces to define a virtual interface. help?
2019-10-06 12:57:46 <AinNero> how did you install it? because setup-alpine calls setup-interfaces
2019-10-06 12:58:13 <la_ensalada> AinNero, it's a docker image so I didn't install it
2019-10-06 12:58:49 <AinNero> when using docker, you configure networking from the host system
2019-10-06 12:59:30 <la_ensalada> ...oh
2019-10-06 13:00:03 <AinNero> docker networks is what you are looking for
2019-10-06 13:00:13 <AinNero> or docker links, but they have been deprecated
2019-10-06 13:04:14 <la_ensalada> AinNero, thank you, I'll start pulling articles
2019-10-06 14:49:43 <mckaygerhard> hi! someone know about a build service&repository that provies support for alpine like OBS suse build service provides for many distros or ppa provides for winbuntu ?
2019-10-06 14:51:05 <Cogitri> Alpine's repositories
2019-10-06 14:54:20 <AinNero> with alpine, generating packages
2019-10-06 14:54:27 <AinNero> ... isn't that difficult
2019-10-06 14:54:40 <AinNero> and hosting them as repository somewhere
2019-10-06 14:59:53 <Cogitri> Or that, but if possible just get them into the repo, it's really not that hard
2019-10-06 14:59:57 <mckaygerhard> alpine repositories with the git apports are huge in every clone operation.. that's are not good NONE for people that dont have enought network connection
2019-10-06 15:00:23 <mckaygerhard> also all the pulls must be acepted to see compilation.. it's a problem
2019-10-06 15:00:36 <mckaygerhard> in obs i can see my compilation without any aceptation of
2019-10-06 15:00:56 <AinNero> mckaygerhard: you can compile locally and publish your local ~/packages via rsync to your own server
2019-10-06 15:01:21 <AinNero> such things exist: https://github.com/c3d2/aports
2019-10-06 15:01:29 <AinNero> alpine-independent aport trees
2019-10-06 15:01:36 <Cogitri> mckaygehrard: People don't have to download all of aports though, you can just download the APKBUILD by itself if you want to
2019-10-06 15:02:19 <Cogitri> Also, adding lots of repos for different packages is annoying and they're hard to discover and their quality usually sucks and they're not trusted
2019-10-06 15:02:28 <Cogitri> We try our best merging PRs somewhat timely, but they're required for QA
2019-10-06 15:02:34 <AinNero> until you explicitly trust them :P
2019-10-06 15:02:40 <Cogitri> You get stuff like CI and reviews for free in return
2019-10-06 15:03:42 <mckaygerhard> AinNero: i will explain you.. in a atom laptop i cannot build the chromiun browser okey! what a nonsense from you
2019-10-06 15:04:20 <Cogitri> AinNero: And trusting random people to provide binary packages is pretty bad in my books
2019-10-06 15:04:33 <mckaygerhard> Cogitri: i cannod to that due if a  must build a newer version of qt5 libs my poor machine will break
2019-10-06 15:05:41 <mckaygerhard> Open suse build service provides a very good service.. but i cannot see the alpine support there.. many people in telegram said that cannot contribute due that, those persons dont have enought resources to build by won a heavy package
2019-10-06 15:05:45 <AinNero> mckaygerhard: im using thinkpads from 2010. whats your point here?
2019-10-06 15:06:30 <AinNero> either you arrange with waiting for the PR's or you get your own build servers
2019-10-06 15:06:31 <mckaygerhard> dyou said "thinkpads" not "thinkpad" i knowed peole that only have one machine and that machine does not have enoguht free space or power..
2019-10-06 15:07:11 <mckaygerhard> if alpoine does not have enought service like OBS contributions to alpine will be always poor (and we need so much that)
2019-10-06 15:07:23 <mckaygerhard> now you have the point AinNero ?
2019-10-06 15:07:50 <AinNero> mckaygerhard: year, you are expecting us to supply computing power to you for free
2019-10-06 15:08:41 <mckaygerhard> AinNero: no.. due alpine community of alpine organization does not have it enought resources but what are need if i have people that can provide powered machine for ?
2019-10-06 15:09:04 <mckaygerhard> someone from alpine infra that can be sense about THAT problem
2019-10-06 15:09:05 <mckaygerhard> ?
2019-10-06 15:12:36 <AinNero> mckaygerhard: if you have people that can provide powerful machines, ask them so you can use them to build your alpine packages
2019-10-06 15:14:13 <AinNero> so you dont need to make your poor laptop suffer from it
2019-10-06 15:16:44 <AinNero> i have a powerful machine in my attic and i an ssh to it, but i rarely need to
2019-10-06 15:16:59 <AinNero> since a decade old laptops are sufficient to build almost everything
2019-10-06 15:17:45 <AinNero> firefox and chromium might be worst offenders on build requirements, though
2019-10-06 15:18:14 <maldridge> nah, libreoffice
2019-10-06 15:18:47 <AinNero> libreoffice is it! or has someone something worse?
2019-10-06 15:19:11 <Cogitri> Nah Electron :P
2019-10-06 15:20:40 <AinNero> Electron is chrome, essentially, right?
2019-10-06 15:20:46 <Cogitri> mckaygehrard: I don't understand your problem, if you submit a PR you don't have to build the heavy deps
2019-10-06 15:21:36 <Cogitri> Chrome, but you can't build from tarballs because that's too easy
2019-10-06 15:21:56 <Cogitri> So it's Chromium, but with fetching at least 50GB of source code via git
2019-10-06 15:28:45 <mckaygerhard> AinNero: that's not only for me.. we need a way to most user that cannot conribute can do that! as other distributions can do..
2019-10-06 15:29:38 <Cogitri> Seeing how low quality those packages usually are I don't think we need that
2019-10-06 15:30:18 <Cogitri> The contribution requirements for Alpine really aren't too strict, just git commit and follow some basic QA and you're all set
2019-10-06 15:30:23 <mps> I built rust and firefox on arm32 with 4GB RAM and FS on microSD card about 9 months ago. takes time but finished
2019-10-06 15:33:15 <mps> and agree with Cogitri, we need more QA than new packages
2019-10-06 15:38:09 <mckaygerhard> Cogitri: low cuality package.. can i cited the amount of problems for desktop mades ?
2019-10-06 15:38:55 <Cogitri> Pardon?
2019-10-06 18:52:00 <CvH> mps is it normal that compiling arm under qemu is HORRIBLE slow ?
2019-10-06 19:00:49 <mps> CvH: depends what you compile and how you run qemu
2019-10-06 19:01:22 <mps> and what is host for qemu arm
2019-10-06 19:03:05 <mps> anyone here know how to pass '-fPIC' for CMAKE_CXX_FLAGS with cmake build
2019-10-06 19:03:31 <mps> looks like '-DCMAKE_CXX_FLAGS="-fPIC"' doesn't work
2019-10-06 19:07:03 <CvH> mps i just compile the kernel for now :)  I run qemu in a VM with ubuntu + https://github.com/alpinelinux/alpine-chroot-install
2019-10-06 19:07:54 <CvH> normally I build an arm build of libreelec incl every single package from scratch in under 3h, crosscompile ofc and not qemu
2019-10-06 19:08:47 <mps> what is libreelec
2019-10-06 19:11:44 <CvH> ah sry ;D  its the distro i am maintaining normally, we build an complete linux + drivers + tools and stuff to run Kodi (an Mediacenter Software) - we support all kind of ARM hardware (Raspberry, Alwinner ...), but we cross compile every package
2019-10-06 19:12:41 <CvH> normally building an kernel takes ~10minutes, currently it takes 4 horus and still not finished
2019-10-06 20:41:02 <kmmndr> Hi all :-)
2019-10-06 20:49:42 <Cogitri> Heya
2019-10-07 06:37:25 <shrizza> ncopa: I noticed "trace bitmap" doesn't work in Inkscape. Ever gotten any feedback on this before?
2019-10-07 06:39:32 <shrizza> ("trace bitmap" I think normally brings up a panel or window with options, but nothing appears)
2019-10-07 06:40:57 <shrizza> No errors or anything in console output when executing from a terminal.
2019-10-07 07:16:20 <CvH> @mps compiling worked, but it is horrible slow :(  50 times slower then cross compiling a kernel, even if I build the kernel at an low end RPi3 it is 7 times faster , doing a test later
2019-10-07 07:17:40 <mps> CvH: first, don't put @ as prefix in nick on irc, this is for channel OP's
2019-10-07 07:18:15 <CvH> it works at webchat for freenode :)
2019-10-07 07:18:32 <TBB> it's a twitterism isn't it
2019-10-07 07:18:39 <mps> you can ask _ikke_ or clandmeter how they set qemu to work faster
2019-10-07 07:19:25 <CvH> yea won't complain if it get fast to at least reach a lowend arm device level
2019-10-07 07:19:35 <mps> '@' doesn't trigger highlight in my irc client, irssi
2019-10-07 07:19:51 <clandmeter> compiling with emulation will always be slow
2019-10-07 07:20:07 <CvH> 50 times slower ?
2019-10-07 07:20:19 <clandmeter> i dont know the X
2019-10-07 07:20:21 <clandmeter> but a lot yes
2019-10-07 07:20:50 <clandmeter> cross compiling is much faster
2019-10-07 07:20:52 <mps> CvH: also, how many CPU you set for qemu
2019-10-07 07:20:57 <CvH> i basically just wan't to compile alpine for arm, crosscompiling is not really working (or i am doing it wrong .. idk)
2019-10-07 07:21:09 <clandmeter> but it introduces issues
2019-10-07 07:21:38 <clandmeter> CvH: you want to compile alpine for arm?
2019-10-07 07:21:41 <clandmeter> what does that mean?
2019-10-07 07:21:43 <mps> most safe option is native compile
2019-10-07 07:22:16 <mps> even qemu can introduce issue
2019-10-07 07:24:16 <mps> CvH: does librelec project can get better arm machine for builders
2019-10-07 07:24:35 <mps> s/does/could/
2019-10-07 07:25:01 <CvH> yea but not a single kernel dev I know uses native compiling for arm in 99% of the time so I consider crosscompiling a proper approach :)  not sure about qemu
2019-10-07 07:25:26 <CvH> we do 100% crosscompiling for all archs and devices
2019-10-07 07:27:16 <CvH> clandmeter, I try to build a kernel from source :D  its pretty difficult tbh, nothing at the wiki and all other info is wide spread, wrong or broken :(
2019-10-07 07:27:20 <mps> nowadays you can get arm box with 4GB RAM and 4 CPU's for not big money
2019-10-07 07:27:41 <TBB> more like ridiculously cheap, mps :)
2019-10-07 07:28:50 <mps> TBB: yes, but 'ridiculously cheap' could be quite different for different projects/people
2019-10-07 07:28:53 <CvH> yea but why using native building if cross compiling works perfectly :)  building nativly at arm is much much slower
2019-10-07 07:29:17 <TBB> cross compiling used to be troublesome
2019-10-07 07:29:22 <TBB> it isn't anymore
2019-10-07 07:29:51 <clandmeter> CvH: you are using the wrong machine to build native
2019-10-07 07:30:46 <clandmeter> our builders are more than fast enough.
2019-10-07 07:31:19 <clandmeter> im not sure about other distros, but i dont think that serious ones do cross compiling.
2019-10-07 07:32:24 <mps> CvH: ok, does the patch I posted to you work
2019-10-07 07:32:28 <TBB> on the other hand, embedded projects do
2019-10-07 07:32:50 <mps> i.e. kernel you build using it works?
2019-10-07 07:32:53 <clandmeter> yes most embedded ones do.
2019-10-07 07:33:49 <clandmeter> and some arch for which you cannot get decent hw (mips?) you will want to cross compile.
2019-10-07 07:35:12 <clandmeter> there are 3 ways to get better performance on real hardware.
2019-10-07 07:36:24 <clandmeter> 1. get a sponsor. 2. get a ARM based vps (relatively cheap) 3. use CI which support ARM (cloud.drone.io)
2019-10-07 07:37:09 <mps> 4. buy super arm box
2019-10-07 07:37:43 <clandmeter> im talking about possible solutions :)
2019-10-07 07:37:58 <clandmeter> for open source projects that dont earn that kind of money.
2019-10-07 07:38:05 <CvH> clandmeter yea i have other "better" arm hw here but actually not even close to my normal build machine ;)  but for alpine and compiling a single package it might be good enough
2019-10-07 07:38:26 <yunfan> clandmeter: why arm's were cheaper?
2019-10-07 07:39:38 <clandmeter> its not that difficult to setup a pipeline on drone.io and let it build on push and push the artifacts somewhere (some local minio)
2019-10-07 07:40:03 <clandmeter> yunfan: ?
2019-10-07 07:53:44 <CvH> basically everyone in the embeded world uses crosscompiling so you just need the native approach if your buildsystem don't handle crosscompile
2019-10-07 07:54:23 <clandmeter> heh, thats another way to look at it.
2019-10-07 08:01:55 <yunfan> clandmeter: i think i had misunderstanded you
2019-10-07 08:11:56 <mps> CvH: could you tell does built kernel works
2019-10-07 08:15:41 <CvH> mps i don't know yet, need to do other work first :/  i report back, no fear
2019-10-07 08:17:32 <mps> ok, np. but don't forget to report
2019-10-07 08:39:10 <midasi> when I try to build a package with 'abuild -r' I always receive the error message 'ERROR: Not committing changes due to missing repository tags. Use --force-broken-world to override.'
2019-10-07 08:40:32 <_ikke_> Try apk fix
2019-10-07 08:40:53 <_ikke_> Might be that you added some pinned packages and later removed the pinned repositories
2019-10-07 08:42:53 <midasi> yes, you're right, that was exactly the case!
2019-10-07 08:42:57 <midasi> thank you
2019-10-07 09:51:03 <clandmeter> anyone here using multipath?
2019-10-07 09:59:09 <misterunknown> clandmeter: Not with alpine... We have several Debian servers with multipath to manage redundant FC connections.
2019-10-07 09:59:57 <clandmeter> right, im trying to make it work on alpine
2019-10-07 10:00:15 <clandmeter> but im not very successful
2019-10-07 10:00:59 <misterunknown> Too many errors? Or no errors and just not working?
2019-10-07 10:01:43 <clandmeter> it doesnt create any devices in /dev/mapper
2019-10-07 10:02:10 <clandmeter> im trying to use it with open-iscsi
2019-10-07 10:03:27 <misterunknown> I'm absolutely not an expert. The dm_multipath kernel module is loaded correctly?
2019-10-07 10:03:35 <clandmeter> yes
2019-10-07 10:03:48 <clandmeter> im not sure what is responcible for creating the devices
2019-10-07 10:03:52 <clandmeter> i guess normally its udev
2019-10-07 10:04:10 <clandmeter> which is not installed (by default)
2019-10-07 10:05:06 <mps> more than 10 years ago I had to setup and test it on one IBM 'monster', but I forgot much about it
2019-10-07 10:07:13 <mps> that was on RHEL, and I think kernel loaded driver automatically and then I used cli tool - multipath iirc
2019-10-07 10:07:22 <mps> for fine tunning
2019-10-07 10:07:24 <misterunknown> clandmeter: Isn't it the kernel device mapper which creates these devices?
2019-10-07 10:09:12 <clandmeter> i have no idea.
2019-10-07 10:09:17 <mps> there are FC drivers for particular boards I think
2019-10-07 10:09:53 <clandmeter> misterunknown: i have no clue how its suppose to work
2019-10-07 10:10:36 <clandmeter> i need to use it to mount my volume on packet.net
2019-10-07 10:10:48 <clandmeter> they provide some scripts, but it seems they are kind of broken on alpine.
2019-10-07 10:12:06 <clandmeter> it looks like iscsi is creating two sdX devices based on two ip's which are provided.
2019-10-07 10:12:29 <clandmeter> how multipath handles those i dont know.
2019-10-07 10:14:41 <mps> iirc, it assign them as members to another 'master' device
2019-10-07 10:15:24 <mps> (but before talking more should refresh my memory)
2019-10-07 10:15:33 <clandmeter> this would be my idea as well, but i dont see it.
2019-10-07 10:16:02 <misterunknown> clandmeter: Maybe you could use "dmsetup create ..." from these sdX devices.
2019-10-07 10:16:16 <mps> if I saved my notes and script :(
2019-10-07 10:16:30 <misterunknown> s/ /to create a multipath device/
2019-10-07 10:19:11 <misterunknown> Btw, I'm actually here to report success on upgrading my alpine lxc-host to 3.10. Worked without any issues.
2019-10-07 10:34:14 <shrizza> Can anyone explain how "abuild -r" is supposed to work? I apk del'd all the dependencies for the APKBUILD I'm working on, but when I issue abuild -r it complains about "builddeps failed".
2019-10-07 10:34:41 <shrizza> I thought it's supposed to just install and uninstall dependencies around the apk building process.
2019-10-07 10:34:49 <_ikke_> shrizza: It should install all required dependencies, given that they are available in the repositories
2019-10-07 10:34:55 <_ikke_> yes, it does
2019-10-07 10:35:14 <_ikke_> shrizza: Can you paste the entire error / log output somewhereA
2019-10-07 10:39:40 <TBB> that's how it works, it installs a virtual package, a metapackage for all build dependencies, which it then removes at the end of build. and you should always use a barebones environment for creating the abuild so that you'll have to specify all of the deps in the APKBUILD file
2019-10-07 10:40:31 <TBB> in that sense building on top of a minimal lxc container, for example, is the best approach for building apk packages
2019-10-07 10:41:02 <_ikke_> Indeed, so that you know that you covered all dependencies and don't rely on something already present in your environment
2019-10-07 10:42:32 <shrizza> _ikke_: https://pastebin.com/2qJ0WsMq
2019-10-07 10:42:59 <shrizza> Yeah, this is a quite fresh copy of Alpine that I'm using only to test these APKBUILDs I'm writing.
2019-10-07 10:43:10 <clandmeter> did you try apk fix?
2019-10-07 10:43:22 <shrizza> No, what does that do?
2019-10-07 10:43:28 <clandmeter> fix things
2019-10-07 10:45:24 <shrizza> For the record, abuild alone works fine when I have all the dependencies I need installed.
2019-10-07 10:45:48 <shrizza> But yeah, I wanted to emulate installing on a somewhat fresh install.
2019-10-07 10:45:51 <clandmeter> so did `apk fix` fix things for you?
2019-10-07 10:46:01 <shrizza> Let me try that.
2019-10-07 10:46:38 <shrizza> OK, I ran it.
2019-10-07 10:46:43 <_ikke_> What did it return?
2019-10-07 10:47:39 <shrizza> https://pastebin.com/VeDPX5Fs
2019-10-07 10:47:50 <shrizza> I tried the abuild -r again afterward, but I still get the same error.
2019-10-07 10:47:55 <_ikke_> Oh, that issue
2019-10-07 10:48:27 <clandmeter> can you remove one of them?
2019-10-07 10:48:33 <clandmeter> or are they both deps?
2019-10-07 10:48:56 <_ikke_> https://gitlab.alpinelinux.org/alpine/aports/issues/10716
2019-10-07 10:48:59 <clandmeter> i think we have an issue for that
2019-10-07 10:49:12 <TBB> yup, that conflict alone is enough to abort the build
2019-10-07 10:49:28 <clandmeter> the problem is that abuild hides it
2019-10-07 10:49:59 <TBB> yup, you kind have to know this (or really only build in a minimal build container where things like these don't happen)
2019-10-07 10:50:05 <shrizza> Oh, this is actually stalling the abuild -r?
2019-10-07 10:50:08 <_ikke_> yes, deps "--quiet --simulate" || return 1
2019-10-07 10:50:12 <_ikke_> shrizza: it is
2019-10-07 10:50:24 <shrizza> Ahhh...
2019-10-07 10:50:27 <TBB> yes, it makes installing make depends return an error code which then aborts abuild
2019-10-07 10:50:38 <clandmeter> it should add an error message
2019-10-07 10:50:46 <clandmeter> failed to install deps or whatever
2019-10-07 10:51:07 <TBB> I would probably be a good idea to push abuild to a container-only build direction (and there already is some support for that)
2019-10-07 10:51:09 <shrizza> Well, I do get an error "builddeps failed".
2019-10-07 10:51:16 <_ikke_> TBB: -1
2019-10-07 10:51:27 <TBB> yeah, it just doesn't tell you why builddeps failed, that's not very useful
2019-10-07 10:51:42 <_ikke_> abuild should just be a generic tool
2019-10-07 10:51:57 <_ikke_> (and ncopa wants to split some functionality out of it)
2019-10-07 10:52:02 <TBB> I actually do all my abuilds exactly that way
2019-10-07 10:52:14 <clandmeter> we already have dockerized abuild
2019-10-07 10:52:35 <_ikke_> docker-abuild
2019-10-07 10:53:01 <shrizza> I've never used docker or containers so I'll have to research that.
2019-10-07 10:53:12 <clandmeter> it wont fix your issue
2019-10-07 10:53:22 <_ikke_> shrizza: what does apk policy hwdata return?
2019-10-07 10:53:24 <clandmeter> not if its pulled both in by deps
2019-10-07 10:54:07 <mps> hwadata is removed from edge
2019-10-07 10:54:10 <shrizza> https://pastebin.com/9LLFLqBT
2019-10-07 10:54:21 <mps> hwdata*
2019-10-07 10:54:30 <_ikke_> mps: Yes, I noticed in that issue
2019-10-07 10:54:35 <_ikke_> Might want to backport it to 3.10
2019-10-07 10:54:41 <_ikke_> Not sure if that would create issues
2019-10-07 10:55:02 <mps> _ikke_: you mean backport hwids
2019-10-07 10:55:48 <shrizza> BTW, I'm not in any hurry with this APKBUILD so please take your time.
2019-10-07 10:56:19 <_ikke_> shrizza: I hope that wasn't meant sarcastic
2019-10-07 10:56:23 <mps> if you backport hwids which provides hwdata it could make issue with few pkg's, chromium and two or three more
2019-10-07 10:56:33 <shrizza> I'm just leisurely learning to build packages and I didn't realize the abuild -r thing would be this deep.
2019-10-07 10:56:40 <shrizza> _ikke_: No.
2019-10-07 10:56:46 <_ikke_> shrizza: ok, good
2019-10-07 10:57:07 <shrizza> (these packages are just games)
2019-10-07 10:58:33 <misterunknown> shrizza: Oha... Which games? :)
2019-10-07 10:59:05 <clandmeter> shrizza: did you install any of the pci pkgs manually?
2019-10-07 10:59:05 <shrizza> I know how rabbit holes can be so I sincerely didn't want to compel you guys into doing something that could take up your time.
2019-10-07 10:59:28 <shrizza> clandmeter: AFAIK, no.
2019-10-07 10:59:44 <shrizza> The hwdata package does not ring a bell.
2019-10-07 11:00:04 <clandmeter> do you depend on one of them?
2019-10-07 11:00:23 <clandmeter> i guess pciutils
2019-10-07 11:02:22 <shrizza> Hm, well "apk info -a hwdata" shows "hwdata-0.318-r0 provides:" with no dependencies it seems.
2019-10-07 11:02:56 <shrizza> misterunknown: Shiromino (https://github.com/FelicityVi/shiromino) and Avara (https://github.com/avaraline/Avara).
2019-10-07 11:04:55 <shrizza> I just went ahead and apk del pciutils.
2019-10-07 11:05:12 <shrizza> And it removed hwids-pci along with it.
2019-10-07 11:05:29 <shrizza> Nice, seems to be working now. Thanks!
2019-10-07 11:23:42 <shrizza> ls -l
2019-10-07 11:24:13 <_ikke_> .
2019-10-07 11:24:15 <_ikke_> ..
2019-10-07 12:18:00 <misterunknown> Quick question: Are there any plans to switch to kernel 5.x?
2019-10-07 12:18:30 <misterunknown> (I think so, but is there a timeline?)
2019-10-07 12:19:01 <mps> misterunknown: when the 5.4 released
2019-10-07 12:19:14 <_ikke_> mps: alpine follows LTS kernels
2019-10-07 12:19:17 <_ikke_> misterunknown: *
2019-10-07 12:20:49 <mps> maybe we should this time start with -rc{4,5} to fix potential issues before alpine 3.11 release
2019-10-07 12:21:37 <mps> 5.4-rc2 is released yesterday which means stable will be in 6 weeks
2019-10-07 12:24:53 <mps> btw, I'm thinking to make linux-mp (multiplatform) for armv7 (for now) in testing with newer releases, but not sure if this a good idea
2019-10-07 12:26:19 <mps> lot of people ask for such thing and I run it on few arm boxes long time
2019-10-07 12:27:22 <mps> that will save me some time to not send it everytime someone ask
2019-10-07 12:28:19 <midasi> I'm playing around with diskless installation. I would like to have /var on a disk while the system is loaded from a readonly media (= setup-disk type 'data'). Now, I have the issue that the disk is already getting mounted on startup to /media (i.e. /dev/xvda2 -> /media/xvda2) and I'm not able to mount it to /var in fstab since it's already mounted. How is this meant to work?
2019-10-07 12:32:31 <AinNero> midasi: /var and /boot in the same place?
2019-10-07 12:33:11 <AinNero> hm, /var and the actual boot partition in the same place
2019-10-07 12:33:51 <midasi> AinNero: /boot and the system is on a read-only media... apkovl and /var on a readwrite disk
2019-10-07 12:37:30 <nepugia> If the problem is that it is mounted ahead of time you may need to change the initramfs script to do what you actually want instead
2019-10-07 12:38:12 <midasi> nepugia: yes, that's clear... I thought it would work out of the box
2019-10-07 12:39:03 <nepugia> Well, you have tested it, and it seams not to :)
2019-10-07 12:39:05 <midasi> perhaps the issue is that apkovl is on the same partition as /var. Due to that it's not possible to remount the partition
2019-10-07 12:39:55 <fredrigu> TBB: let's not make abuild any slower than we've to
2019-10-07 12:41:01 <hightower2> Hey folks, does alpine have binary package upgrades? And if so, how did you solve the issue of musl and it's dynamic linker not having symbol versioning?
2019-10-07 12:42:42 <nepugia> Uhh, do you mean something else with binary package upgrades than the package manager already does?
2019-10-07 12:44:20 <hightower2> no, no, I mean just regular package upgrades
2019-10-07 12:44:34 <hightower2> how do they work in absence of symbol versioning in musl, and/or how you solved it
2019-10-07 12:57:06 <fredrigu> hightower2: you mean that musl is missing soname?
2019-10-07 13:00:38 <hightower2> I mean the symbol/function versions like event_strlcpy_@Base 2.1.8-stable etc.
2019-10-07 13:02:47 <fredrigu> hightower2: that's out of my expertise but it sounds like something that _ikke_ might has a look at
2019-10-07 13:03:11 <fredrigu> *had
2019-10-07 13:03:28 <_ikke_> This is also kind of out of my expertise :)
2019-10-07 13:04:19 <_ikke_> But it would be expected that if you change the ABI, the soname changes as well
2019-10-07 13:05:06 <fredrigu> hightower2: I'm going to take the chance to learn something new here. How does this work with glibc?
2019-10-07 13:13:29 <hightower2> I assume its runtime linker supports it
2019-10-07 13:15:13 <hightower2> https://sourceware.org/glibc/wiki/FAQ#What_is_symbol_versioning_good_for.3F__Do_I_need_it.3F
2019-10-07 13:20:19 <PICCORO> hey guys.. any web panel that can be installed on alpine to manage hosting services?
2019-10-07 13:20:51 <_ikke_> hightower2: ah, so it's not about the ABI perse, but about different implentations for the same interface
2019-10-07 13:31:26 <misterunknown> PICCORO: Not to my knowledge. The common web GUIs don't support alpine yet, AFAIK.
2019-10-07 13:31:57 <misterunknown> I just checked ispconfig, cpanel and webmin. None of them honor alpine atm.
2019-10-07 13:32:22 <misterunknown> PICCORO: Why would you want to use such a thing?
2019-10-07 13:32:30 <fredrigu> hightower2: I see, thanks! So it should be a non issue if you just recompile all programs that depends on musl, right? The only thing this is good for is to update glibc (with in incompatible new version) without recompiling packages that depends on glibc
2019-10-07 14:02:11 <hightower2> fredrigu, so alpine solves it by ignoring the issue?
2019-10-07 14:04:13 <mps> hightower2: alpine is distribution with not much developers
2019-10-07 14:05:16 <mps> we do our best to not break things and fix issues
2019-10-07 14:05:58 <mps> but we cannot solve all 'computer world' issues
2019-10-07 14:07:03 <AinNero> hightower2: if you want symbol versioning, then you should use a glibc based distro
2019-10-07 14:07:31 <AinNero> if you need (working) static linking or small binary sizes, you are right with alpine
2019-10-07 14:08:54 <hightower2> don't get me wrong, I didn't mean to imply any solution was bad
2019-10-07 14:09:05 <nepugia> Hmm, i wonder if alpine will get a nightly rust, or if the patches might get upstreamed to rust in some way or another
2019-10-07 14:09:15 <hightower2> I am just genuinely interested in how you did it, not judging any particular approach
2019-10-07 14:10:00 <AinNero> Alpine is downstreaming musl, so you'd better ask in #musl
2019-10-07 14:10:49 <hightower2> yes but I'm asking here because I already know about musl (musl does not implement symbol versioning itself, in any form... either in musl or in its dynamic linker)
2019-10-07 14:11:49 <hightower2> so, does your approach sometimes/periodically require you to "rebuild world"?
2019-10-07 14:12:45 <AinNero> why should it?
2019-10-07 14:13:30 <hightower2>  I'd expect breakage to happen from time to time, requiring you to do so
2019-10-07 14:13:48 <hightower2> (with no versioning, that is)
2019-10-07 14:13:53 <_ikke_> depends on the stability of musl
2019-10-07 14:14:02 <AinNero> musl has interface versioning
2019-10-07 14:14:22 <AinNero> but until now, it stayed on *.so.1
2019-10-07 14:15:09 <AinNero> the version number in the soname is expected to increase if a breaking change occurs
2019-10-07 14:21:00 <hightower2> Cool, ok great, thank you for input
2019-10-07 14:22:00 <hightower2> I'll need a musl env soon so might try with alpine. I'd use an aarch64 image, do you know offhand if I'd work on Asus TP370QL <AinNero> the version number in the soname is expected to increase if a breaking change occurs?
2019-10-07 14:22:26 <hightower2> I'll need a musl env soon so might try with alpine. I'd use an aarch64 image, do you know offhand if I'd work on Asus TP370Q?
2019-10-07 14:26:28 <mps> is that arm64
2019-10-07 14:28:13 <mps> alpine works on my two arm64 chromebooks about two years now
2019-10-07 14:29:10 <mps> and on one arm32 chromebook, which now mostly keeping dust
2019-10-07 14:30:23 <nepugia> that's kinda sad :)
2019-10-07 14:31:43 <mps> well, I clean it from time to time when have to build something locally needed for my SBC's arm32 boxex
2019-10-07 14:32:17 <mps> boxes*
2019-10-07 14:34:56 <hightower2> mps,  yes arm64, nice thanks
2019-10-07 14:44:36 <markand> I'm surprised to see that almost nobody implement pax, neither busybox or coreutils
2019-10-07 14:50:39 <hiro> good morning
2019-10-07 14:51:20 <hiro> i wonder if i'm doing something wrong, i am unable to use gdb here...
2019-10-07 14:51:22 <hiro> received signal SIGSEGV, Segmentation fault.
2019-10-07 14:51:22 <hiro> [Switching to LWP 7295]
2019-10-07 14:51:22 <hiro> 0x00007ffff7fb7041 in ?? () from /lib/ld-musl-x86_64.so.1
2019-10-07 14:51:22 <hiro> #0  0x00007ffff7fb7041 in ?? () from /lib/ld-musl-x86_64.so.1
2019-10-07 14:51:22 <hiro> (gdb) bt
2019-10-07 14:51:22 <hiro> #1  0x0000000000000000 in ?? ()
2019-10-07 14:52:05 <hiro> is there anything special i need to do to install some unstripped libraries, debug symbols, etc. ?
2019-10-07 14:59:30 <dalias> hiro, install the -dbg packages for the app/libs
2019-10-07 14:59:38 <dalias> for libc that would be musl-dbg
2019-10-07 15:01:35 <hiro> dalias: thank you so much. i installed a bunch of -dbg but missed musl!
2019-10-07 15:01:44 <hiro> dalias: somehow i searched for libc and not musl ;)
2019-10-07 15:01:55 <hiro> dalias: everything showing up fine now
2019-10-07 15:10:10 <dalias> yay
2019-10-07 15:10:16 <dalias> glad to help
2019-10-07 15:15:59 <fredrigu> hightower2: thanks for the link. I don't have an answear for you. But my understanding is that alpine wholy depends on soname and that sometimes breaks
2019-10-07 15:40:18 <TBB> fredrigu: I don't mind it the way it is, I can in any case setup a bare build root in about 0.008 seconds and do my builds there without the issues that were talked about earlier
2019-10-07 16:59:28 <PICCORO> misterunknown: so due i need to sell services to clients .. and clients don are epxerts in alpine right?
2019-10-07 16:59:28 <PICCORO> that's why i need any web panel that can be installed on alpine to manage hosting services?
2019-10-07 17:21:15 <_ikke_> PICCORO: look for ACF
2019-10-07 17:40:20 <mps> dalias: did you solved your microSD card issue
2019-10-07 17:43:19 <mps> to me it didn't appeared with kernel 5.3.4
2019-10-07 17:44:42 <mps> ohm, 5.3.5 is released
2019-10-07 17:47:18 <mps> these syztools will force few releases every day
2019-10-07 18:14:35 <dalias> mps, i haven't updated kernel in a while
2019-10-07 18:15:06 <dalias> alpine doesn't have 5.3 packaged does it?
2019-10-07 18:15:32 <mps> no, it doesn't, I build new ones for me
2019-10-07 18:15:38 <Cogitri> Nop
2019-10-07 18:15:57 <Cogitri> I guess we should get a 5.4rc soon-ish to get 5.4 into Alpine 3.11 though
2019-10-07 18:16:33 <mps> Cogitri: you reading my mind, I just changing APKBUILD to try -rc2
2019-10-07 18:17:00 <Cogitri> Hehe :P
2019-10-07 18:17:49 <mps> actually I tried -rc1 but didn't had time to test it
2019-10-07 18:18:08 <mps> first will check on armv7
2019-10-07 18:19:02 <Cogitri> I don't think ZFS 0.8.2 works with 5.4 yet, soooo :c
2019-10-07 18:22:00 <Cogitri> On the topic of ZFS, if someone's here who happens to use ZFS without encryption it'd be nice of you if you could test https://github.com/alpinelinux/mkinitfs/pull/58
2019-10-07 18:22:11 <mps> hmm, this will be problem then, does it work with any in 5.x series
2019-10-07 18:22:25 <Cogitri> It works with 5.3 already
2019-10-07 18:22:54 <Cogitri> Maybe it works with 5.4 too, but they can't verify that before it's out :)
2019-10-07 18:22:57 <mps> aha, maybe they will upgrade to 5.4 RSN
2019-10-07 18:25:57 <Cogitri> Hopefully
2019-10-07 18:28:47 <mps> I don't have any real experience with ZFS so couldn't test, last night I noticed your changes to enable encryption on boot
2019-10-07 18:28:55 <mps> so I can't help
2019-10-07 18:36:56 <_ikke_> mps: would it make sense to change the url for ncurses to https://invisible-island.net/ncurses/, which same to be the actual homepage?
2019-10-07 18:37:54 <mps> just walked out of house, will look when I come back
2019-10-07 18:38:26 <_ikke_> Ok, no worry, it's not very important
2019-10-07 18:38:28 <mps> but yes, I thought so
2019-10-07 18:39:18 <mps> but I don't like to make big changes to package were I'm not in header, maintainer or contributor
2019-10-07 18:39:47 <mps> I make just minimum of changes which is needed
2019-10-07 18:40:12 <_ikke_> I consider these kinds of changes trivial
2019-10-07 18:40:14 <mps> also I looked at that url and thought it should be changed
2019-10-07 18:40:22 <_ikke_> ok
2019-10-07 18:41:06 <_ikke_> https://repology.org/project/ncurses/information#Homepages
2019-10-07 18:41:40 <mps> looked at that few weeks ago, and I saw that we need to change a lot of them
2019-10-07 18:42:10 <mps> most perl pkgs, iirc
2019-10-07 18:43:28 <_ikke_> Where do you even find the releases for ncurses?
2019-10-07 18:43:53 <_ikke_> This only contains the patches: ftp://ftp.invisible-island.net/ncurses/6.0/
2019-10-07 18:44:48 <mps> gnu site have url to invisible-island.net
2019-10-07 18:44:58 <mps> link*
2019-10-07 18:45:25 <mps> but I'm subscribed to ncurses ML
2019-10-07 18:45:34 <_ikke_> ah ok
2019-10-07 18:46:17 <mps> I follow development there
2019-10-07 18:49:50 <_ikke_> hmm, perl-io tests are now passing
2019-10-07 18:50:42 <mps> ah, why i didn't tried, shame on me
2019-10-07 18:50:46 <_ikke_> np
2019-10-07 18:50:58 <_ikke_> I read the comment and wondered wheter that was still teh case
2019-10-07 18:51:00 <_ikke_> the case
2019-10-07 18:52:03 <mps> I worked in parallel on it and pdns, and on zig
2019-10-07 18:52:17 <mps> should do things one by one
2019-10-07 18:52:55 <_ikke_> I tend to multitask a lot as well
2019-10-07 18:54:09 <mps> yes, those who use awesome wm naturally multitask :)
2019-10-07 18:54:40 <_ikke_> :D
2019-10-07 20:38:39 <tdtrask> PICCORO: acf provides a web interface for a variety of services. What services are you planning to host?
2019-10-07 20:42:01 <PICCORO> just web. dns that0s all
2019-10-07 20:42:23 <PICCORO> tdtrask: just web hosting and dns (the dns only to provide subdomain to clients)
2019-10-07 20:44:05 <PICCORO> so clients only will access to panel to see their hosting service as clients. that's can be possible?
2019-10-07 22:57:26 <tdtrask> PICCORO: generally, ACF is used to manage an entire box, not set up for shared services
2019-10-07 22:58:22 <tdtrask> tinydns was set up to add permissions by domain, so that multiple users could use the same ACF but only be able to manage certain domains
2019-10-07 22:58:35 <tdtrask> most ACF modules are not like that
2019-10-08 05:38:55 <ArchitectZ> Is there a method to autorun a script when booting the ISO?  I'm trying to create an unattended installation disc.
2019-10-08 05:45:42 <jmcgnh> ArchitectZ:-->  I don't have a recipe, but I'd think the place to look is https://wiki.alpinelinux.org/wiki/Initramfs_init
2019-10-08 05:47:09 <jmcgnh> It means creating your own ISO, but for your purposes, that should not be a barrier.
2019-10-08 05:48:16 <ArchitectZ> I've got an ISO created already, I'm just not sure how to have the script execute on start up.
2019-10-08 05:49:04 <jmcgnh> ArchitectZ:-->  You get the script to execute on startup by building the call to it into the initramfs.
2019-10-08 05:49:22 <ArchitectZ> On the arch linux ISO you can add `script=/root/install.sh` to the kernel parameters in the bootloader.  Just wondering if alpine has something similar.
2019-10-08 05:57:39 <jmcgnh> ArchitectZ:-->  It looks like there's an init= boot parameter, but it wants an executable, not a script
2019-10-08 06:04:49 <ArchitectZ> yeah, that should just point to the init script - which is busybox on the ISO.
2019-10-08 06:12:28 <jmcgnh> ArchitectZ:-->  Well, yes, but it seems you can have an alternate init executable and you could probably hook your script into that, if boot params is the way you want to go. Seems like it would be simpler to build the call to your script into the "normal" init system by customizing your initramfs.
2019-10-08 06:21:52 <ArchitectZ> ok, thanks.  If I'm going to have to alter the init script I could just have it read the script arg from the kernel params.
2019-10-08 06:26:53 <ArchitectZ> There is an `alpine_start` parameter but I can't find any documentation on what it's for.  https://github.com/alpinelinux/mkinitfs/blob/master/initramfs-init.in#L316
2019-10-08 06:37:32 <TBB> not sure if you want to put this stuff into initramfs really
2019-10-08 06:38:26 <TBB> if you're to install stuff you probably want the whole system up and then run the installer; I've done this before and our solution was to add agetty to a base install, make it autologin root and run the installer directly from root's profile
2019-10-08 06:45:43 <ArchitectZ> I could run it from inittab.  I also could set roots shell to be the script and setup autologin.  But it just doesn't seem like the correct way to do this.
2019-10-08 06:48:29 <TBB> yeh, you have to set up agetty for autologin in inittab as well. but if your installer depends on specific things OS services provide, for example in our case all installs were done over the network, those are easiest setup using those services. I don't think there's one right way to do this stuff really...
2019-10-08 06:48:59 <TBB> initramfs-init has some functionality but by no means it is a complete environment
2019-10-08 06:57:09 <TBB> but all that said, our requirements were very special so we did things in a way probably more complex than what's needed for your case, so I'm not implying the way we did it is the only right way :)
2019-10-08 07:03:03 <ArchitectZ> I also tried to put the script in /etc/local.d but the local service is not enabled on the iso.  I'd still need to run a command to enable it.
2019-10-08 07:03:51 <ArchitectZ> cron is enabled though.  That might be the easiest method.
2019-10-08 07:26:06 <fredrigu> PICCORO: then you can always ask if alpine really is the right choice for you. Are there any special software you do want to administrate? Many have their own web interface like cups and ldap
2019-10-08 07:27:41 <fredrigu> PICCORO: and I should learn to read the whole log before asking questions...
2019-10-08 07:28:19 <fredrigu> PICCORO: please describe more what you expect this panel to do
2019-10-08 07:29:19 <fredrigu> I understand it as you want to let a client be able to add new subdomains and to upload files, so something in between an apache/nginx config and ftp service?
2019-10-08 10:36:44 <mps> anyone have installed and running 4.19.76-0-vanilla on x86_64
2019-10-08 10:37:38 <tomponline> mps: yes i do
2019-10-08 10:37:49 <mps> need help about one issue
2019-10-08 10:37:52 <tomponline> sure
2019-10-08 10:38:09 <mps> tomponline: I'm trying to solve your issue
2019-10-08 10:38:20 <tomponline> mps: I thought you may be :)
2019-10-08 10:38:50 <mps> I don't have 4.19.76 so can't check it
2019-10-08 10:39:07 <tomponline> mps: how can I help?
2019-10-08 10:40:13 <mps> sounds stupid but can you look in /boot/config-vanilla and see what is on second line, i.e. version
2019-10-08 10:40:42 <tomponline> #
2019-10-08 10:40:44 <tomponline> # Automatically generated file; DO NOT EDIT.
2019-10-08 10:40:46 <tomponline> # Linux/x86_64 4.19.76 Kernel Configuration
2019-10-08 10:41:38 <mps> ok, my question was stupid
2019-10-08 10:42:43 <mps> can't understand what's happened, here is my 'Linux/x86_64 4.19.77 Kernel Configuration'
2019-10-08 10:44:02 <MY-R> mps, no hash or what? :D
2019-10-08 10:44:06 <mps> and 'grep CONFIG_BLK_DEV_THROTTLING boot/config-vanilla' => CONFIG_BLK_DEV_THROTTLING=y
2019-10-08 10:44:56 <mps> MY-R: no, he posted version with comments and I without
2019-10-08 10:45:21 <MY-R> oook
2019-10-08 10:45:40 <tomponline> mps: im trying a fresh install on a vm (but with vanilla iso)
2019-10-08 10:48:12 <tomponline> mps: same result, just installed on a VM from http://dl-cdn.alpinelinux.org/alpine/v3.10/releases/x86_64/alpine-standard-3.10.2-x86_64.iso, I use lvm followed by sys mode during setup-alpine (if that makes a difference), and /boot/config-vanilla doesn't have CONFIG_BLK_DEV_THROTTLING
2019-10-08 10:49:03 <mps> what uname -a shows on this release
2019-10-08 10:50:08 <MY-R> tomponline, done apk update; apk upgrade?
2019-10-08 10:50:49 <tomponline> localhost:~# uname -a
2019-10-08 10:50:51 <tomponline> Linux localhost 4.19.76-0-vanilla #1-Alpine SMP Thu Oct 3 12:00:48 UTC 2019 x86_64 Linux
2019-10-08 10:50:53 <tomponline> localhost:~# apk upgrade -U
2019-10-08 10:50:55 <tomponline> fetch http://dl-cdn.alpinelinux.org/alpine/v3.10/main/x86_64/APKINDEX.tar.gz
2019-10-08 10:50:57 <tomponline> OK: 674 MiB in 141 packages
2019-10-08 10:51:02 <MY-R> https://gitlab.alpinelinux.org/alpine/aports/blob/3.10-stable/main/linux-vanilla/config-vanilla.x86_64
2019-10-08 10:51:19 <MY-R> in 3.10 # CONFIG_BLK_DEV_THROTTLING is not set
2019-10-08 10:51:29 <tomponline> that would explain it then
2019-10-08 10:51:33 <MY-R> yep
2019-10-08 10:51:43 <mps> no, wait
2019-10-08 10:52:15 <mps> 4.19.76 should have CONFIG_BLK_DEV_THROTTLING=y
2019-10-08 10:52:36 <MY-R> but doesnt :)
2019-10-08 10:53:24 <mps> that is the question, why
2019-10-08 10:53:27 <MY-R> somebody bump kernel version without changing configs
2019-10-08 10:53:44 <mps> this isn't changed for some time
2019-10-08 10:56:02 <MY-R> # Linux/x86_64 4.19.73 Kernel Configuration in EDGE vs # Linux/x86_64 4.19.67 Kernel Configuration in 3.10
2019-10-08 10:57:34 <mps> this is enable in commit 6aef2d983bfc353adbf8d4e215c371140cdbe137
2019-10-08 10:57:47 <mps> Mon Aug 26 07:39:44 2019
2019-10-08 10:59:39 <mps> hmm, strange
2019-10-08 11:09:01 <tomponline> mps: how does it get into the 3.10 branch?
2019-10-08 11:12:22 <mps> let me see in git log for 3.10 stable
2019-10-08 11:13:35 <mps> ah, that explains, it is not enabled on 3.10 kernels
2019-10-08 11:13:52 <mps> only in edge kernels
2019-10-08 11:15:47 <mps> i.e. it is not applied to 3.10-stable branch when it is applied to edge
2019-10-08 11:21:36 <tomponline> mps: I see, so these things will not be backported until a newer stable is released?
2019-10-08 11:26:37 <mps> if no one change this in 3.10-stable branch it will not be enabled in any next 3.10 release
2019-10-08 11:27:01 <mps> looks like is time for some cherry-picking in aports
2019-10-08 11:28:12 <tomponline> mps: ok thanks
2019-10-08 11:30:16 <mps> np
2019-10-08 12:40:55 <midasi> how can I load a kernel module permanently in a diskless system?
2019-10-08 12:41:19 <midasi> 'modprobe xy' doesn't survive a reboot
2019-10-08 12:42:14 <qa3Txu0iak0F> putit into /etc/modules and then persist it with lbu
2019-10-08 12:42:25 <Cogitri> modprobe never survives a reboot
2019-10-08 12:42:50 <midasi> oh ok, thanks
2019-10-08 14:36:02 <mps> tomponline: I posted !302 hope it will be applied for next update to 3.10
2019-10-08 14:49:02 <tomponline> mps: thanks again
2019-10-08 14:50:42 <mps> np, again :)
2019-10-08 14:52:08 <zetuno> is it possible to use jemalloc from 3.3 on edge?
2019-10-08 14:52:30 <_ikke_> zetuno: depends on the dependencies
2019-10-08 16:50:13 <KennethDodrill> Hi all, just upgraded my server today and have run into some PHP issues. Has anyone had issues with the pgsql and pdo_pgsql extensions?
2019-10-08 16:50:53 <KennethDodrill> They are both installed, enabled in php.ini...I can send the error log for it if necessary
2019-10-08 16:54:52 <KennethDodrill> Oh, btw it's mostly a Nextcloud issue. Does anyone know if upgrading the nextcloud package will just remove all of my config?
2019-10-08 16:56:14 <KennethDodrill> this is what happens when I run the `occ upgrade` command for nextcloud. `UnexpectedValueException: The files of the app "cloud_federation_api" were not correctly replaced before running the update`
2019-10-08 16:57:10 <_ikke_> KennethDodrill: files in /etc should not be touched by ugprading
2019-10-08 16:57:13 <_ikke_> upgrading*
2019-10-08 16:57:42 <KennethDodrill> nextcloud is installed to the /usr/share/webapps folder
2019-10-08 16:58:00 <_ikke_> And you keep you config files in there as well?
2019-10-08 16:58:15 <_ikke_> (which is not recommended just because of this reason)
2019-10-08 16:58:18 <_ikke_> your*
2019-10-08 16:59:09 <KennethDodrill> oh oops, sorry it's been awhile since I've done this. Yes, config is in etc
2019-10-08 16:59:45 <KennethDodrill> interesting...I have a config.php and config.php.apk-new
2019-10-08 17:00:26 <_ikke_> right
2019-10-08 17:03:19 <KennethDodrill> Hmm, I'm not really seeing anything in the docs about the upgrade process. I can't seem to find much online either
2019-10-08 17:04:12 <_ikke_> I think there is an occ migrate command you need to execuate
2019-10-08 17:04:14 <_ikke_> execute
2019-10-08 17:10:04 <KennethDodrill> checked all app migrations, all went through - no pending ones
2019-10-08 17:12:40 <KennethDodrill> updated all apps (few of them had updates), same issue. Ran repair as well...
2019-10-08 17:12:51 <_0x40_> Hi. After rebooting alpine after installing, it gets stuck after " * Starting busybox crond ... [ ok ]"
2019-10-08 17:14:04 <_0x40_> This happens even if I boot from the install usb
2019-10-08 17:29:46 <_0x40_> I got it to boot by mashing keys on the keyboard to generate entropy.
2019-10-08 17:32:00 <mps> _0x40_: installing haveged and enabling it to run on boot can help
2019-10-08 17:32:06 <MY-R> _0x40_, apk -i add haveged; rc-update add haveged boot
2019-10-08 17:32:34 <_0x40_> Thanks, I'll give that a try
2019-10-08 17:33:25 <mps> it helped from waiting 15 minutes to 6 seconds on headless arm box
2019-10-08 17:35:51 <MY-R> in /etc/init.d/haveged can set HAVEGED_OPTS="-w 1024" but I think haveged set it by default or no?
2019-10-08 17:36:13 <MY-R> ^ /etc/conf.d/haveged
2019-10-08 17:38:14 <MY-R> mps, what is your "cat /proc/sys/kernel/random/write_wakeup_threshold" ?
2019-10-08 17:39:44 <mps> on arm32 box => 896
2019-10-08 17:40:10 <MY-R> ah so isnt set by default
2019-10-08 17:40:54 <mps> hmm, this is on patched kernel so not sure if it works as default
2019-10-08 17:41:28 <MY-R> but it can be set by using haveged in conf.d
2019-10-08 17:42:02 <mps> yes, but I didn't
2019-10-08 17:42:13 <MY-R> just do it! ;)
2019-10-08 17:42:22 <mps> patched kernel to pre 4.19 crng changes
2019-10-08 17:42:28 <MY-R> maybe will drop from 6s to 2s!
2019-10-08 17:42:38 <MY-R> ah
2019-10-08 17:43:02 <mps> no, this box doesn't have any boot wait time
2019-10-08 17:43:36 <mps> on another one I testing alpine kernels and there is installed haveged
2019-10-08 17:44:05 <MY-R> on openwrt with kernel HZ=100 random: crng init done is after maybe 30 seconds on poor device when with HZ=1000 and dynamic ticks taking just a few seconds
2019-10-08 18:14:39 <KennethDodrill> I'm just going to restore from a backup because I can't get Nextcloud to update. Is it advised to go with the web updater instead of through the package?
2019-10-08 18:20:16 <alex1010101> Hello. I am configuring WireGuard on Alpine Linux.
2019-10-08 18:20:46 <alex1010101> Following the guide (https://wiki.alpinelinux.org/wiki/Configure_a_Wireguard_interface_(wg)), my client device can connect to the interface on the server.
2019-10-08 18:21:01 <alex1010101> However, the client cannot reach the Internet.
2019-10-08 18:21:53 <alex1010101> Moreover, the line "post-up ip route add x.x.x.x/24 dev wg0" seems incorrect. When I run this line (without post-up) directly, Alpine Linux says "ip: RTNETLINK answers: Invalid argument".
2019-10-08 18:22:26 <alex1010101> Did I miss something?
2019-10-08 18:43:45 <MY-R> alex1010101, did you try with "iproute2" package installed?
2019-10-08 18:44:46 <MY-R> by default "ip" point to /bin/busybox
2019-10-08 18:49:42 <KennethDodrill> Finally got everything updated - ran the updater through updater.phar in the nextcloud dir, then updated all packages for alpine, then ran the maintenance repair script for the occ command. Everything works now :)
2019-10-08 21:07:47 <nixme> I'm having issues running an application and it seems like I'm getting Operation not permitted -- it seems like the app tries to mmap /dev/mem -- Is there a way to disable this restriction? Or can I add an account to access that?
2019-10-08 21:08:52 <AinNero> nixme: just curious, which application is that?
2019-10-08 21:09:31 <nixme> AinNero It is just a golang app that is trying to talk to the raspberry pi's gpios
2019-10-08 21:10:24 <AinNero> you could make /dev/mem world-writable via chmod
2019-10-08 21:10:29 <AinNero> or run the application as root
2019-10-08 21:10:56 <AinNero> the second is a less bad thing that the first
2019-10-08 21:28:24 <nixme> AinNero I had to add iomem=relaxed to the kernel params for it to work
2019-10-08 21:28:48 <AinNero> sounds like a different problem
2019-10-08 21:32:11 <nixme> yeah.. I was clearly wrong with what the issue was so I'll have to figure out why since that's not the best fix
2019-10-08 21:37:40 <AinNero> if you need 'just' iomem=relaxed, it seems fine
2019-10-08 23:17:35 <nicolaus> im trying to start up a murmur (mumble) service, it's crashing on start; always asking for lsb_release
2019-10-08 23:20:16 <MY-R> nicolaus, it is just a warning
2019-10-08 23:20:27 <nicolaus> good!
2019-10-08 23:20:31 <nicolaus> MY-R: thanks :)
2019-10-08 23:20:48 <nicolaus> so the real problem is this one: SQL Error [INSERT INTO `channels` (`server_id`, `channel_id`, `parent_id`, `name`) VALUES (?, ?, ?, ?)]: attempt to write a readonly database Unable to fetch row
2019-10-08 23:21:03 <MY-R> nicolaus, try use first default config cus it changed since previous release
2019-10-08 23:21:23 <nicolaus> i'm using the default /etc/mumble.ini
2019-10-08 23:21:29 <nicolaus> without any edit
2019-10-08 23:21:46 <MY-R> attempt to write a readonly database Unable to fetch row
2019-10-08 23:21:52 <nicolaus> and running the server via "rc-service murmur start" in root
2019-10-08 23:22:00 <MY-R> it cant write there so check your permissions
2019-10-08 23:22:21 <MY-R>  /var/lib/murmur/
2019-10-08 23:23:00 <nicolaus> -rw-r----- murmur.sqlite
2019-10-08 23:23:37 <MY-R> is it owned by murmur:murmur?
2019-10-08 23:23:51 <nicolaus> by root
2019-10-08 23:24:32 <MY-R> so "chown murmur:murmur murmur.sqlite"
2019-10-08 23:24:48 <MY-R> and be sure that dir in which is that file is owned by murmur too
2019-10-08 23:28:08 <nicolaus> ok i'll try :) thanks
2019-10-08 23:28:32 <MY-R> gl!
2019-10-08 23:40:05 <nicolaus> MY-R: it's better to run murmur as root or murmur (user)?
2019-10-09 00:25:05 <nicolaus> MY-R: it is running with uname=murmur in the /etc/murmur.ini ; and murmur permissions in the sqlite file
2019-10-09 00:25:48 <nicolaus> i can connect without problems via mumble client, yet the "rc-status" is showing "murmur crashed"; very strange
2019-10-09 00:26:13 <nicolaus> doing tail -f /etc/murmur.log shows me that is running fine lol
2019-10-09 03:41:36 <priontology> Did we ever figure out if I'm just an idiot regarding the nmap scripting stuff?
2019-10-09 03:53:52 <nicolaus> where i can contact the package maitainer of murmur?
2019-10-09 04:05:41 <priontology> nicolaus: invite me to your mumble server
2019-10-09 04:06:01 <nicolaus> mumble://mumble.miratengointernet.com:64738
2019-10-09 04:06:07 <nicolaus> i think it's working
2019-10-09 04:06:50 <nicolaus> even though open-rc is listing the service as "crashed"
2019-10-09 10:00:02 <nul1__> o/! im trying to access rbd images from 4.19.78-1-vanilla but the kernel module is not found. is there a way to make it work? (no issues on 4.19.78-0-virt) thanks in advance!
2019-10-09 10:14:44 <jontto> Hi guys and gals, I'm new to the alpine community so a quick questions about bug reports: I have a package (encodings) that fails to install in Docker container when build is performed on a Windows 10 machine. Where should I post the bug report in this case?
2019-10-09 10:15:52 <Cogitri> Https://gitlab.alpinelinux.ort/Alpine/aports/issues I guess
2019-10-09 10:16:50 <ncopa> nul1__: did you reboot?
2019-10-09 10:21:21 <nul1__> ncopa, its not that easy this time :) im missing /lib/modules/4.19.78-1-vanilla/kernel/drivers/block/rbd.ko
2019-10-09 10:22:32 <jontto> Cogitri: thanks, I was a bit puzzled by which project, but I'll post it there.
2019-10-09 10:22:47 <ncopa> nul1__: huh...
2019-10-09 10:22:49 <nul1__> ill look into compiling it later, if nobody has done it already
2019-10-09 10:23:04 <AinNero> ncopa: the module is not in the package
2019-10-09 10:23:26 <ncopa> so the only thing that changes from 4.19.78-r0 -> -r1 was that i enabled CONFIG_DM_UEVENT
2019-10-09 10:23:46 <AinNero> config-vanilla.x86_64:# CONFIG_BLK_DEV_RBD is not set
2019-10-09 10:23:56 <AinNero> config-virt.x86_64:CONFIG_BLK_DEV_RBD=m
2019-10-09 10:24:23 <nul1__> it was also an issue in  4.19.73-0-vanilla
2019-10-09 10:24:47 <ncopa> 3ba52b18b4d728383adb08f947814b7346c33e11
2019-10-09 10:25:35 <ncopa> nul1__: what architecture? x86_64?
2019-10-09 10:26:26 <nul1__> yup
2019-10-09 10:26:41 <AinNero> ncopa: yes, look in aports/main/linux-vanilla/, grep RBD config-v*
2019-10-09 10:27:03 <AinNero> it seems like vanilla and virt have been different on CONFIG_BLK_DEV_RBD for quite a while now
2019-10-09 10:27:06 <AinNero> not a recent change
2019-10-09 10:31:12 <ncopa> ah. its -virt -> -vanilla
2019-10-09 10:31:24 <ncopa> that makes the diff, not -r0 -> -r1
2019-10-09 10:36:25 <vkar> hello
2019-10-09 10:36:34 <vkar> guyz! need help with alpine linux ulimit configuration. I found a similar forum post: https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum/general-discussion/set-system-wide-limits.html Anybody can help with that?
2019-10-09 10:37:05 <nul1__> i tried using the rbd.ko from the virt but it errors: [ 3190.584517] rbd: disagrees about version of symbol module_layout
2019-10-09 10:38:59 <AinNero> nul1__: using kernel modules from one kernel built in another kernel build is not supported by the kerne;
2019-10-09 10:39:17 <AinNero> this is a limitation from the kernel itself
2019-10-09 11:08:43 <vkar> anybody knows why linux-pam module doesnt have any affect in alpine linux?
2019-10-09 11:09:45 <Cogitri> What do you mean?
2019-10-09 11:09:50 <Cogitri> wfm
2019-10-09 11:13:14 <vkar> I mean when I configure limits in "/etc/security/limits.conf" file it doesnt make any affect
2019-10-09 11:14:50 <vkar> having entry "*	-	 nofile	10001" in "/etc/security/limits.conf" I still have 1024 for "ulimit -n" when re-login
2019-10-09 11:17:00 <MY-R> vkar, most probably some of the configs in /etc/pam.d/ need a tweak like additional line with: session required pam_limits.so
2019-10-09 11:17:37 <vkar> ye I already played with it as well, but that doesnt help either
2019-10-09 11:21:05 <MY-R> shouldnt be two entries like "* hard nofile 16384" and "* soft nofile 16384" ?
2019-10-09 11:22:35 <vkar> I tried both ways as well
2019-10-09 11:23:17 <vkar> I am curious if that works in your system :)
2019-10-09 11:24:04 <MY-R> I dont use pam on alpine
2019-10-09 11:24:16 <MY-R> still you can set I think ulimit -n 4096
2019-10-09 11:24:27 <MY-R> try set if you can more
2019-10-09 11:24:40 <vkar> yes it works, but only for current session. And I want it to be present after reboot
2019-10-09 11:24:40 <MY-R> try set more if you can*
2019-10-09 11:25:10 <MY-R> ye, but can you set more than 4096?
2019-10-09 11:26:57 <vkar> if I am root then I can set more, If I am basic user I have "sh: error setting limit: Operation not permitted" when I try "ulimit -n 4097"
2019-10-09 11:43:31 <MY-R> hmm :\
2019-10-09 11:56:38 <PICCORO> any panel software for alpine ? just web hosting and dns (the dns only to provide subdomain to clients) ?
2019-10-09 11:58:27 <AinNero> PICCORO: I used Froxlor in the past
2019-10-09 11:59:16 <AinNero> its distribution-independent tho, so not specific to alpine
2019-10-09 11:59:21 <TBB> isn't ACF for that purpose specifically?
2019-10-09 11:59:47 <TBB> 
ACTION has no idea about the development status of Alpine Configuration Framework

2019-10-09 12:00:25 <AinNero> acf is more for general system administration
2019-10-09 12:00:29 <AinNero> compated to luci on openwrt
2019-10-09 12:00:36 <AinNero> *compared
2019-10-09 12:01:05 <AinNero> Froxlor is a thing focused on web & mail hosting
2019-10-09 12:01:37 <TBB> a lot of what it's usable for depends on what modules have been written for it; this is not my area of expertise really, I just like the look of text I've written myself I suppose ;)
2019-10-09 12:01:41 <AinNero> with a interface for customers instead of admin-only
2019-10-09 12:32:41 <PICCORO> AinNero: that's was i talking about.. only for those clients that does nto not know nothig about ..  no db admin due each client only will have one db
2019-10-09 12:34:02 <AinNero> PICCORO: yeah then look into froxlor or their alternatives
2019-10-09 12:57:43 <vkar> so I am a bit confused. In ubuntu it just works :\
2019-10-09 13:01:20 <vkar> and no info in internet except of forum post https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum/general-discussion/set-system-wide-limits.html, but there is still no any answer
2019-10-09 13:59:11 <workodera> Hey, I changed my default shell from ash to fish by changing the path in passwd, however now when I hit ctrl+c during input it logs me out...
2019-10-09 13:59:18 <workodera> does anyone know how to fix this?
2019-10-09 14:09:21 <PICCORO> hi! who created https://wiki.alpinelinux.org/wiki/Fonts ? for some merge work to!?
2019-10-09 14:10:03 <PICCORO> workodera: yo does not change shell in fact.. seems your shell are executed on top of some kind of subprocess
2019-10-09 14:10:13 <PICCORO> maybe pam settings
2019-10-09 14:10:44 <workodera> should I just install and use shadow/chsh then?
2019-10-09 14:11:34 <workodera> that did the trick lol
2019-10-09 14:59:36 <mrnoone> I broke my APK database. Whatever I try to do it returns: ERROR: unsatisfiable constraints: linux-vanilla-4.19.78-r1: breaks: world[linux-vanilla><Q1WmB2andpWSof1Xl+vmQyjYrorZ4=]. How can I manually fix it?
2019-10-09 15:06:02 <_ikke_> mrnoone: check /etc/apk/world
2019-10-09 15:19:24 <midasi> where is in alpine the default log configuration when I don't have sysklogd or rsyslog installed (i.e. there is no syslog.conf)
2019-10-09 15:21:50 <MY-R> midasi, /etc/conf.d/syslog
2019-10-09 15:22:50 <MY-R> syslogd --help
2019-10-09 15:24:59 <midasi> MY-R: I see, thanks. It seems to be very limited though...
2019-10-09 15:25:24 <MY-R> busybox
2019-10-09 15:26:02 <midasi> yes... I wonder why my iptables LOG entries are not logged, that's why I was looking for the configuration
2019-10-09 15:26:17 <MY-R> midasi, they are, do dmesg :P
2019-10-09 15:26:28 <midasi> oh
2019-10-09 15:26:31 <midasi> there they are :D
2019-10-09 17:57:44 <mateusz`> Hi guys, I have a problem with telnet on alpine. I have an access to some service where my IP is whitelisted. I can telnet from my VM which is ubuntu. On this VM I have few docker instances, and I can also telnet this service from ubuntu docker instance, but I can't access it from my docker alpine instance. I get "Connection closed by foreign host" error
2019-10-09 17:58:35 <mateusz`> telnet generally works on alpine docker instance, I can telnet some free public hosts, but not this particular service which I need
2019-10-09 18:05:03 <mateusz`> I forgot to mention, alpine version 3.10.2
2019-10-09 18:11:13 <KatolaZ> mateusz`: can you reach the service via nc?
2019-10-09 18:32:33 <eldowan> Where can I find more information about the different download versions?
2019-10-09 18:32:45 <eldowan> extended, netboot, virtual, etc?
2019-10-09 18:39:42 <programmerq> more info than their short descriptions on the download page?
2019-10-09 18:39:58 <programmerq> what's your usecase?
2019-10-09 18:44:44 <eldowan> Yes, I'm looking for more info than the short description if it's available. Currently I'm deploying to an esxi server for use solely as a chrony time server, so I'm starting witht the virtual download.
2019-10-09 18:46:13 <programmerq> virtual is what I would have chosen for an esxi guest.
2019-10-09 18:46:42 <mps> is there vim syntax plugin for zig somewhere
2019-10-09 18:50:03 <eldowan> thanks programmerq, that's what I thought.
2019-10-09 18:52:32 <programmerq> https://git.alpinelinux.org/alpine-iso/tree/ - it looks like the scripts used to generate those is in this git repo
2019-10-09 18:52:54 <eldowan> oh awesome. most excellent
2019-10-09 18:55:38 <telmich> hey guys
2019-10-09 18:55:55 <telmich> I just wanted to give a bit of feedback to the UEFI instructions on the wiki
2019-10-09 18:56:06 <programmerq> looks like virt has a different kernel flavor and serial/boot options are different
2019-10-09 18:57:04 <telmich> The new Dell xps 13" 2-in-1 really does not start with the included grub; then I tried to boot alpine on a dell latitude 5285 (tablet-y thing)
2019-10-09 18:57:26 <telmich> that one sees the uefi boot medium, but just goes back to the boot prompt when I select alpine
2019-10-09 19:16:06 <PICCORO> Adding a lockdown to the Linux kernel was now aproved with linus flavor, and 5.4 will have a lockdown module, which blocks and locks the kernel (to modifications) even if it is root, that means each distro will have to activate it manually which is a nonsence separating the distros more and more and complicating distros with personal kernels. That's are no so good news at all
2019-10-09 19:25:17 <telmich> PICCORO: how does that work / how is it locked down against what?
2019-10-09 19:28:10 <AinNero> PICCORO: i think not each distro, but each user will have to activate it manually
2019-10-09 19:28:19 <AinNero> or it might be done via the update-grub mechanism
2019-10-09 19:28:52 <AinNero> opt-out is not an option here, that would break systems that rely on the now blocked functionality
2019-10-09 19:29:31 <AinNero> there has been a similar case with sysctl values that you cannot reset after setting them to specific values
2019-10-10 02:51:02 <shrizza> Anybody here use Inkscape on Alpine?
2019-10-10 02:51:45 <shrizza> Trying to figure out why "Trace Bitmap" does not work.
2019-10-10 02:51:53 <shrizza> (for the Alpine package)
2019-10-10 02:54:54 <mckaygerhard> wiki page for newbie users https://wiki.alpinelinux.org/wiki/Alpine_newbie_apk_packages#fonts_and_fonts_configuration about font adn configurations of fonts for each language desktop case most commonly
2019-10-10 07:22:26 <mungustas> hi
2019-10-10 07:22:31 <mungustas> running alpine on docker
2019-10-10 07:23:24 <mungustas> root users crontabs are fine, I can't seem to figure how to run tasks under different user
2019-10-10 07:24:26 <mungustas> bash-4.4# ls -al /var/spool/cron/crontabs/www-data
2019-10-10 07:24:27 <mungustas> -rw-------    1 www-data www-data      4682 Oct 10 07:10 /var/spool/cron/crontabs/www-data
2019-10-10 07:24:51 <mungustas> I have crontabs for user www-data which I want to be run
2019-10-10 07:57:35 <iggy> mungustas: under docker, cron wouldn't be running
2019-10-10 08:07:39 <mungustas> https://github.com/aptible/supercronic okay trying this one now
2019-10-10 08:13:38 <CvH> Hi, could anyone point me to a solution
2019-10-10 08:13:44 <CvH> I try to build "./scripts/mkimage.sh --arch armv7 --profile uboot --extra-repository /home/cvh/repo/ --repository http://dl-cdn.alpinelinux.org/alpine/v3.10/main/"
2019-10-10 08:14:02 <CvH> But I get "WARNING: Ignoring /home/cvh/repo/armv7/APKINDEX.tar.gz: UNTRUSTED signature"
2019-10-10 08:14:38 <CvH> the content of cvh/repo/armv7 was created from building an aports package
2019-10-10 08:14:41 <mps> CvH: I filled bug report about this
2019-10-10 08:16:42 <CvH> mps: is there a hack/workaround for now ?
2019-10-10 08:17:41 <mps> I don't know, this one holds me to make better uboot tarballs and test them
2019-10-10 08:20:48 <CvH> :(  can't test the kernel either
2019-10-10 08:28:42 <mps> I install current tarball and then upgrade kernels to test them
2019-10-10 08:32:15 <mps> CvH: you can try to add your abuild key to /etc/apk/keys, maybe it can help
2019-10-10 08:55:45 <CvH> mps: cp the pub file into /etc/apk/keys worked , tx
2019-10-10 08:56:45 <mps> nice, when you write guide please inform me
2019-10-10 08:58:30 <CvH> its more an "how to workaround the problems"
2019-10-10 09:00:19 <mps> whatever it is it will be useful I think
2019-10-10 09:00:19 <CvH> but if somewhere would be stated that cross compiling is not working and the dirty diry hack with qemu is horrible slow would be a nice start ;D  thats nothing obvious that just native compiling is properly supported at all
2019-10-10 09:01:09 <CvH> yea ... If I need to redo this I would fall into all potholes again
2019-10-10 09:01:42 <mps> CvH: do you run qemu host as arm32 or arm64
2019-10-10 09:04:45 <CvH> whatever the official alpine script did :D  "Linux ubuntu 5.0.0-32-generic #34-Ubuntu SMP Wed Oct 2 02:06:48 UTC 2019 armv7l Linux"
2019-10-10 09:06:57 <mps> no, I mean qemu-system-aarch64 or qemu-system-arm
2019-10-10 09:07:22 <mps> and, sorry, not host but guest
2019-10-10 09:12:34 <CvH> host is ubuntu x86_64 vm, guest is "armv7l" -> arm32
2019-10-10 09:14:37 <mps> you can try to run arm64 guest and setarch to linux32, maybe it will build faster then
2019-10-10 09:53:34 <mateusz`> KatolaZ: yes, I can reach via nc. I execute "nc -zv <host> <ip>" and I get "open"
2019-10-10 10:16:17 <PureTryOut[m]> Why does Busybox's `ping` require root?
2019-10-10 10:16:47 <Cogitri> I guess because it doesn't have the right capabilities set (or SUID) to do it without root
2019-10-10 10:16:55 <AinNero> raw socket
2019-10-10 10:16:58 <mps> PureTryOut[m]: how to setcap to busybox applet
2019-10-10 10:17:10 <AinNero> there is a sysctl to work around it
2019-10-10 10:17:22 <PureTryOut[m]> I doubt it should use SUID, as you'd then give those rights to the entirety of Busybox
2019-10-10 10:17:30 <AinNero> sysctl net.ipv4.ping_group_range=999 59999
2019-10-10 10:18:00 <Cogitri> PureTryOut: Yes, It's just what coreutils ping does, so I mentioned it
2019-10-10 10:18:08 <Cogitri> I wouldn't do that, AinNero
2019-10-10 10:18:12 <PureTryOut[m]> Ah yes
2019-10-10 10:18:28 <PureTryOut[m]> In case of coreutils that's possible due to them having a separate binary for it
2019-10-10 10:18:29 <Cogitri> Applying the capability sounds like the best way to solve this as it only allows that one binary to run ping
2019-10-10 10:19:20 <mps> ping requires CAP_NET_RAW, iirc
2019-10-10 10:19:38 <PureTryOut[m]> I'm not sure what you mean with capability in this context
2019-10-10 10:20:03 <Cogitri> the `libpcap` capabilities :)
2019-10-10 10:20:17 <PureTryOut[m]> Ah
2019-10-10 10:21:02 <mps> Cogitri: you misspeled it
2019-10-10 10:21:08 <mps> libcap
2019-10-10 10:21:37 <mps> pcap is packet capture, and I know that you know that
2019-10-10 10:21:41 <AinNero> Cogitri: i quoted the alpine defaults
2019-10-10 10:22:05 <AinNero> cat /etc/sysctl.d/00-alpine.conf
2019-10-10 10:22:21 <AinNero> this is the supported way for unprivileged ping to work on alpine
2019-10-10 10:26:43 <mps> and, btw, standalone ping is in iputils
2019-10-10 10:35:18 <CvH> mps: any idea why the extra repo is not used and still the default kernel from the alpine repo is used ?
2019-10-10 10:37:09 <PureTryOut[m]> Hmpf, I just upgraded my system, rebooted, and now network manager doesn't want to connect to my wifi anymore...
2019-10-10 10:37:22 <PureTryOut[m]> Cogitri: does wifi still work for you on latest network manager?
2019-10-10 10:37:52 <Cogitri> Yup, using it in this second
2019-10-10 10:37:57 <Cogitri> What doesn't work?
2019-10-10 10:38:21 <PureTryOut[m]> Well like I said, it just refuses to connect to my network. No error message given
2019-10-10 10:38:27 <mps> CvH: that is question I have no answer, here is issue I opened some time ago https://gitlab.alpinelinux.org/alpine/aports/issues/10543
2019-10-10 10:38:37 <PureTryOut[m]> `nmcli c up <ssid> --ask` just keeps asking for my password
2019-10-10 10:38:59 <PureTryOut[m]> Ah now it reports "The Wi-Fi network could not be found"
2019-10-10 10:39:12 <PureTryOut[m]> But it is, it shows up in `nmtui` and `plasma-nm`...
2019-10-10 10:42:08 <CvH> mps: no my issue is that linux-vanilla is used from the official repo instead what I have at extra-repo
2019-10-10 10:42:57 <mps> also, my issue is the same
2019-10-10 10:43:29 <steinex> hi, i've just discovered that the ip address behind git.alpinelinux.org is owned by the Jehovah's Witnesses in .NL. Why is this? :D
2019-10-10 10:43:30 <mps> actually I should/could close #10543
2019-10-10 10:43:54 <CvH> mps: then I don't understand the issue at your linked isse
2019-10-10 10:44:39 <Cogitri> PureTryOut: huh, odd
2019-10-10 10:45:01 <Cogitri> Try starting NM with --foreground and enable debugging to see more info
2019-10-10 10:45:20 <mps> steinex: business as usual :D
2019-10-10 10:50:05 <PureTryOut[m]> How do I enable debugging? And `NetworkManager` doesn't support `--foreground`, although it does know `--no-daemon`
2019-10-10 10:58:18 <PureTryOut[m]> Ok wtf after 15 minutes of trying it now just connected, and I basically changed nothing...
2019-10-10 10:58:19 <PureTryOut[m]> ¯\_(ツ)_/¯ computers
2019-10-10 11:04:33 <mps> PureTryOut[m]: AI finally works ;)
2019-10-10 11:05:27 <PureTryOut[m]> 🎉
2019-10-10 11:27:40 <Cogitri> PureTryOut: ah yes, getting the option name right when everything uses something different is hard :P
2019-10-10 11:27:46 <Cogitri> Good that it works now :)
2019-10-10 11:34:20 <CvH>  mps: found no way in making it work :(  how the heck is kernel dev/packaging done at alpine if there is no working way in creating the package ?
2019-10-10 11:34:54 <CvH> uboot is in the same boat
2019-10-10 11:40:38 <mps> CvH: not sure how it works, and no one answered my questions on IRC for some time so I gave up
2019-10-10 11:42:50 <CvH> maybe ncopa can answer this ? it is rather strange that you can't even build properly alpine and nobody cares or uses it O_o
2019-10-10 11:44:30 <mps> CvH: short time, I don't have time to debug this problem
2019-10-10 11:49:31 <CvH> i have too less understanding how alpine handles that
2019-10-10 11:50:54 <tankf33der> o/
2019-10-10 11:51:58 <tankf33der> https://pastebin.com/zZg0yCZt
2019-10-10 11:52:20 <tankf33der> this is edge. what i need to install to be able to run this one.
2019-10-10 11:52:22 <tankf33der> thanks.
2019-10-10 11:53:46 <mps> CvH: I understand it superficially but don't have time to look deeply
2019-10-10 12:12:15 <PICCORO> hey guys.. new kernel 5.4 now has lockdown.. how will be the position of alpine respect that?
2019-10-10 12:16:04 <mps> PICCORO: this should be end user decision
2019-10-10 12:17:38 <AinNero> it has to be
2019-10-10 12:17:58 <AinNero> enabling it per default would just break usecases like kexec
2019-10-10 12:18:00 <qman__> I would imagine it would be compiled in but turned off by default, allowing you to enable it by adjusting configuration
2019-10-10 12:18:12 <AinNero> its surely be useful for TPM/uefi setups
2019-10-10 12:18:49 <mps> lockdown places no restrictions by default
2019-10-10 12:19:58 <mps> if user wants then it can add kernel parameter 'lockdown=confidentiality' (for example)
2019-10-10 12:20:43 <AinNero> what did you think i mean with "enabling it per default" ?
2019-10-10 12:22:41 <mps> to me having option to do 'cat /dev/zero > /dev/mem' is feature
2019-10-10 12:23:34 <mps> also, 'cat /dev/zero > /dev/sda' is good to have
2019-10-10 12:23:54 <AinNero> ?
2019-10-10 12:24:37 <CvH> mps: so to workaround the whole mess i rsynced the repo, replaced the linux package with my selfbuild one and signed the APKINDEX according to the wiki, looks working but now i get `ERROR: busybox-initscripts-3.1-r7: package mentioned in index not found (try 'apk update')`
2019-10-10 12:25:20 <mps> CvH: probably signature (key) problem
2019-10-10 12:26:09 <CvH> so https://wiki.alpinelinux.org/wiki/Apkindex_format is not up to date ?
2019-10-10 12:26:13 <mps> maybe to sign all downloaded apk's with your key
2019-10-10 12:28:05 <CvH> `apk index -o APKINDEX.unsigned.tar.gz *.apk` -> `abuild-sign -k ~/.abuild/cvh-5d990079.rsa /home/cvh/repo/armv7/APKINDEX.unsigned.tar.gz` -> `mv APKINDEX.unsigned.tar.gz APKINDEX.tar.gz`
2019-10-10 12:31:00 <ncopa> CvH: how can i help you?
2019-10-10 12:31:13 <ncopa> i didnt get the question sorry
2019-10-10 12:32:31 <CvH> i build kernel-vanilla, then I try to build `./scripts/mkimage.sh --arch armv7 --profile uboot` with this self build kernel
2019-10-10 12:32:34 <CvH> `./scripts/mkimage.sh --arch armv7 --profile uboot --extra-repository /home/cvh/repo/ --repository http://dl-cdn.alpinelinux.org/alpine/v3.10/main/`
2019-10-10 12:32:50 <CvH> ^^^that works but does not pick up the kernel from my "extra repo"
2019-10-10 12:33:33 <CvH> key etc is okay and no error message
2019-10-10 12:33:49 <ncopa> do you run the mkimage.sh script on an armv7 machine?
2019-10-10 12:35:11 <CvH> my setup is x86_64 ubuntu vm -> qemu from alpinelinux/alpine-chroot-install -> tried the script from the armv7 qemu instance
2019-10-10 12:36:29 <ncopa> what version is your your linux-vanilla package in /home/cvh/repo and what version is on dl-cdn.a.o v3.10/main?
2019-10-10 12:37:14 <ncopa> it will likely take the "newest", so if version is lower in your local repo, it may take the package from dl-cdn instead
2019-10-10 12:38:39 <ncopa> can you please also run: apk audit /home/cvh/repo/armv7/linux-vanilla*.apk
2019-10-10 12:38:40 <mps> ncopa: it didn't worked also in my tries
2019-10-10 12:38:55 <ncopa> mps: i think you reported it too
2019-10-10 12:39:07 <ncopa> do you remember the bug number?
2019-10-10 12:39:15 <mps> yes, but that was about absolute paths
2019-10-10 12:39:23 <CvH> ncopa both packages have the same version
2019-10-10 12:39:35 <mps> #10543
2019-10-10 12:40:06 <CvH> i guess the whole point at an extra repo is to use stuff additional to the normal repo where nobody has write access to
2019-10-10 12:41:06 <mps> but I didn't had time to deeply look at it because I solved the needs by installing new kernel after installing uboot tarball
2019-10-10 12:41:36 <CvH> maybe there is a way how i can tell `./scripts/mkimage.sh --arch armv7 --profile uboot` to use an particular package
2019-10-10 12:41:56 <CvH> otherwise you can always just build whatever is at the repo
2019-10-10 12:44:02 <CvH> `ubuntu:~/aports$ apk audit /home/cvh/repo/armv7/linux-vanilla-4.19.76-r0.apk` outputs nothing
2019-10-10 12:45:12 <ncopa> CvH: sorry, it was supposed to be `apk verify`, not `apk audit`
2019-10-10 12:46:03 <ncopa> can you please also try: apk verify /home/cvh/repo/armv7/APKINDEX.tar.gz
2019-10-10 12:46:56 <CvH> 3x OK
2019-10-10 12:47:06 <ncopa> ok
2019-10-10 12:47:25 <ncopa> thats a proof that the repo index is ok, and that the packages are ok
2019-10-10 12:47:33 <ncopa> that keys are found etc
2019-10-10 12:48:34 <ncopa> both packages has same version, and im not sure how apk will handle that
2019-10-10 12:48:45 <CvH> can i set a priority ?
2019-10-10 12:49:02 <ncopa> basically, apk will detect that they are different (since checksum is different), but im not sure how it prioritizes them
2019-10-10 12:49:11 <CvH> btw what happen if i rename it to some higher version ?
2019-10-10 12:49:13 <ncopa> it will treat it as same package, since pkgname is the ame
2019-10-10 12:49:39 <ncopa> it reads version from .PKGINFO inside the .apk
2019-10-10 12:50:07 <ncopa> i'd bump pkgrel and rebuild
2019-10-10 12:50:34 <CvH> meh
2019-10-10 12:50:44 <CvH> 17h ...
2019-10-10 12:50:48 <ncopa> !
2019-10-10 12:50:50 <CvH> instead of 10min
2019-10-10 12:51:05 <ncopa> right, in qemu...
2019-10-10 12:51:23 <ncopa> if you still have the builddir
2019-10-10 12:51:29 <ncopa> from previous build
2019-10-10 12:52:01 <CvH> nope :D
2019-10-10 12:52:07 <ncopa> then you can bump pkgrel in APKBUILD, rm -rf pkg and then run  `abuild rootpkg index`
2019-10-10 12:52:52 <CvH> what does `abuild rootpkg index` ?
2019-10-10 12:53:30 <ncopa> abuild rootpkg will re-exec abuild in fakeroot, and run package() function in fakeroot
2019-10-10 12:53:43 <ncopa> and then split the packages (run the sptil funktions)
2019-10-10 12:53:47 <ncopa> split*
2019-10-10 12:53:56 <ncopa> and finally it will create the .apks
2019-10-10 12:54:16 <CvH> isnt `abuild -r`enough ?
2019-10-10 12:54:20 <ncopa> abuild index will update the APKINDEX.tar.gz
2019-10-10 12:54:23 <ncopa> it is
2019-10-10 12:54:47 <mps> CvH: 'abuild -r' cleans and rebuild from scratch
2019-10-10 12:54:53 <ncopa> but if you had the previous build, then you wouldnt need rebuild it from scratch
2019-10-10 12:54:56 <CvH> ah okay
2019-10-10 12:55:00 <ncopa> just run the "make install" phase
2019-10-10 12:55:11 <CvH> i guess i setup some build setup at a native arm client
2019-10-10 12:55:50 <ncopa> it was for saving you for 17h build...
2019-10-10 12:55:53 <CvH> hopefully ancient 4.19 kernel supports my devises :D
2019-10-10 12:56:08 <CvH> yea i already nuked the directory
2019-10-10 12:56:23 <mps> CvH: tell me your invocation of mkimage.sh, I can try on native hardware
2019-10-10 12:57:31 <CvH> `./scripts/mkimage.sh --arch armv7 --profile uboot --extra-repository /home/cvh/repo/ --repository http://dl-cdn.alpinelinux.org/alpine/v3.10/main/`
2019-10-10 12:58:06 <ncopa> oh, i think there is a way
2019-10-10 12:58:16 <CvH> ubuntu:~/aports$ ls -l /home/cvh/repo/armv7/
2019-10-10 12:59:05 <mps> thanks
2019-10-10 12:59:38 <ncopa> yeah
2019-10-10 12:59:40 <PICCORO> help please: lxc-start a1 20191010125713.315 ERROR    cgroup - cgroups/cgroup.c:cgroup_init:49 - Failed to initialize cgroup driver
2019-10-10 13:00:24 <ncopa> PICCORO: maybe start /etc/init.d/cgroups?
2019-10-10 13:00:28 <PICCORO> mps: but end user are no preciselly "user" in alpine...
2019-10-10 13:01:03 <ncopa> CvH: i think there may be a workaround
2019-10-10 13:01:09 <ncopa> i havent tested it though
2019-10-10 13:01:39 <PICCORO> ncopa:  i'll try when arrives at home..  thanks
2019-10-10 13:02:08 <ncopa> if you do: apk add /home/cvh/repo/armv7/linux-vanilla-4.19.76-r0.apk
2019-10-10 13:02:19 <mps> CvH: didn't helped, I have linux-vanilla-5.3.5-r0.apk in local repo but it installed 4.19.76 kernel from mirror
2019-10-10 13:02:57 <ncopa> you get in your /etc/apk/world a line with linux-vanilla><$hash
2019-10-10 13:03:35 <ncopa> you can in thery specify the ><$hash to force apk to pick that specific package
2019-10-10 13:05:19 <mps> PICCORO: this about 'user' is debated for decades and without final consensus or conclusions what is 'user'
2019-10-10 13:09:58 <PICCORO> well .. as i can see it! in mint or winbuntu it's just a person without enoght knowledge to understand if must activate or not that! and in alpine it's a person with some knowledge but little time to make it enought perfect so that's the problem!
2019-10-10 13:10:24 <PICCORO> in our context we do not have enought time to make the things we dessire!
2019-10-10 13:10:42 <PICCORO> in my possition i dont know if i really need in fact a lock down in the kernel!
2019-10-10 13:11:01 <PICCORO> i nerver have security problem in alpine neither debian!
2019-10-10 13:11:28 <PICCORO> i think that's it's a stupid desktop feature.. from gnome-related guys and redhat/systemd infection!
2019-10-10 13:11:50 <PICCORO> ah! ncopa seem works tjhanks!
2019-10-10 13:11:51 <AinNero> its for TPM/uefi
2019-10-10 13:11:52 <mps> I don't want it enabled by default
2019-10-10 13:12:08 <PICCORO> mps: i dont know if are necesary!
2019-10-10 13:12:17 <AinNero> a locked down kernel makes sense if you have like, corporate laptops that are corporate-managed linux
2019-10-10 13:12:53 <mps> option will be there and those who want it could enable it in kernel cmd line
2019-10-10 13:13:09 <Cogitri> TPM is nice for other things like verifying boot and keys and others too
2019-10-10 13:13:29 <mps> if you trust TPM vendor
2019-10-10 13:13:49 <PICCORO> AinNero:  as i said! stupid desktop feature.. from gnome-related guys and redhat/systemd infection! you seem confirmed!
2019-10-10 13:13:56 <PICCORO> * seems
2019-10-10 13:17:31 <Cogitri> wat
2019-10-10 13:19:22 <AinNero> its a thing from microsoft
2019-10-10 13:19:31 <AinNero> EFI is essentially windows boot
2019-10-10 13:19:37 <PICCORO> OF COURSE!
2019-10-10 13:19:52 <AinNero> and GPT partition table share the idiodic mixed endianess from microsoft
2019-10-10 13:20:00 <AinNero> *mixed endianess uuid's
2019-10-10 13:20:15 <PICCORO> https://wiki.alpinelinux.org/wiki/Alpine_and_UEFI
2019-10-10 13:21:09 <AinNero> 
ACTION scratches head

2019-10-10 13:21:09 <PICCORO> uefi explained.. not all are due mycosoft.. old bios are 16 bit so there's some other problems but essentially mayor "features" are windos' shit
2019-10-10 13:22:10 <PICCORO> as i posted in the head of the wiki page: "When something new comes.. all the bad things become good and news become bad!"
2019-10-10 13:23:07 <PICCORO> As example: a top of the range Skylake i7-6700k still has an 80286  embedded in it - for the people who insist on using BIOS. Indeed. But  BIOS strictly only supports 16-bit 8088-derivative processors. All of this crap are just due redmon's company operating system  limitations in fact! due intel, amd and others can easyle provide for  many more years an 80286 embedded in each processor
2019-10-10 13:30:38 <TBB> for what it's worth, I've used EFI systems before Microsoft had any support for it
2019-10-10 13:30:59 <TBB> big HP iron and Apple were both there way before
2019-10-10 13:36:20 <AinNero> ls
2019-10-10 13:36:25 <AinNero> oops
2019-10-10 13:39:33 <CvH> ncopa: so it is basically not possible to build alpine with an changed kernel ? what about creating an local repo with changed packages
2019-10-10 13:39:41 <CvH> sadly it errors too
2019-10-10 13:40:09 <CvH> `ERROR: busybox-initscripts-3.1-r7: package mentioned in index not found (try 'apk update')`
2019-10-10 13:49:49 <qa3Txu0iak0F> CvH just install it directly with `apk add --allow-untrusted <pathto>/some.apk`
2019-10-10 13:50:21 <CvH> i don't want to install the package, I want to create an "image"
2019-10-10 13:50:41 <CvH> `./scripts/mkimage.sh --arch armv7 --profile uboot`
2019-10-10 13:50:58 <qa3Txu0iak0F> ah. sorry. didn't read back far enough
2019-10-10 13:51:15 <CvH> np :)  any feedback is welcome
2019-10-10 13:51:19 <qa3Txu0iak0F> you might need to setup your own repo then for this one package
2019-10-10 13:51:30 <CvH> yea this won't work too
2019-10-10 13:51:37 <qa3Txu0iak0F> what do you call this repo?
2019-10-10 13:51:37 <CvH> `ERROR: busybox-initscripts-3.1-r7: package mentioned in index not found (try 'apk update')`
2019-10-10 13:51:51 <qa3Txu0iak0F> don't call it main/community/testing
2019-10-10 13:52:09 <CvH> `./scripts/mkimage.sh --arch armv7 --profile uboot --repository /home/cvh/repo/`
2019-10-10 13:52:31 <qa3Txu0iak0F> oh. ok i'm out, this is above my paygrade. sorry for diversion.
2019-10-10 13:54:27 <PICCORO> CvH: you must tell to apk to trus untrusted repositories
2019-10-10 13:54:37 <PICCORO> * trust
2019-10-10 13:54:57 <CvH> there is no option for that at mkimage.sh
2019-10-10 13:55:35 <CvH> btw maybe it works if i add it at the source commandlines
2019-10-10 13:55:38 <CvH> lets try
2019-10-10 13:59:13 <TBB> maybe the correct way is to add the key you used to package custom stuff into the image itself
2019-10-10 14:02:04 <CvH> `--hostkeys` don't fix it
2019-10-10 14:08:51 <CvH> so manual changing the commandlines is also not working -.-
2019-10-10 14:10:02 <CvH> is there a manual way in creating `initramfs-vanilla`, `modloop-vanilla`, `System.map-vanilla`, `vmlinuz-vanilla` ? basically to workaround the broken buildsystem for now
2019-10-10 14:14:51 <mps> install new kernel and update-kernel should do that automatically if you have mkinitfs installed
2019-10-10 14:17:49 <CvH> hmm thats sounds like an solution,  the amount of workarounds getting quickly out of hand :D
2019-10-10 14:19:26 <mps> that's how I'm testing new kernels and new features enabled in kernel, for now, till I learn how to make uboot tarball
2019-10-10 14:33:51 <CvH> I guess there is nothing to learn, its just plain broken :)
2019-10-10 14:42:51 <ncopa> CvH: can you please file a bug and I'll have a look at it when i get a chance
2019-10-10 14:43:25 <ncopa> this thing should work
2019-10-10 14:43:28 <ncopa> but it doesnt
2019-10-10 14:48:37 <CvH> where is the official bugtracker, github or the selfhosted gitlab ?
2019-10-10 14:56:06 <ncopa> gitlab
2019-10-10 15:07:45 <mps> CvH: I'm sure it can be fixed or we don't understand how it works
2019-10-10 15:08:09 <CvH> if nobody knows how it works i guess it is broken :)
2019-10-10 15:08:23 <CvH> fix looks easy tbh but I have sadly no idea about alpine internals
2019-10-10 15:09:46 <mps> I don't agree with this stance, for example it took me few days with --workdir will ncopa told that I should use absolute path
2019-10-10 15:11:01 <mps> and we all have similar experience with a lot of things
2019-10-10 15:18:05 <MY-R> nicolaus, about your second comment from https://gitlab.alpinelinux.org/alpine/aports/issues/10863
2019-10-10 15:18:39 <MY-R> nicolaus, set pid and log paths in config to: logfile=/var/log/murmur.log pidfile=/var/run/murmur/murmur.pid
2019-10-10 15:19:48 <MY-R> nicolaus, if somebody will fix APKBUILD then it will be done automaticly
2019-10-10 15:40:00 <CvH> mps: if you need a workaround for a workaround to make something somehow work it is likely not working correctly :)  ofc you can also say it is not supported
2019-10-10 15:45:35 <mps> that's not my POV
2019-10-10 19:45:20 <gillesmauve> How do I run multiple monitors with alpine?
2019-10-10 19:45:27 <gillesmauve> I'm on a thinkpad with a dsub connector on the side.
2019-10-10 19:45:57 <_ikke_> I've no experience with it on Alpine, but did you look at xrandr?
2019-10-10 19:46:15 <gillesmauve> Oh, is that how you usually do it?
2019-10-10 19:46:30 <_ikke_> That's how *I* usually do it :)
2019-10-10 19:46:44 <_ikke_> Some DMs have their own tools, but xrandr should work in any case
2019-10-10 19:47:06 <nicolaus> you can use xrandr or arandr which is the gui program for it
2019-10-10 19:47:35 <gillesmauve> Interesting. I've got bspwm and polybar, but I'm fine just moving windows over to the other monitor.
2019-10-10 19:47:59 <gillesmauve> It claims that VGA1 is disconnected. Weird.
2019-10-10 19:48:09 <gillesmauve> Oh, maybe I have to use the one on the back of the dock. Sec.
2019-10-10 19:48:51 <gillesmauve>  Sweet. Now it's connected. Back to docs. (Yeah, had to plug it in on the back)
2019-10-10 19:49:58 <_ikke_> This is an example I use: xrandr --output DVI-0 --mode 1680x1050 --output DVI-1 --mode 1920x1080 --right-of DVI-0
2019-10-10 19:50:15 <gillesmauve> DVI-0 being your actual DVI port, I assume?
2019-10-10 19:50:25 <_ikke_> Yes, in this case it's not a laptop
2019-10-10 19:51:23 <gillesmauve> Ah, so I'd have to use LVDS1 and VGA1, since those are the only two connected screens. Is there a way I can tell xrandr to use the native resolutions for each screen that's available and assign one to each port without manually specifying it?
2019-10-10 19:51:33 <mps> here is my 'xrandr --output LVDS1 --auto --output VGA1 --auto --left-of LVDS1'
2019-10-10 19:51:44 <gillesmauve> mps: Oh, just like that. Loller.
2019-10-10 19:51:56 <_ikke_> Yeah, in my case I have to set the resolution for some reason
2019-10-10 19:52:08 <gillesmauve> Now let's see if this screen wakes the hell up
2019-10-10 19:52:31 <gillesmauve> How the hell does bspwm play with this?
2019-10-10 19:53:08 <mps> don't used bspwm but awesome wm do that automatically
2019-10-10 19:53:17 <_ikke_> yeah, similar experience ;-)(
2019-10-10 19:53:47 <gillesmauve> lmao
2019-10-10 19:54:37 <gillesmauve> Ah, looks like I have to set up another monitor specification and reload
2019-10-10 19:56:30 <gillesmauve> Oh, maybe bspc will let me.
2019-10-10 19:59:51 <gillesmauve> Dood, it's so easy. The FUCK
2019-10-10 20:00:05 <_ikke_> :)
2019-10-10 20:06:41 <gillesmauve> I was wondering why I couldn't drag my firefox window over to the other instance, but I then realized that I was just being a moron.
2019-10-10 20:15:02 <AndyAndyBoBandy> Why do posix man pages from kernel.org (e.g. awk), at least as packaged by say alpine or arch, use '−' (U+2212 MINUS SIGN) for flags, when those are not valid to use in invocation, are not found on keyboards, and also make searching the page inconvenient? What's the advantage over using the actual '-' (U+002D HYPHEN-MINUS)?
2019-10-10 20:16:19 <_ikke_> AndyAndyBoBandy: I think you have to ask the authors of those man pages
2019-10-10 20:17:06 <shrizza> Anyone know what this is when I do apk fix? (1/1) [APK unavailable, skipped] Reinstalling man-db (2.8.5-r0)
2019-10-10 20:17:56 <shrizza> I searched around on the net but couldn't find any similar issue.
2019-10-10 20:21:05 <mps> shrizza: do you testing repo in /etc/apk/repositories
2019-10-10 20:21:08 <shrizza> Hm, nevermind. I found that there was some contention over /usr/bin/man between man-db and mdocml.
2019-10-10 20:21:22 <shrizza> Yeah.
2019-10-10 20:21:39 <mps> yes, mdocml is better supported on alpine
2019-10-10 20:22:41 <AndyAndyBoBandy> _ikke_: according to my chat in ##posix, the authors of those man page authors don't use u+2212, only ascii. Do you know if kernel.org hosts modified versions, or if it's something about the rendering process?
2019-10-10 20:24:12 <_ikke_> AndyAndyBoBandy: no, I have no idea myself
2019-10-10 20:25:06 <gillesmauve> What character is it on your local machine, AndyAndyBoBandy, (inside of the manpages)?
2019-10-10 20:25:58 <_ikke_> I verified it's not a hyphen on my machine (arch linux, man awk)
2019-10-10 20:30:37 <AndyAndyBoBandy> _ikke_: '−' (U+2212 MINUS SIGN)
2019-10-10 20:52:36 <gillesmauve> When I unplug that monitor, can I just tell xrandr and bspwm to reconfigure based on my old setup?
2019-10-10 20:53:09 <gillesmauve> I'm still not even certain why AndyAndyBoBandy is asking this question.
2019-10-10 20:56:00 <mps> xrandr --output VGA-1 --off
2019-10-10 20:58:32 <gillesmauve> Ah, nifty.
2019-10-10 20:58:46 <gillesmauve> Thanks, mps.
2019-10-10 20:58:56 <mps> gillesmauve: np
2019-10-11 01:07:47 <AndyAndyBoBandy> gillesmauve: asking the question so that 1. I may learn the answer and 2. I may rectify the problem on my own systems, as I can think of no sane reason for the current state of things and haven't heard any offered
2019-10-11 16:22:32 <jlieberg> Hello! I have a hopefully quick question. I'm trying to compile postgis from source and am getting the error `could not find libproj - you may need to specify the directory of a PROJ.4 installation using --with-projdir`. Was the proj-dev package updated recently to move the files, or has something else broken? proj-dev is currently on the edge
2019-10-11 16:22:32 <jlieberg> branch, so I expect occasional issues but I can't figure this one out
2019-10-11 16:24:09 <jlieberg> (I also previously had issues with the protobuf package, so I had to use the `--without-protobuf` when configuring while building)
2019-10-11 17:19:28 <javashin> hi
2019-10-11 17:19:42 <javashin> flatpak works ? using openrc ?
2019-10-11 17:25:46 <javashin> i do not have sys-apps/bubblewrap installed
2019-10-11 19:10:43 <tmhoang> javashin: alpine not gentoo ?
2019-10-11 19:11:23 <javashin> i fixed everything thanks
2019-10-11 19:11:28 <javashin> for flatpaks
2019-10-11 19:11:57 <javashin> i patched the version 3.3 with patches from void
2019-10-11 19:12:26 <javashin> replaced bubblewrap  on flatpak with the new one
2019-10-11 19:12:28 <javashin> fixed
2019-10-11 19:28:14 <tmhoang> +1
2019-10-11 23:11:35 <strfry> AinNero: i got this question at my talk today: "does alpine have reproducible builds?". my follow-up to that is, are the exact setups of the official alpine build server document, and themselves reproducible?
2019-10-11 23:19:20 <mps> strfry: short answer no, but there are some work on this
2019-10-11 23:27:57 <strfry> i'm currently reading backlog, and that discussion about an OBS-style build server reminded me
2019-10-11 23:29:00 <strfry> there still is quite a gap between abuild being rather easy to use, and having a reproducible setup that will convert your APKBUILD repo to installable packages ;)
2019-10-12 02:34:08 <tommyrot1977> hi folks, loving alpine linux so far. I've got an issue with installing a specific package on my raspberry pi however. getting this error when I attempt to install xbindkeys ERROR: unsatisfiable constraints:
2019-10-12 02:34:09 <tommyrot1977>   xbindkeys (missing):
2019-10-12 02:34:09 <tommyrot1977>     required by: world[xbindkeys]
2019-10-12 02:36:01 <tommyrot1977> I should add that I've been able to install other packages (eg. the links browser) with no issues and the xbindkeys appears to be in the package list for my pi architecture (https://pkgs.alpinelinux.org/packages?name=xbindkeys&branch=edge&arch=armhf)
2019-10-12 04:03:17 <maxice8> tommyrot1977: did you add the community repository ?
2019-10-12 06:38:50 <AinNero> ufff
2019-10-12 06:38:54 <AinNero> good morning str
2019-10-12 06:39:04 <AinNero> oh he left?
2019-10-12 16:07:07 <doggone> Is anyone here using dovecot on alpine?
2019-10-12 16:08:56 <AinNero> yes
2019-10-12 16:09:47 <doggone> if possible, could you test restarting the service with ssh as such: ssh user@server service dovecot restart
2019-10-12 16:10:05 <doggone> this hangs the ssh session for me, as well as: echo service dovecot restart | ssh -T user@server
2019-10-12 16:10:21 <doggone> although it does restart dovecot without issues.
2019-10-12 16:10:36 <AinNero> i cannot test it like that, root login is not allowed in general here..
2019-10-12 16:10:45 <AinNero> (via SSH)
2019-10-12 16:11:11 <doggone> ah, thanks anyway
2019-10-12 16:11:39 <doggone> I use a script to deploy my LetsEncrypt certs to different servers, so I need to restart the services automatically
2019-10-12 19:31:51 <circ-user-jYvUq> nick arash
2019-10-13 16:18:14 <vsc> Hello #alpine-linux, I was wondering if I could get some assistance. I was installing alpine on my machine for desktop use, and for some reason no browser seems to work out of the box. I've tried chromium, firefox, midori, adn surf, but all seem to have a flickering effect and don't appear to allow for user interaction. Do you know if this is a
2019-10-13 16:18:14 <vsc> known issue and is there presently a workaround?
2019-10-13 16:21:39 <mps> vsc: what WM you use
2019-10-13 16:21:55 <mps> and video driver
2019-10-13 16:22:00 <vsc> mps: I use dwm.
2019-10-13 16:23:57 <mps> sometimes I have similar effect with awesome WM and firefox, but restarting firefox solve it for some time
2019-10-13 16:24:39 <mps> most of that mainstream browsers are not made to work well with tilling WM's
2019-10-13 16:25:09 <vsc> mps: As for video driver I'm not sure, and I believe that may be the source of the problem. I've installed `xf86-video-intel`, but the problem seems to have persisted.
2019-10-13 16:25:59 <mps> do you have installed linux-firmware-i915
2019-10-13 16:26:12 <mps> and linux-firmware-intel
2019-10-13 16:27:06 <vsc> Ahh those I've have not added unless they are installed by default via the setup-alpine or setup-xorg utilities. Let me see about adding those.
2019-10-13 16:27:34 <mps> under xfce firefox and chromium works well on two other notebooks of my familly members
2019-10-13 16:38:40 <_ikke_> mps: I don't have a lot of issues with the browser on arch + awesome
2019-10-13 16:38:47 <_ikke_> (re tiling)
2019-10-13 16:46:02 <mps> also I didn't had, but in previous week I had to restart firefox two times because main menu and popup menu 'flicks' out (disappears instantly) when I click 'for them'
2019-10-13 16:46:55 <mps> and that also happened long time ago when I used debian
2019-10-13 16:47:21 <mps> _ikke_: you switched to alpine as workstation
2019-10-13 16:51:10 <_ikke_> not yet
2019-10-13 17:00:13 <pavan> hey everyone just downloaded virt image of alpine and after selecting sys it says installing system what is going on?
2019-10-13 17:00:38 <pavan> How can it be only 35 mb and do all this?
2019-10-13 17:01:34 <_ikke_> pavan: there is no kernel / firmware included
2019-10-13 17:21:29 <qa3Txu0iak0F> i've never had any problem with i3wm and firefox/chromium
2019-10-13 17:21:48 <qa3Txu0iak0F> (re tiling)
2019-10-13 17:22:18 <mps> also I didn't had it on alpine till last week
2019-10-13 17:23:35 <mps> and only on x86_64 box, on aarch64 it still works quite fine
2019-10-13 17:24:08 <mps> and awesome config is nearly same
2019-10-13 17:38:47 <priontology> Why do I typically get the same IP address on the same network every time I run udhcpc?
2019-10-13 17:39:10 <priontology> Is the request for that address generated from my side, or the dhcp server's side?
2019-10-13 17:39:45 <_ikke_> priontology: dhcp servers usually keep a reservation by mac-address
2019-10-13 17:40:03 <mps> priontology: dhcp servers tend to remember previously assigned IP and resend it again for same clients
2019-10-13 17:40:10 <priontology> ah, i see
2019-10-13 17:40:14 <_ikke_> And asside from that, a client can try to request the same address again
2019-10-13 17:40:42 <priontology> Actually, I've been curious about this. Is there an easy way to scramble my mac on every boot?
2019-10-13 17:41:53 <_ikke_> I would not know any solution for that
2019-10-13 17:41:58 <mps> randmize it by using some pkg's, forgot exact name
2019-10-13 17:42:05 <priontology> Was thinking I could do something like macchanger on boot
2019-10-13 17:42:43 <mps> for example iwd have option to randomiza MAC on every dhcp request
2019-10-13 17:44:12 <priontology> Oh, nifty. I didn't know about that.
2019-10-13 17:44:16 <qa3Txu0iak0F> machchanger
2019-10-13 17:44:27 <qa3Txu0iak0F> but you can also set the hwaddress with ifconfig/ip
2019-10-13 17:44:38 <KatolaZ> yeah but the actual assignment still depends on the server
2019-10-13 17:44:45 <KatolaZ> unless the client asks for a specific IP
2019-10-13 17:45:14 <KatolaZ> i.e., if you are the first client asking for a lease, you might always get the first available IP
2019-10-13 17:45:14 <priontology> That's what I was most curious about. Interesting.
2019-10-13 17:45:38 <KatolaZ> then the server might or might not decide to reuse the same lease for the same MAC
2019-10-13 17:50:46 <TBB> I imagine they have a preference for that, actually. Only reassign an IP when out of "free" addresses.
2019-10-13 17:51:44 <priontology> Huh. What response does a server give when its pool is exhausted?
2019-10-13 17:52:26 <_ikke_> It might not give one at all
2019-10-13 18:02:56 <KatolaZ> priontology: it might just ignore any DHCPDISCOVER
2019-10-13 18:03:33 <KatolaZ> so the client won't make a DHCPREQUEST
2019-10-13 22:43:10 <ndowens> Does anyone have a guide to a manual zfs install with gpt. Followed a guide on the wiki and when I boot it, dmesg says didn’t load zfs module and ofc causes boot to fail
2019-10-13 22:53:14 <nicolaus> is there any pastebin service in the alpine repo so I can host my own?
2019-10-13 22:54:30 <qa3Txu0iak0F> ndowens: you want to add zfs as a feature to /etc/mkinitfs/mkinitfs.conf
2019-10-13 22:54:36 <qa3Txu0iak0F> and then rerun mkinitfs
2019-10-13 22:55:00 <ndowens> I believe it was there but going to recheck
2019-10-13 22:56:00 <ndowens> Yup it is
2019-10-13 22:56:52 <qa3Txu0iak0F> and did you rebuild your initfs with mkinitfs?
2019-10-13 22:57:11 <ndowens> Previously yes but doing it again.
2019-10-13 22:57:16 <qa3Txu0iak0F> and if boot fails, doesn't it drop you into a shell?
2019-10-13 22:57:34 <qa3Txu0iak0F> can you try in that shell if you can load it manually?
2019-10-13 23:04:36 <ndowens> It will let me modprobe zfs fine in shell after fail
2019-10-13 23:06:28 <qa3Txu0iak0F> hmmm. that is strange
2019-10-13 23:07:38 <ndowens> Here is my setup. I created a separate boot partition as setup-disk complains zfs isn’t supported unless I had a compatible FS mounted inside /mnt I am also using rEFInd
2019-10-13 23:10:13 <ndowens> Got it
2019-10-13 23:10:47 <ndowens> Had rEFInd not setup correctly
2019-10-14 10:30:44 <kalamaki> mc
2019-10-14 12:05:12 <AinNero> heh
2019-10-14 12:05:13 <AinNero> https://f0ck.me/b/fc2c6760.png
2019-10-14 12:05:23 <AinNero> the site is NSFW but this post is SFW and alpine-related
2019-10-14 15:01:13 <KH405> Should I learn busybox-syslog? I used to use syslog-NG. will I have problem switching from one to another?
2019-10-14 15:03:30 <KH405> And also AinNero, Linux uses some GNU code so yeah ...
2019-10-14 15:06:19 <AinNero> neither apk, musl, busybox nor linux itself are GNU projects, which constitutes the alpine linux core
2019-10-14 15:07:59 <AinNero> a
2019-10-14 15:08:11 <AinNero> ... and 'uname -o' does speak the truth :D
2019-10-14 15:09:45 <KH405> Maybe you're right, but I though there was parts from GNU in the linux kernel
2019-10-14 15:11:07 <qa3Txu0iak0F> actually apk uses gnu-tar...
2019-10-14 15:11:45 <KH405> also just try cp --version for me
2019-10-14 15:13:09 <AinNero> qa3Txu0iak0F: no? apk uses an own tar implementation, see apk-tools -> src/archive.c
2019-10-14 15:13:39 <AinNero> KH405: if you see GNU coreutils there, you have coreutils installed over busybox
2019-10-14 15:14:05 <qa3Txu0iak0F> sorry i meant abuild.
2019-10-14 15:15:18 <KH405> Anyway, what do you guys think about busybox syslog
2019-10-14 15:15:27 <KH405> Can it replace 100% something like syslog-ng ?
2019-10-14 15:15:57 <AinNero> KH405: can busybox X 100% replace X? usually never
2019-10-14 15:16:15 <AinNero> busybox syslogd can send the data via network, but the format is not configurable
2019-10-14 15:16:18 <AinNero> and does not include the hostname
2019-10-14 15:16:35 <AinNero> which restricts its usuability to local networking
2019-10-14 15:17:08 <AinNero> so, it works fine for basic needs
2019-10-14 15:17:20 <AinNero> but as soon as you need something special, you need rsyslog or syslog-ng
2019-10-14 15:18:43 <KH405> Thanks AinNero, really appreciate your input :)
2019-10-14 15:19:21 <AinNero> qa3Txu0iak0F: consider how many alpine machines have abuild installed^^
2019-10-14 15:26:59 <qa3Txu0iak0F> 100% of those who build official or inhouse packages
2019-10-14 15:35:32 <AinNero> qa3Txu0iak0F: on every machine?
2019-10-14 15:35:41 <AinNero> for me, alpine runs in docker containers and servers
2019-10-14 15:36:45 <PureTryOut[m]> Not on your laptop/desktop? 😢
2019-10-14 15:37:48 <AinNero> X is a server for me :P
2019-10-14 15:38:20 <PureTryOut[m]> Btw, GCC is GNU, and is used to compile like 95% of Alpine's packaging
2019-10-14 15:38:35 <PureTryOut[m]> Yeah not GNU/Linux, but I wouldn't say it doesn't use GNU code either
2019-10-14 15:38:55 <AinNero> i didn't claim otherwise
2019-10-14 15:39:10 <AinNero> > neither apk, musl, busybox nor linux itself are GNU projects, which constitutes the alpine linux core
2019-10-14 15:39:34 <AinNero> ah, i see, but the meme claimed
2019-10-14 15:40:09 <PureTryOut[m]> Yes
2019-10-14 17:03:25 <mps> hmm, on 3.10 'apk add openssl-dev' shows 'ERROR: Failed to create usr/lib/libcrypto.a: I/O error'
2019-10-14 17:04:23 <mps> nvm, apk-cache is broken
2019-10-14 17:31:32 <mps> MY-R: nicolaus: mumble is upgraded in 3.10-stable and edge. please test it and report if it works or not
2019-10-14 20:37:41 <mps> musl 1.1.24 is released
2019-10-14 20:44:07 <tomponline> sudo might need updating: https://thehackernews.com/2019/10/linux-sudo-run-as-root-flaw.html
2019-10-14 20:44:27 <tomponline> ah its been flagged already :)
2019-10-14 20:49:31 <_ikke_> tomponline: Maybe you could open an issue for that on gitlab.a.o?
2019-10-14 20:50:26 <tomponline> _ikke_: will do
2019-10-14 20:50:34 <_ikke_> thanks!
2019-10-14 20:50:41 <mps> do we have runas
2019-10-14 20:50:45 <_ikke_> yes
2019-10-14 20:50:52 <_ikke_> hmm
2019-10-14 20:50:53 <_ikke_> no
2019-10-14 20:50:57 <_ikke_> we have doas
2019-10-14 20:51:05 <mps> yes
2019-10-14 20:51:35 <mps> I looked this news about sudo and thought that we are safe
2019-10-14 20:52:24 <tomponline> _ikke_: shall I still create an issue, if so, which project in gitlab do you want it under?
2019-10-14 20:52:33 <_ikke_> tomponline: under aports
2019-10-14 20:52:47 <_ikke_> and yes, it's still necessary, as alpine still uses / offers sudo
2019-10-14 20:52:53 <tomponline> will do thanks
2019-10-14 20:55:03 <mps> anyway, it will not hurt to add this fix
2019-10-14 20:55:08 <tomponline> https://gitlab.alpinelinux.org/alpine/aports/issues/10874
2019-10-14 21:43:12 <ahills> resolvconf does this weird thing where it tries to execute every single executable in /usr/lib as a script when it runs?
2019-10-14 21:43:28 <ahills> this is because LIBEXECDIR is set to /usr/lib, and that's where it looks…
2019-10-14 21:43:54 <ahills> line 20 of aports/main/openresolv/APKBUILD:    --libexecdir=/usr/lib \
2019-10-14 21:44:24 <ahills> and line 93 of /usr/sbin/resolvconf: for script in "$LIBEXECDIR"/*; do
2019-10-14 21:44:54 <ahills> line 923*
2019-10-14 21:45:16 <ahills> on line 931 you can see it simply execute blindly:       "$script" "$cmd" "$iface"
2019-10-14 21:45:30 <_ikke_> does $LIBEXECDIR/resolvconf exist?
2019-10-14 21:45:41 <ahills> nope
2019-10-14 21:45:57 <_ikke_> I wonder what it's expected to exeute then?
2019-10-14 21:46:15 <ahills> openresolv does include some stuff in /usr/lib
2019-10-14 21:46:23 <ahills> damed, dnsmasq, etc
2019-10-14 21:46:30 <ahills> but one would not want to execute those in this manner either
2019-10-14 21:46:39 <ahills> so I think /usr/lib is not supposed to be the value of libexecdir
2019-10-14 21:47:22 <ahills> in the Makefile, it defaults to "/libexec/resolvconf"
2019-10-14 21:47:52 <ahills> perhaps its for other packages to integrate?
2019-10-14 21:47:59 <ahills> it's*
2019-10-14 21:48:24 <_ikke_> /usr/lib/resolvconf would make sense
2019-10-14 21:48:28 <ahills> maybe the openresolv package could own /usr/libexec/resolvconf
2019-10-14 21:48:32 <ahills> sure, or that
2019-10-14 21:48:37 <_ikke_> yeah, either way
2019-10-14 21:48:58 <mps> ahills: it is fixed in edge
2019-10-14 21:50:23 <ahills> mps: so it is; this fix seems easy to backport to 3.10, is that possible?
2019-10-14 21:50:41 <ahills> one-line change in APKBUILD
2019-10-14 21:50:47 <_ikke_> Sounds like it should be possible
2019-10-14 21:51:27 <mps> yes, I though at that time when made fix to backport it but forgot
2019-10-14 21:51:43 <ahills> great, what can I do to get the backport in?
2019-10-14 21:51:51 <mps> hope will not forgot tomorrow
2019-10-14 21:52:19 <mps> if you fill issue on gitlab.a.o it could help
2019-10-14 21:52:20 <_ikke_> ahills: Make a merge request against 3.10 on aports
2019-10-14 21:52:49 <_ikke_> ahills: (don't forget to bump pkgrel)
2019-10-14 21:52:51 <ahills> is gitlab.a.o the canonical repository?
2019-10-14 21:53:20 <_ikke_> ahills: git.a.o is the cannonical source, but gitlab.a.o is synced
2019-10-14 21:53:27 <ahills> ok
2019-10-14 22:16:32 <mps> ahills: I posted it as !934
2019-10-14 22:16:59 <mps> !472
2019-10-14 22:17:42 <ahills> thank you!
2019-10-14 22:18:50 <mps> np, actually I had to clean some mess I made on gitlab, but pshhhh, _ikke_ shouldn't hear ;)
2019-10-14 22:19:11 <_ikke_> 
ACTION pretends he didn't hear

2019-10-14 22:19:31 <mps> :D
2019-10-14 22:21:50 <_ikke_> mps: If you have a bit patience, we will have armv7 support on gitlab soon
2019-10-14 22:23:53 <mps> np, I know how hard you both works to setup all these things
2019-10-14 22:24:44 <mps> and, good night to all, time for bed
2019-10-14 22:24:51 <_ikke_> good night
2019-10-15 06:10:08 <ncopa> any volunteer to create favicons for our different sites? https://gitlab.alpinelinux.org/alpine/infra/infra/issues/10640
2019-10-15 06:10:21 <ncopa> its a simple job
2019-10-15 06:12:34 <oliv3r[m]> I can take a look at it tonight
2019-10-15 06:12:55 <oliv3r[m]> what repo would these icons/sources be stored normally?
2019-10-15 06:13:26 <oliv3r[m]> thinking that this may be scriptable (use the main svg; find/replace the color, generate icon from svg
2019-10-15 06:14:39 <mps> oliv3r[m]: +1 for scriptable
2019-10-15 06:15:22 <maldridge> ncopa: I opened an issue some time ago in the infrastructure repo about upstreaming work my company did to have both amazon machine images and vagrant VMs for alpine, do you think this is something alpine would be interested in?
2019-10-15 06:17:59 <ncopa> maldridge: possibly
2019-10-15 06:18:17 <maldridge> https://gitlab.alpinelinux.org/alpine/infra/infra/issues/10638 <- context
2019-10-15 06:18:43 <ncopa> seems like not many people have asked for it (yet)
2019-10-15 06:19:08 <ncopa> and we are kind of struggeling keeping our head over the surface with the normal maintenance
2019-10-15 06:19:18 <maldridge> anything I can do to help?
2019-10-15 06:19:23 <maldridge> my org runs alpine at scale
2019-10-15 06:19:39 <ncopa> so i dont know if we have capacity to add stuff that requires maintenance of us
2019-10-15 06:20:03 <maldridge> I don't have much capacity beyond my work with void, but I'm happy to upstream whatever I can to lighten the load of running the fleet
2019-10-15 06:21:06 <ncopa> thanks, i dont have any specific things in mind atm that we need help with
2019-10-15 06:21:16 <ncopa> i guess respond to tickets may help, and review patches
2019-10-15 06:21:41 <maldridge> well, if I have bandwidth I'll see what I can do there
2019-10-15 11:38:32 <mps> dalias: you can save me some time if you tell me is there difference in SIGSYS behavior in musl from glibc
2019-10-15 13:05:25 <dalias> mps, ?
2019-10-15 13:05:37 <dalias> mps, what do you mean by "SIGSYS behavior" ?
2019-10-15 13:07:36 <Cogitri> Apparently the sudo tests are failing
2019-10-15 13:16:44 <maribu> Hi. In branch "kicad" in https://gitlab.alpinelinux.org/maribu/aports I have added APKBUILDS for the KiCad EDA suite and its dependencies. I would like to create a merge request to alpine/aports. But the web interface does not provide the "new merge request" button in https://gitlab.alpinelinux.org/alpine/aports/merge_requests
2019-10-15 13:17:28 <maribu> Am I missing privileges, or is something wrong with my repo in maribu/aports, so that GitLab is not able to create merge requests?
2019-10-15 13:18:05 <Cogitri> <maribu "Hi. In branch "kicad" in https:/"> You have to make a Merge Request from https://gitlab.alpinelinux.org/maribu/aports/merge_requests IIRC
2019-10-15 13:18:36 <Cogitri> It's a bit confusing. GitLab also prints a link to directly open a MR when you push to a fork's branch that isn't master too though
2019-10-15 13:18:45 <Cogitri> (also, sorry for the quote, didn't mean to do that)
2019-10-15 13:19:47 <maribu> If I do so, I'm only able to create a merge request for maribu/aports (master branch) :-/
2019-10-15 13:20:25 <Cogitri> https://docs.gitlab.com/ee/gitlab-basics/add-merge-request.html
2019-10-15 13:21:31 <dalias> cogitri, details?
2019-10-15 13:21:41 <maribu> OK. I guess the issue that GitLab didn't identify maribu/aports as fork of alpine/aports. I guess I have to re-create the repo...
2019-10-15 13:22:25 <Cogitri> Ah yes, you need to explicitly fork it and not just create a new repo
2019-10-15 13:24:49 <Cogitri> dalias: mps said: `building sudo 1.8.28 I got 'strsig_test: FAIL: str2sig(SIGSYS): -1 != 0'. we lost SIGSYS, or what`
2019-10-15 13:24:56 <Cogitri> Not sure of the details though
2019-10-15 13:26:23 <mps> dalias: sry, I was afk. I had to patch sudo 1.8.28 to pass test, here is patch http://tpaste.us/barZ
2019-10-15 13:26:42 <mps> maribu: nice to see you here
2019-10-15 13:26:54 <maribu> @Coqitri: Thanks for your help! I think I should be able to open the Merge Request now. (But I have to wait for a build to complete before I can switch branches and push to the now correctly forked repo.)
2019-10-15 13:28:43 <dalias> mps, that looks odd
2019-10-15 13:29:15 <mps> and here is file from sudo I patched http://tpaste.us/7vrw
2019-10-15 13:30:47 <mps> now I will rebuild it again and post error message
2019-10-15 13:31:02 <dalias> what is sudo_sig2str ?
2019-10-15 13:31:28 <mps> here is err 'strsig_test: FAIL: str2sig(SIGSYS): -1 != 0'
2019-10-15 13:32:01 <dalias> yes but the function that's failing is something sudo itself defined
2019-10-15 13:32:56 <mps> yes, and before asking them I just wanted to hear what you think about that
2019-10-15 13:33:11 <Cogitri> maribu: Feel free to ask if you need more help and thanks for your contribution :)
2019-10-15 13:33:24 <dalias> mps, well unless i see that function i have no idea what to think
2019-10-15 13:33:35 <dalias> and apparently sudo is 80s-era crap with no website or git repo...
2019-10-15 13:35:40 <mps> include/sudo_compat.h, http://tpaste.us/pWjY
2019-10-15 13:35:50 <dalias> ok it has a website but its like bottom of search results or something >_<
2019-10-15 13:36:02 <mps> site is fast, but like snail :)
2019-10-15 13:36:26 <dalias> mps, that's not anything. it's just a header
2019-10-15 13:36:56 <dalias> need the .c file defining the function
2019-10-15 13:37:21 <maribu> @mps, @Coqitri: Thanks for the warm welcome :-)
2019-10-15 13:37:29 <mps> lib/util/str2sig.c is here http://tpaste.us/gMLx
2019-10-15 13:38:38 <mps> and I agree that sudo should be retired
2019-10-15 13:39:06 <dalias> ok this is sudo's fault
2019-10-15 13:39:33 <mps> dalias: thank for confirming my thinking
2019-10-15 13:40:03 <dalias> musl does not have the broken legacy extern const char *const sys_signame[]; array
2019-10-15 13:40:19 <mps> and thanks for hint
2019-10-15 13:40:29 <dalias> so they use their own:
2019-10-15 13:40:30 <dalias> extern const char *const sudo_sys_signame[NSIG];
2019-10-15 13:40:36 <dalias> and apparently theirs is incomplete
2019-10-15 13:40:45 <dalias> (need to go look and see what file it's defined in)
2019-10-15 13:43:32 <mps> lib/util/signame.c , here http://tpaste.us/ee9X
2019-10-15 13:44:31 <mps> SYS is not there
2019-10-15 13:46:15 <mps> and, my patch is good, despite I wandered in dark
2019-10-15 13:46:34 <mps> s/despite/despite the fact/
2019-10-15 13:48:45 <dalias> UNUSED is SYS
2019-10-15 13:49:19 <dalias> SYS is the correct naem for it; UNUSED should be an alias in the previous file you showed me
2019-10-15 13:49:23 <dalias> however
2019-10-15 13:49:29 <dalias> this whole thing is utterly broken
2019-10-15 13:49:33 <mps> yes, enumerated, but why they didn't named it correctly
2019-10-15 13:49:36 <dalias> it assumes the x86 order for signal numbers
2019-10-15 13:49:55 <dalias> the whole thing needs to be an array with designated initializers
2019-10-15 13:49:56 <dalias> like
2019-10-15 13:50:11 <dalias> [SIGHUP] = "HUP", [SIGINT] = "INT", ...
2019-10-15 13:50:22 <dalias> this code is just utter shit
2019-10-15 13:50:33 <dalias> and should absolutely never be installed suid anywhere
2019-10-15 13:50:50 <dalias> but everyone teaches everyone to use sudo
2019-10-15 13:50:55 <dalias> despite it being awful and wrong
2019-10-15 13:51:44 <mps> I know, and can't wait when we switch to 'doas' (at least)
2019-10-15 13:52:22 <mps> changing UNUSED to SYS fixes test
2019-10-15 13:53:42 <mps> tested on x86_64, later will on armv7 and aarch64 to see if it pass build
2019-10-15 13:53:53 <mps> dalias: thank you for help
2019-10-15 13:55:04 <dalias> but it's still broken on other archs
2019-10-15 13:55:11 <dalias> that number signals differently
2019-10-15 13:55:27 <dalias> not sure if alpine has any affected ones
2019-10-15 13:56:14 <mps> will look later when I come to home, from where I can access arm builders
2019-10-15 14:01:26 <mps> dalias: FYI, yesterday looked at changes for 5.4-rcX kernel and noticed that they lowered speed for some mmc cards
2019-10-15 14:01:50 <mps> maybe this would help for your problem with mmc
2019-10-15 14:03:16 <mps> clock, not speed, but end result is same, slow is safe ;)
2019-10-15 14:05:26 <AinNero> for mmc-spi, clock and speed are the same anyways?
2019-10-15 14:11:54 <mps> maribu: so I can stop to work on kicad apk :) , thanks
2019-10-15 14:14:56 <maribu> mps: I would love if you could test and review the APKBUILDs I created. So far they seem to work for me, but I could very well missed some dependencies, as my working machine has quite a lot installed...
2019-10-15 14:17:10 <mps> maribu: when you create MR it will go to our (new and shiny) CI which will show if something is bad
2019-10-15 14:45:04 <oliv3r[m]> Hmm, I thought in shell EXIT was 0 i found somewhere
2019-10-15 15:22:07 <tmhoang> maldridge: re amz images : ping or cc me in MR/issue please. I would love to review and/or just thumbs up.
2019-10-15 15:25:24 <tmhoang> my current approach to use Alpine on cloud providers is : setup qemu *raw* image locally, then use a glibc (or rescue shell) system in the cloud provider, simply dd if=alpine-raw.img of=/dev/sdb , then reboot into /dev/sdb as Alpine root disk. Quite similar to those in Alpine EC2 wiki page, but less interaction/dependencies on remote hosts -
2019-10-15 15:25:24 <tmhoang> because some hosters don't even allow attach another disk or so ...
2019-10-15 15:26:26 <dalias> :-)
2019-10-15 15:26:39 <tmhoang> and dd approach doesn't require Alpin/musl chroot - much simpler. Everything should be configured in local dev machine i.e. running setup-alpine.
2019-10-15 15:26:54 <dalias> i really don't like how so many cloud providers want you to install their OS images
2019-10-15 15:27:29 <tmhoang> it's convenient to use their images - using cloud-init I guess - for some users who like fast food
2019-10-15 15:28:13 <tmhoang> people in this channel eat healthy :P
2019-10-15 15:29:03 <maribu> mps: The KiCad Merge Request is open: https://gitlab.alpinelinux.org/alpine/aports/merge_requests/483
2019-10-15 15:29:54 <mps> tmhoang: fasting is healthy not eating :P
2019-10-15 16:31:59 <maldridge> tmhoang: I linked the issue above
2019-10-15 16:32:40 <tmhoang> maldridge: yeah so if you open MR, please also link that issue too, which is enough I think
2019-10-15 16:39:22 <tmhoang> _ikke_ do you have x86 box I can login ?
2019-10-15 16:41:20 <_ikke_> We run x86_64 with linux32
2019-10-15 16:42:48 <_ikke_> tmhoang: do you have vpn access?
2019-10-15 16:43:11 <tmhoang> so I guess if I prefix all my shell command with "linux32 -c "my command"' on x86_64, I should be ok ?
2019-10-15 16:43:29 <_ikke_> tmhoang: or linux32 /bin/sh
2019-10-15 16:43:39 <tmhoang> ok
2019-10-15 18:58:18 <oliv3r[m]> ncopa !10640 has now a script with it
2019-10-15 18:58:23 <oliv3r[m]> i wonder if the link works properly to go to 'infra' :)
2019-10-15 18:59:35 <oliv3r[m]> lets help out poor algitbot: is what I meant https://gitlab.alpinelinux.org/alpine/infra/infra/issues/10640
2019-10-15 19:01:30 <mps> #10640
2019-10-15 19:02:38 <rsmarples> hello!
2019-10-15 19:02:46 <nepugia> Hi
2019-10-15 19:03:02 <rsmarples> I'm trying to use gcc-9.2 with lubsan .... but gcc says it can't find libubsan to link against
2019-10-15 19:03:05 <mps> oliv3r[m]:  '!' is for MR's and '#' for issues numbers
2019-10-15 19:03:06 <rsmarples> what did I miss?
2019-10-15 19:03:47 <nepugia> i would guess the headers? but since i dont know what you actually installed :)
2019-10-15 19:04:21 <mps> rsmarples: do you lib installed
2019-10-15 19:05:41 <rsmarples> gcc-libs (missing):
2019-10-15 19:05:48 <rsmarples> required by: world[gcc-libs]
2019-10-15 19:06:10 <rsmarples> I've set my repo to edge and upgraded
2019-10-15 19:07:30 <mps> do we have gcc-libs package
2019-10-15 19:19:10 <oliv3r[m]> mps; true, but it's from a different repo anyway
2019-10-15 19:21:20 <oliv3r[m]> Is there a go-expert that could give me some insight on !482 ? I am trying to build something that uses cgo, but no matter what I try, it keeps failing (also on the CI). Most probably due to the same reason, I'm running it inside of a docker container, so maybe some binfmt thing goes wrong. The container I am using is 'priviledged' and should actually do binfmting just fine (i use it also to build for arm)
2019-10-15 20:13:48 <maldridge> tmhoang: there doesn't seem to be interest from the project, so I'm unlikely to split out the vagrant components from their existing repo
2019-10-15 20:15:20 <maldridge> algitbot: that's a really weird error, since CGO shouldn't even be active on that build
2019-10-15 20:15:24 <maldridge> oliv3r[m]: ^
2019-10-15 20:39:19 <crosbymichael> has anyone tried things like zoom on an alpine desktop?
2019-10-15 20:39:39 <crosbymichael> thinking about installing alpine+xfce on my laptop but I need things like zoom for work
2019-10-15 20:42:27 <nepugia> Uhh, what is zoom?
2019-10-15 20:42:43 <crosbymichael> video chat
2019-10-15 20:42:45 <AinNero> like, zooming in on monitor contents
2019-10-15 20:42:48 <AinNero> uh
2019-10-15 20:42:57 <crosbymichael> https://zoom.us/
2019-10-15 20:42:59 <crosbymichael> sorry
2019-10-15 20:43:02 <_ikke_> Chat application with a free backdoor
2019-10-15 20:43:10 <crosbymichael> yes! thats the one
2019-10-15 20:43:37 <nepugia> well, you can install a glibc chroot of another distro if your nice app doens't run on alpine proper
2019-10-15 20:43:54 <AinNero> dunno.. running propietary software compiled against glibc on alpine is tricky
2019-10-15 20:43:57 <crosbymichael> ya, that is always an option
2019-10-15 20:43:59 <AinNero> sometimes works, sometimes doesnt
2019-10-15 20:44:28 <AinNero> 
ACTION has a dedicated ubuntu system just for the stuff that doesn't run on alpine

2019-10-15 20:44:40 <AinNero> like propietary software
2019-10-15 20:45:09 <crosbymichael> i'm debating on alpine in a VM on my laptop or as the root and run ubuntu in a vm for this type of stuff
2019-10-15 20:47:38 <AinNero> a chroot or some sort of container usually suffices, that makes data sharing easier
2019-10-15 20:47:47 <AinNero> since you can bind-mount directories
2019-10-15 20:50:33 <crosbymichael> sounds like a plan
2019-10-15 20:51:48 <PICCORO> hey guys how it's the sudo behavour in alpine?
2019-10-15 20:52:39 <_ikke_> that's a very broad question
2019-10-15 20:54:01 <maldridge> crosbymichael: you could also just use zoom in a browser
2019-10-15 20:55:04 <PICCORO> that' why i asked _ikke_ ! due sudo works so like windoshit "runs as" mechanish
2019-10-15 20:55:20 <PICCORO> also due only windo-like only distributions promote that tool
2019-10-15 20:56:05 <_ikke_> PICCORO: sudo is a generic linux tool, not specific to Alpine Linux at all
2019-10-15 20:58:19 <PICCORO> of course.. but well.. as i said works like "runs as"... also only user-friendly linux distros promote it! so what about alpine? we sould promote that in official alpine docs  and wiki tutorials?
2019-10-15 20:58:55 <PICCORO> and soo yo knowed _ikke_ .. systemd seems will be integrated into/with sudo
2019-10-15 20:59:45 <mps> PICCORO: we have 'doas' also
2019-10-15 21:01:40 <PICCORO> interesting.. documentation ?
2019-10-15 21:02:16 <PICCORO> and that "doas" alternative are due i mentioned "systemd" ?
2019-10-15 21:02:58 <mps> no, hope it will replace sudo in alpine (especially abuild)
2019-10-15 21:09:46 <PICCORO> mps that's its a official position or just that you wants to happened?
2019-10-15 21:10:44 <mps> not only I but also some other people
2019-10-15 21:12:22 <mps> and I think author is here or some other alpine channel
2019-10-16 00:22:24 <electrite> Hello! does anyone know how to install tor on alpine via $ apk add tor? it is saying "ERROR: unsatisfiable constraints: tor (missing): required by: world[tor]"
2019-10-16 00:40:46 <iggy> electrite: have you enabled the community repository?
2019-10-16 00:40:59 <electrite> How do I check?
2019-10-16 00:41:24 <iggy> check in /etc/apk/repositories
2019-10-16 00:41:35 <electrite> ok please give me a second.
2019-10-16 00:43:04 <electrite> I have this and this: http://dl-cdn.alpinelinux.org/alpine/v3.10/community & http://dl-cdn.alpinelinux.org/alpine/edge/community
2019-10-16 00:43:32 <iggy> not a great idea to mix 3.10 and edge
2019-10-16 00:43:53 <electrite> How would I come about fixing that?
2019-10-16 00:44:35 <iggy> it depends how you got to this situation in the first place
2019-10-16 00:44:46 <electrite> What do you mean?
2019-10-16 00:45:00 <electrite> I downloaded Alpine the other day and installed it in virtualbox today.
2019-10-16 00:45:06 <iggy> are you expecting to be running on edge or 3.10?
2019-10-16 00:45:41 <electrite> please do forgive me as I do not know which. Only the one of which will help me acquire tor.
2019-10-16 00:46:06 <iggy> it's in either really, that's why I said it depends
2019-10-16 00:46:31 <electrite> Very well, how would I come about appending edge?
2019-10-16 00:46:46 <iggy> if this is a system that you can easily recreate, I'd comment out the 3.10 lines and just stick with edge
2019-10-16 00:47:59 <iggy> do an `apk update ; apk upgrade`
2019-10-16 00:48:20 <electrite> Oh nice! it works now. I didn't have to do apk update.
2019-10-16 00:48:44 <electrite> Silly mistake, I didn't notice the comments. However thanks for leading me to that realization.
2019-10-16 00:49:16 <iggy> glad you got it working
2019-10-16 03:16:27 <kaey> oliv3r[m]: thats a spurious error, you can safely ignore it
2019-10-16 05:15:31 <oliv3r[m]> kaey you mean the cgo thing? hmm, so your saying it was compiled despire the error?
2019-10-16 05:36:46 <kaey> oliv3r[m]: yes and yes
2019-10-16 05:42:16 <oliv3r[m]> stupid error; why don't people on the internet say this :)
2019-10-16 05:43:32 <kaey> oliv3r[m]: there https://github.com/golang/go/issues/31544
2019-10-16 05:44:09 <kaey> alpine is affected because our go is patched to use external linking by default
2019-10-16 05:45:21 <oliv3r[m]> ah, ok that was something I didn't know; but then, I know nothing of go ;)
2019-10-16 06:33:51 <oliv3r[m]> Ok, so then !482 is ready :)
2019-10-16 07:27:47 <tmhoang> mps: was this sudo thing you talked about yesterday/few days ago ?
2019-10-16 07:27:52 <tmhoang> https://thehackernews.com/2019/10/linux-sudo-run-as-root-flaw.html
2019-10-16 07:28:06 <tmhoang> CVE-2019-14287
2019-10-16 07:30:01 <mps> tmhoang: yes
2019-10-16 08:27:23 <ncopa> rsmarples: hi! nice to see you around here!
2019-10-16 08:31:49 <rsmarples> ncopa: hi! just the man :D
2019-10-16 08:34:28 <rsmarples> was hunting for a gcc-9.2 based distro and ofc alpine was the answer .... but no libubsan seems to be around?
2019-10-16 08:35:18 <rsmarples> sadly i can't repro my issue on alpine but it would still be handy to have them libs
2019-10-16 08:41:29 <ncopa>  # musl does not support mudflap, or libsanitizer
2019-10-16 08:42:41 <ncopa> https://git.alpinelinux.org/aports/tree/main/gcc/APKBUILD#n239
2019-10-16 08:42:45 <ncopa> thats the reason why
2019-10-16 08:42:58 <ncopa> what is your issue?
2019-10-16 08:43:46 <ncopa> also note that alpine's gcc enables -fstack-protector-strong by default
2019-10-16 08:43:56 <ncopa> and build with PIE by default
2019-10-16 08:52:03 <kaey> rsmarples: archlinux ships gcc92
2019-10-16 08:59:27 <ncopa> or you could use docker gcc:9.2, which imho is the most convenient way to get an environment
2019-10-16 10:01:45 <rsmarples> kaey, ncopa: sadly neither alpine nor arch gcc 9.2 can replicate the issue - only gentoo and devuan miscompile dhcpcd with gcc-9.2
2019-10-16 10:01:59 <rsmarples> so it's a little wierd
2019-10-16 10:07:34 <ncopa> how do you know its miscompiled? did you compare the asm?
2019-10-16 10:41:53 <rsmarples> ncopa: i didn't comapre the asm
2019-10-16 10:42:09 <rsmarples> it fails by not mis computing ipv4 checksums
2019-10-16 10:42:17 <rsmarples> s/not//
2019-10-16 10:42:47 <rsmarples> i can build the same code on the same box with gcc-8.3 or another other compile with the same options and it works fine
2019-10-16 12:34:18 <xeems> Hi folks, I'm new to Alpine and am fiddling around a bit with it to see if it could fit my needs. As a part of this, I've tried to install ZFS, by installing the packages zfs and zfs-vanilla. When I try to modprobe zfs it finds no kernel module. Any ideas of what might be missing?
2019-10-16 12:35:01 <qa3Txu0iak0F> probably you run a different kernel than the modules you installed, you might want to reboot
2019-10-16 12:35:20 <qa3Txu0iak0F> but you can check first with uname -a and compare it to /lib/modules if the versions match...
2019-10-16 12:35:31 <crosbymichael> Have you seen the zfs docs? https://wiki.alpinelinux.org/wiki/Setting_up_ZFS_on_LUKS
2019-10-16 12:35:54 <crosbymichael> shows you atleast how to get the module
2019-10-16 12:36:26 <xeems> qa3Txu0iak0F the versions do match
2019-10-16 12:37:42 <xeems> crosbymichael I've seen that it exists, but was kind of put off by the mention of LUKS, which I'm not presently interested in, but I can take a look anyway
2019-10-16 12:38:31 <xeems> crosbymichael actually I had read it, and the packages I do not have seem to be LUKS-related to me
2019-10-16 12:38:41 <crosbymichael> ok
2019-10-16 12:39:10 <ncopa> hey crosbymichael! welcome!
2019-10-16 12:39:13 <xeems> Is there something like "yum whatprovides" in Alpine?
2019-10-16 12:39:13 <crosbymichael> i run zfs on alpine, and havn't had any issues.  zfs + zfs-"your kernel"
2019-10-16 12:39:19 <KatolaZ> xeems: have you `depmod -a` ?
2019-10-16 12:39:36 <xeems> KatolaZ no I haven't
2019-10-16 12:40:48 <KatolaZ> try it
2019-10-16 12:41:23 <xeems> `depmod -a` generates a bunch of errors complaining about read-only file system. Could this be related to the fact that I'm running it disklessly?
2019-10-16 12:41:33 <KatolaZ> yeah, it could
2019-10-16 12:44:42 <ncopa> xeems: i think the alpine-extended iso ships with zfs
2019-10-16 12:46:02 <crosbymichael> ncopa, it does
2019-10-16 12:53:32 <xeems> I think for my purposes right now I can as well install it to disk so I'll do that and hope for the best
2019-10-16 14:37:28 <mrgreen> How do I add apks to world? when iso building?
2019-10-16 14:56:14 <oliv3r[m]> _ikke_: kun je look at !482 it's failing as it's not able to access the runner; I see the x86_64 is offline; I think your personal runner is hooked up to the CI, but we (I) can't use it
2019-10-16 14:57:41 <_ikke_> lol, dutch/english :D
2019-10-16 14:58:17 <PureTryOut[m]> "kun je look", nice
2019-10-16 14:59:33 <_ikke_> oliv3r[m]: I think the runner was not started again after an unexpected server reboot this morning
2019-10-16 14:59:45 <mps> 'kun yu luk'
2019-10-16 14:59:59 <oliv3r[m]> ah, fair enough
2019-10-16 15:00:14 <_ikke_> oliv3r[m]: Now it's running again
2019-10-16 15:00:35 <oliv3r[m]> i'll restart the job then
2019-10-16 15:00:40 <_ikke_> not necessary
2019-10-16 15:00:45 <_ikke_> it's already continueing
2019-10-16 15:01:07 <_ikke_> oh, it's canceled now
2019-10-16 15:01:24 <oliv3r[m]> i was to fast! or slow
2019-10-16 15:01:34 <_ikke_> oliv3r[m]: the runner I host doesn't do CI build jobs, only simple things like linting
2019-10-16 15:01:44 <oliv3r[m]> it's trying to start now
2019-10-16 15:01:59 <oliv3r[m]> there we go!
2019-10-16 15:02:49 <oliv3r[m]> success! !482 is ready for review and merge
2019-10-16 15:34:54 <futro> Hey folks, i wonder if there is any disadvantage in creating a partition after diskless-install and mounting it as /var via fstab?
2019-10-16 15:35:14 <AinNero> you mean data mode?
2019-10-16 15:35:20 <_ikke_> manual data mode
2019-10-16 15:35:33 <futro> yes _ikke_
2019-10-16 15:35:40 <AinNero> i... i have such a setup on my Siemens Futro, futro
2019-10-16 15:35:59 <AinNero> is that were your nick cames from? :P
2019-10-16 15:36:07 <futro> data mode takes a whole disk to mount as /var.
2019-10-16 15:36:32 <AinNero> it didn't for me,
2019-10-16 15:37:01 <futro> have you been able to supply a parition like /dev/sda2 to setup-disk?
2019-10-16 15:37:30 <AinNero> ahh i see, yes, it takes a whole disk, but only a partition gets used as /var
2019-10-16 15:37:36 <futro> yes
2019-10-16 15:37:40 <AinNero> because it also allocated a swap partition
2019-10-16 15:37:47 <futro> yes
2019-10-16 15:38:23 <futro> i took the first gb to create a bootable partition containing the iso
2019-10-16 15:39:06 <futro> so i got rid of using an usb-stick and still have the boot-to-ram setup
2019-10-16 15:39:31 <AinNero> that sounds unproblematic to me
2019-10-16 15:39:51 <AinNero> i wonder, how did you put the ISO on the partition? dd of=/dev/sda and then edit the partition table?
2019-10-16 15:41:40 <futro> https://wiki.alpinelinux.org/wiki/Create_a_Bootable_USB#Add_Alpine_Linux_to_the_USB_stick
2019-10-16 15:42:42 <AinNero> ah okay, no iso involved there, only bootloader and boot files
2019-10-16 15:42:47 <futro> I booted from usb and ran setup-alpine once, created the partitions in my favour, added syslinux and finally ran setup-bootable
2019-10-16 15:44:06 <futro> yes. it's quite replacing the usb-stick by one partitions of the cf
2019-10-16 15:44:48 <AinNero> sounds okay to me
2019-10-16 15:44:48 <futro> this way i could format the /var partition to f2fs
2019-10-16 15:45:06 <AinNero> i have no experience with f2fs
2019-10-16 15:45:18 <futro> me2...i give it a try.
2019-10-16 15:45:34 <AinNero> futro: if you have a Siemens Futro, which version of it?
2019-10-16 15:46:33 <AinNero> im also interested in the /proc/cpuinfo for the transmeta CPU and the pci id's
2019-10-16 15:47:02 <futro> in fact i have just finished testing this scenario in vbox.
2019-10-16 15:47:19 <AinNero> oh okay
2019-10-16 15:47:25 <AinNero> keep me updated
2019-10-16 15:47:33 <mps> most of my SSD and SD card have f2fs as FS
2019-10-16 15:47:42 <futro> but there is a s520 and as900 (i think) waiting to get the same treatment
2019-10-16 15:48:59 <AinNero> ah, okay
2019-10-16 15:49:08 <futro> yesterday i installed debian on the s520. took me 3 attempts, because the ascii-installer crashed to times without a warning
2019-10-16 15:49:22 <AinNero> a different generation... i have a S300, im sorta into retrocomputing
2019-10-16 15:49:44 <futro> and it was ***** slow - i think thats the poor write speed of the cf
2019-10-16 15:49:51 <AinNero> S520 has AMD cpu + EFI, while the S300 has i585-compatible by Transmeta, so neither Intel nor AMD
2019-10-16 15:50:29 <AinNero> i run Alpine on the S300 because there aren't much linuxes left that support it
2019-10-16 15:52:16 <futro> s520 is quite modern and doesn't cost more than a Raspberry Pi 4 (that has USB 3)
2019-10-16 15:55:01 <AinNero> uhm, used s520 from ebay vs new RPi 4?
2019-10-16 16:01:14 <futro> I stumbled upon the (possible?) lack of hardware crypto support with RPi's arm cpu
2019-10-16 16:02:29 <futro> So that became an occasion to have a look at the s520 ^^
2019-10-16 16:04:58 <futro> I will report back, when the futros are taken care of.
2019-10-16 16:14:54 <crosbymichael> i'm having issues with alpine on an NVME drive.  bios does not see it at all
2019-10-16 16:15:03 <crosbymichael> any tips?
2019-10-16 16:23:54 <mps> crosbymichael: does motherboard have uefi enabled
2019-10-16 16:25:22 <mps> about two months ago I had similar issue with NVME drive, enabling UEFI boot in bios solved it
2019-10-16 16:27:50 <crosbymichael> it does not have UEFI enabled
2019-10-16 16:28:17 <crosbymichael> how did you do your install?
2019-10-16 16:28:44 <mps> enabled UEFI and boot Alpine from usb
2019-10-16 16:28:59 <crosbymichael> ok
2019-10-16 16:29:04 <mps> rest you probably know
2019-10-16 16:29:25 <mps> it installed grub with uefi support
2019-10-16 16:29:28 <crosbymichael> GPT partions?
2019-10-16 16:29:35 <mps> yes, GPT
2019-10-16 16:29:42 <crosbymichael> i'm was trying to use syslinux
2019-10-16 16:30:21 <crosbymichael> well, setup with cryptsetup+zfs
2019-10-16 16:30:41 <crosbymichael> but the real problem was bios was not seeing any boot drive
2019-10-16 16:31:08 <mps> I had same till enabled uefi
2019-10-16 16:37:58 <qa3Txu0iak0F> i have no uefi enabled, but my nvme works well.
2019-10-16 16:39:15 <mps> maybe that depends on bios on motherboard
2019-10-16 16:39:48 <mps> I mean, quality and options in  bios
2019-10-16 16:58:27 <mps> restarting job doesn't help
2019-10-17 01:29:56 <cim209> where is the ssh log file?
2019-10-17 01:31:04 <cim209> nvm it's in a shared log file
2019-10-17 02:15:01 <qman__> anyone using Xorg on a radeon HD 7xxx on alpine 3.10? I'm trying to install it, but as soon as I run setup-xorg-base it breaks my display on all subsequent reboots - I get the prompt to enter my luks password and see a few openrc start messages, then it goes to no signal
2019-10-17 02:17:43 <qman__> if I don't run setup-xorg-base, the framebuffer works fine, at the full resolution of the display (which happens to be 4k30)
2019-10-17 03:31:12 <TBB> do you by any chance have both integrated graphics and a discrete gfx card in that machine?
2019-10-17 03:32:12 <TBB> I don't remember the exact kernel boot parameter magic, but I've run into that problem and I remember the solution being disabling one of them
2019-10-17 03:33:05 <TBB> 
ACTION needs to go back to sleep but hopefully that was helpful.

2019-10-17 05:38:57 <User__> Hey. I have a problem. After installing alpine linux, the computer does not shut down by the commands poweroff, reboot, etc. the Problem is observed on a new installation. The screen turns off, the computer continues to work, but does not turn off or restart.
2019-10-17 08:51:03 <wowaname> uh poweroff should work
2019-10-17 08:51:18 <wowaname> vanilla kernel?
2019-10-17 08:52:36 <AinNero> uh, that guy has already left
2019-10-17 08:53:29 <KH405> Dumb question, how does alpine puts new version in his repo 6
2019-10-17 08:54:00 <KH405> I feel like some packages are hold back
2019-10-17 08:54:37 <wowaname> AinNero: sorry i cant see parts here
2019-10-17 08:54:44 <wowaname> well i turned them off anyway
2019-10-17 08:56:14 <AinNero> i dont have that disabled, but i didn't read it
2019-10-17 08:56:23 <AinNero> i learned because auto-complete for hilighting failed
2019-10-17 09:05:31 <KH405> Like why did python just go updated to 3.7.5 when 3.8 was released yesterday ?
2019-10-17 09:05:48 <KH405> I'm on edge branch btw
2019-10-17 09:07:01 <_ikke_> KH405: Because a minor update is easier than a major ugprade
2019-10-17 09:07:26 <KH405> "easier" ?
2019-10-17 09:07:26 <_ikke_> a major upgrade requires all python modules to and other dependencies to be rebuilt as well
2019-10-17 09:07:28 <_ikke_> yes
2019-10-17 09:07:45 <_ikke_> ^^
2019-10-17 09:08:14 <KH405> I see, so when will 3.8 be available ?
2019-10-17 09:19:49 <Cogitri> KH405: Once someone does the work required
2019-10-17 09:20:03 <Cogitri> But IMHO we shouldn't upgrade just yet, lots of stuff still doesn't work on Py3.8
2019-10-17 09:40:59 <KH405> Gotcha, Cogitri
2019-10-17 11:30:44 <crockwork> Does anyone here commonly use fzf?
2019-10-17 11:30:50 <crockwork> I'm not sure what its main purpose is.
2019-10-17 11:31:49 <AinNero> crockwork: i3 is using dmenu
2019-10-17 11:32:00 <AinNero> i used to use fzf in a terminal inplace of dmenu
2019-10-17 11:32:06 <AinNero> with the intent to launch programs
2019-10-17 11:32:18 <_ikke_> crockwork: I use it
2019-10-17 11:32:39 <crockwork> AinNero: How the hell does that work?
2019-10-17 11:32:43 <crockwork> _ikke_: What do you use it for?
2019-10-17 11:33:10 <_ikke_> Several usecases: with tmux to quickly switch sessions, as a replacement for ctrl+r
2019-10-17 11:33:21 <_ikke_> to switch wireless networks via CLI
2019-10-17 11:33:36 <crockwork> Oh, big question: Today, a friend of mine deleted some files he didn't want to delete. We failed to do what I wanted to do beforehand, which is to basically move into a ramdisk or something and then remount the main drive (only drive) as read only. Do all linux distros have this functionality (to basically move the working system into memory)?
2019-10-17 11:33:41 <AinNero> crockwork: like the regular i3+dmenu workflow?
2019-10-17 11:33:45 <_ikke_> Any usecase where you need to select items from a list and want fuzy logic to select it
2019-10-17 11:34:09 <crockwork> _ikke_: So can you basically search for something that might be in that tmux buffer to switch windows?
2019-10-17 11:34:21 <crockwork> How the HELL do you do wireless network switching with fzf? that sounds sick
2019-10-17 11:34:40 <_ikke_> crockwork: The idea is that you can feed any list to fzf, and it just returns what you selected
2019-10-17 11:34:55 <_ikke_> so fzf does not need to know the details, you just create a shell alias for that
2019-10-17 11:35:13 <crockwork> That's so weirdly interesting
2019-10-17 11:35:21 <crockwork> Can you select multiples in fzf?
2019-10-17 11:35:32 <_ikke_> Yes
2019-10-17 11:35:37 <_ikke_> fzf -m / --multi
2019-10-17 11:36:14 <crockwork> what the literal hell
2019-10-17 11:36:16 <crockwork> how interesting
2019-10-17 11:36:27 <_ikke_> crockwork: http://tpaste.us/RgZb
2019-10-17 11:36:38 <crockwork> it seems so powerful--was there an alternative to this before it came around?
2019-10-17 11:36:54 <_ikke_> Maybe, I don't know
2019-10-17 11:37:09 <crockwork> seems like it makes interactive terminal stuff really easy
2019-10-17 11:38:42 <_ikke_> crockwork: this is what I use to switch sessions in tmux: http://tpaste.us/9gmV
2019-10-17 11:39:33 <crockwork> what's the || true for?
2019-10-17 11:39:46 <_ikke_> To prevent it from erroring out when you select nothing
2019-10-17 11:39:57 <crockwork> Ah, interesting
2019-10-17 11:40:29 <qman__> TBB: no, that's the only video adapter in the machine, it's an FX-8350 - thanks for the suggestion though
2019-10-17 11:40:45 <qman__> maybe I'll test the HDMI port and see if it's trying to use that, even though my displays are on the displayports
2019-10-17 11:47:18 <qman__> I broke it worse trying to fix it so I have to redo the install to test
2019-10-17 12:20:23 <midasi> is somebody using cryptodev with alpine? I don't find any package providing that
2019-10-17 12:47:25 <steven> whats the difference between docker-php-ext-{configure|install} and using apk (running alpine in docker) ?
2019-10-17 13:32:41 <midasi> how can I get rid of this error?
2019-10-17 13:32:41 <midasi> ERROR: unsatisfiable constraints:
2019-10-17 13:32:41 <midasi>   linux-vanilla-dev-4.19.78-r0:
2019-10-17 13:32:41 <midasi>     breaks: .makedepends-mypkg-linux-vanilla-20191017.152937[linux-vanilla-dev=4.19.76-r0]
2019-10-17 13:33:25 <midasi> I tried apk fix and apk del .makedepends.... both didn't work
2019-10-17 15:54:54 <misterunknown> Hi guys. I have a question, which is not alpine related. Does anyone has a rule-of-thumb-like value how much slower a syscall is, if I trace it with strace?
2019-10-17 17:18:49 <dalias> misterunknown, well you need a couple round-trip context switches to strace and back, plus all the syscalls strace uses in performing the trace
2019-10-17 17:19:10 <dalias> (for each syscall)
2019-10-17 17:19:42 <dalias> do i'd guess it takes well over 5x as long, probably 10x-20x as long
2019-10-17 17:20:28 <dalias> that's assuming the syscall overhead dominates. if you have a syscall that's genuinely slow because of what it's doing in kernelspace, difference could be much less
2019-10-17 18:53:04 <bbuccianti> why alpine-sdk needs sudo?
2019-10-17 18:54:08 <_ikke_> bbuccianti: installing dependencies requires root access, but building itself should be done as normal user
2019-10-17 18:54:17 <_ikke_> (and other things require root as well)
2019-10-17 18:54:24 <Cogitri> Because abuild by default uses sudo to install stuff
2019-10-17 19:15:32 <bbuccianti> would be too difficult to change that to doas?
2019-10-17 19:18:03 <brass> Doas is excellent
2019-10-17 19:19:40 <mps> there is plan to use doas instead of sudo, iirc
2019-10-17 19:23:25 <_ikke_> THe question is when
2019-10-17 19:24:28 <mps> when it is implemented in abuild, tested and work well, as usual ;)
2019-10-17 19:28:54 <joneskoo> Looks like on the rpi alpine you can't apk add wireguard-rpi for the kernel modules because apk can't write lib/modules/...: ERROR: Failed to create lib/modules/4.19.76-0-rpi/extra/wireguard.ko: No such file or directory
2019-10-17 19:29:14 <joneskoo> because of modloop basically
2019-10-17 19:33:29 <joneskoo> https://pkgs.alpinelinux.org/contents?file=&path=&name=wireguard-rpi&branch=v3.10&repo=community&arch=armhf vs. uname -a shows 4.19.58-0-rpi - perhaps that's the problem
2019-10-17 19:37:31 <joneskoo> hm, found https://wiki.alpinelinux.org/wiki/Upgrading_Alpine ; I think this macro is broken
2019-10-17 19:37:34 <joneskoo> setup-bootable -u {{#latestalp:alpine|url}} /media/$LBU_MEDIA
2019-10-17 20:37:38 <jailbox> hmm, did a apk upgrade, it installed linux-vanilla-4.19.79-r0  after reboot uname -r is still ..-4.19.78-r0 so it cant find modules, how to fix ?
2019-10-17 20:38:30 <jailbox> it seems, from file-date, that kernel et.al. in /boot where not updated/replaced with new version
2019-10-17 20:40:38 <jailbox> (did a temp-fix  ln-s 4.19.79-0-vanilla 4.19.78-0-vanilla in /lib/modules to get on)
2019-10-17 20:47:12 <crockwork> Is there a way to boot a 'live' version of my disk (where the disk is then read-only) from a shut-down state?
2019-10-17 20:48:24 <crockwork> Essentially what I want to do is learn if there is a way to 1) remount the system r/o in case of a big messup 2) shut down 3) boot up and change something in grub, maybe, to be able to then install packages in-memory and not do anything to the system while I work on it without booting from external media.
2019-10-17 21:16:49 <mps> not alpine issue, but anyone have solution to disable Ctrl-Q in firefox
2019-10-17 21:17:45 <mps> it is quite annoying because I press it sometimes instead of Ctrl-W
2019-10-17 21:22:06 <jailbox> oops sorry, not a bug, files in /boot where updated, but due to my multiboot-setup it keept booting with the old version
2019-10-17 21:22:24 <Cogitri> Pretty sure key shortcuts are hardcoded in GTK applications
2019-10-17 21:33:05 <mps> Cogitri: hardcoded, so no solution except change source and build locally
2019-10-17 21:34:41 <Cogitri> Either that or use workarounds like https://superuser.com/questions/1318336/how-to-disable-ctrlq-shortcut-in-firefox-on-linux
2019-10-17 21:36:06 <mps> read it earlier, and didn't managed to set it to work :(
2019-10-17 21:36:23 <mps> thanks for trying to help me
2019-10-17 23:00:50 <shodan45> hmm, anyone know if RH's cockpit works in alpine?
2019-10-17 23:01:22 <shodan45> I fear it may require systemd, but haven't seen that mentioned yet
2019-10-17 23:01:37 <bougyman> I thought that as well, shodan45.
2019-10-17 23:03:29 <shodan45> I'm just trying to find a nice front end for managing kvm/qemu
2019-10-17 23:03:33 <shodan45> hopefully web based
2019-10-17 23:03:52 <bougyman> libvirt doesn't do it for ya?
2019-10-17 23:27:18 <shodan45> I'll try virtual machine manager again
2019-10-17 23:28:10 <shodan45> I have a headless home server that I'd like to run VMs on... I just want a nice, simple frontend
2019-10-17 23:28:28 <shodan45> and that home server runs alpine :)
2019-10-18 01:46:18 <ndowens> Anyone have a recommendation of a SoB kit, fairly cheap, just intend for it to be a wireless router
2019-10-18 01:46:31 <ndowens> with alpine as the os
2019-10-18 01:55:09 <ndowens> kinda seems rpi zero w model would be fine
2019-10-18 02:36:29 <dalias> does pi zero w's wifi work well in ap mode?
2019-10-18 02:42:38 <ndowens> i found an article on how to do it
2019-10-18 02:42:59 <ndowens> step said to enable something before something lol, i forget, have it bookmarked, one sec
2019-10-18 02:43:41 <ndowens> https://wiki.alpinelinux.org/wiki/Raspberry_Pi_Zero_W_-_Installation
2019-10-18 05:19:23 <grayhatter> anyone have resume from hibernation working?
2019-10-18 05:20:26 <grayhatter> do you have resume in you boot config -> yes... do you have a kernel with hibernation support -> yes
2019-10-18 05:33:31 <mainiak_cz> Hello. Can please anybody point me to documentation how to create correctly rc scripts if I want to create my own service?
2019-10-18 05:33:45 <mainiak_cz> I found https://wiki.alpinelinux.org/wiki/Writing_Init_Scripts but some more examples would be great
2019-10-18 05:35:09 <grayhatter> mainiak_cz: what are you trying to create one for
2019-10-18 05:35:11 <grayhatter> ?
2019-10-18 05:40:10 <mainiak_cz> Hi grayhatter! I would like to do control my docker conatiners. With systemd it was pretty easy. I don't need anything like pid control (now), but I would like dependency management if possible - ie. start after docker daemon and network ...
2019-10-18 05:40:46 <mainiak_cz> (I assume systemd is not and maybe won't be ever supported ... right?)
2019-10-18 05:45:47 <mainiak_cz> Also if I am not mistaken - using s6 (as recommended on documentation page above) - is only recommendation, but not really necessary, is it?
2019-10-18 05:47:30 <Cogitri> Just use OpenRC, that's definitely the easiest way to do this for as long as Alpine uses it
2019-10-18 05:47:48 <Cogitri> As for how to make a service - please refer to OpenRC's docs for that
2019-10-18 05:51:20 <mainiak_cz> Hi Cogitri. Aren't there any discrepancies between OpenRC - which was mainly designed and used for Gentoo - and Alpine?
2019-10-18 05:51:41 <Cogitri> Sorry?
2019-10-18 05:53:34 <mainiak_cz> Is there anything I have to be careful about? Something that is different on Alpine then what is written on project OpenRC (which equals to Gentoo in my mind) official documentation?
2019-10-18 05:54:07 <mainiak_cz> Or Alpine uses OpenRC fully without and modifications and caveats?
2019-10-18 05:57:30 <Cogitri> We just use OpenRC
2019-10-18 05:57:59 <Cogitri> There aren't any "Alpine modifications"
2019-10-18 06:10:10 <mainiak_cz> thanks cogitri!
2019-10-18 06:11:50 <Cogitri> 👍
2019-10-18 07:15:49 <misterunknown> dalias: Thanks for the explanation :) You're obviously right: As syscalls need different times, the overhead with strace could not simply be calculated by a factor. Would have been too easy...
2019-10-18 08:14:24 <CvH> ncopa: hi, should the RPi4 work with Alpine 3.10.2 ? if not is there a testimage anywhere ?
2019-10-18 08:14:47 <dngray[m]> hmm i was using https://pkgs.alpinelinux.org/package/edge/main/armv7/ppp-pppoe looks like i will have to change if i start using https://en.wikipedia.org/wiki/IPoE
2019-10-18 08:15:01 <_ikke_> CvH: no, it's not supported, the rpi4 was released after 3.10
2019-10-18 08:15:43 <dngray[m]> is it recommended we use https://pkgs.alpinelinux.org/package/edge/main/ppc64le/rp-pppoe instead?
2019-10-18 08:15:47 <CvH> just because I saw the rpi4 dtb files at 3.10.2 :)   - is there anywhere an edge image or something ?
2019-10-18 08:36:02 <mps> dn ime, rp-pppoe is better than ppp-pppoe
2019-10-18 08:36:12 <mps> dngray[m]: ^
2019-10-18 08:36:45 <mps> (didn't press TAB hard enough :) )
2019-10-18 09:21:16 <ncopa> CvH: i dont think alpine 3.10 works with rpi
2019-10-18 09:21:24 <ncopa> i saw some PR for rpi4 support
2019-10-18 09:21:33 <ncopa> needs another kernel config for that
2019-10-18 09:21:41 <CvH> so there is no working edge/somewhat for alpine yet ?
2019-10-18 09:21:56 <ncopa> i havent had time to look at the rpi4 work yet
2019-10-18 09:22:03 <ncopa> and apparently noone else has either
2019-10-18 09:22:22 <ncopa> but we'll try get it in soon
2019-10-18 09:22:31 <ncopa> a couple of weeks or so
2019-10-18 09:22:46 <CvH> how do you build the "initramfs-rpi2" etc ?
2019-10-18 09:22:53 <_ikke_> There have been people working on RPI4 support
2019-10-18 09:24:17 <ncopa> CvH: mkinitfs builds the initramfs's
2019-10-18 09:24:44 <ncopa> i think they are generated with update-kernel script which is called from scripts/mk.img*
2019-10-18 09:24:49 <ncopa> or something like that
2019-10-18 09:26:35 <CvH> hmm as I tried it for allwinner (generic arm image) it failed because it used the kernel that is at the alpine repo and it was impossible to choose an self build one
2019-10-18 09:27:06 <CvH> likewise the apline build system is still broken for it so rpi is likely not building there too ?
2019-10-18 09:28:40 <ncopa> i need to: update kernel to 4.19.80 (include review and test the kernel patches in queue), test that it works, update kernel in 3.10-stable, fix 3.10 builders (go fails), get 3.10.3 release out, tag new edge snapshot release, revert wxgtk (merged a patch that i shouldnt), tag new abuild release (include review the patch queue for abuild), update
2019-10-18 09:28:41 <ncopa> python to 3.8 and rebuild every single py-* package, bootstrap all the 3.11 builders
2019-10-18 09:28:48 <ncopa> and then i may be able to have a look at rpi4 kernel
2019-10-18 09:29:02 <ncopa> i doubt that i will be able to reach everything today
2019-10-18 09:29:35 <ncopa> oh, i may also need to merge some PRs in between the above so the PR queue dont grow too big
2019-10-18 09:30:03 <ncopa> i also need to follow up the discussion about the alpine governance
2019-10-18 09:30:52 <CvH> ah okay so you workaround the buildsystem problem with pushing new files to the repo ;)  so i likely can't build it my self in the meantime
2019-10-18 09:54:38 <ncopa> CvH: https://github.com/alpinelinux/aports/pull/10260
2019-10-18 09:57:19 <mps> oh, someone want to remove armv7 arch from alpine, iiuc what is on above link
2019-10-18 10:07:21 <CvH> yea only aarch64 is not really something you wanne have for RPi at the moment, ofc you also don't want armhf for > RPi1
2019-10-18 10:08:13 <CvH> but aarch64 is going to work "best" at RPi, kernel devs are just not there yet
2019-10-18 10:09:43 <mps> ime, on arm64 most armv7 armhf userspace works fine
2019-10-18 10:11:18 <CvH> yea, but there isn't a full functional aarch64 for rpi yet, armv7 is the standard atm - armhf is just slower
2019-10-18 10:11:57 <mps> if I do not develop some things on arm64 I would switch my arm64 boxes to armv7 userspace
2019-10-18 10:13:15 <CvH> !if! aarch64 drivers etc are in good shape you are likely won't go back to armv7, you are basically loose performance
2019-10-18 10:13:17 <mps> CvH: looks like there is full functional aarch64 but missing part are boot loader and kernel, ime
2019-10-18 10:14:09 <mps> i'm talking about alpine, not other distros
2019-10-18 10:14:52 <mps> and performance could be even better on arm32 than arm64
2019-10-18 10:16:14 <mps> and compiler/assembler (and rest build tools) are better tested and fixed for arm32 than arm64
2019-10-18 10:18:48 <CvH> yes from that perspective you want arm32, overall arm32 is basically "dead" and it gets a lot shifted to aarch64 - ofc work in progress
2019-10-18 10:20:15 <CvH> but aarch64 has too much benefits to stay at armv7 for longer then needed - I am not sure that the typical user would care
2019-10-18 10:21:07 <mps> tipical arm users have a lot arm32 devices 'at work' and sure, they care
2019-10-18 10:22:14 <mps> CvH: ask pmOS people
2019-10-18 10:29:29 <CvH> yea you can't change legacy devices, but basically every modern arm is 64 only btw there are just blobs for 64
2019-10-18 10:58:45 <Piraty> does alpine have crosscompilers in the package repo?
2019-10-18 10:58:57 <_ikke_> No
2019-10-18 10:59:10 <Cogitri> Nope
2019-10-18 10:59:34 <Cogitri> Crossocompilation generally isn't really supported by Alpine
2019-10-18 11:00:38 <Piraty> oh
2019-10-18 11:00:44 <Piraty> thanks
2019-10-18 11:00:49 <Piraty> hi Cogitri
2019-10-18 11:00:57 <Piraty> hi maxice8
2019-10-18 11:01:36 <TBB> there's something in testing
2019-10-18 11:02:14 <Piraty> so alpine is bootstrapped and built on the targets ?
2019-10-18 11:02:43 <Cogitri> Yes, we have native builders
2019-10-18 11:03:35 <Cogitri> And hi, Piraty :)
2019-10-18 11:04:19 <maxice8> Piraty: yes?
2019-10-18 11:04:54 <Piraty> just hi, don't bother me
2019-10-18 11:05:02 <CvH> beware of https://github.com/alpinelinux/alpine-chroot-install that is basically not useable slow - an rpi1 is an performance wonder in comparision
2019-10-18 11:08:04 <Piraty> ah yeah, i had the idea of abusing qemu for avoid crosscompile scenarios for stuff that has malconfigured buildsystems that make crossbuilding too much of a pain
2019-10-18 11:08:18 <Piraty> now i have a reference implementation to look at ;)
2019-10-18 11:08:54 <Piraty> thanks CvH
2019-10-18 11:11:38 <CvH> we are talking about 80-100 slower in comparison to corosscompile AND you won't get exact results
2019-10-18 11:11:47 <CvH> cross compile
2019-10-18 11:13:19 <ncopa> we do native compile only because we cannot run the tests when crosscompile
2019-10-18 11:14:57 <CvH> a big giveaway for so less gain
2019-10-18 11:15:00 <Cogitri> ncopa: And cross can be kind of a pain at times
2019-10-18 11:15:17 <Cogitri> 
ACTION makes intense stares at GIR

2019-10-18 11:19:47 <CvH> try to crosscompile samba with python3 :D  good luck
2019-10-18 11:48:53 <TBB> I guess you could run tests in qemu, but would it really be reliable enough...
2019-10-18 11:53:04 <CvH> or run tests at arm hw and compile at x86_64 :)
2019-10-18 11:53:26 <CvH> not sure that this scales well ;)
2019-10-18 11:54:38 <TBB> even though I use cross compiling in embedded projects, somehow real hardware still sounds like a better idea
2019-10-18 11:58:08 <CvH> if you have some real hardware and this hardware is pretty fast etc ... so only really working if you have rented arm severs somewhere
2019-10-18 11:58:33 <CvH> or you have a lot $ or time :)
2019-10-18 12:12:12 <Cogitri> Lots ot upstreams just assume that their project is built on native systems, so lots of stuff doesn't work on cross sadly :/
2019-10-18 12:12:25 <Cogitri> So using native builds is the best option if you have the builders available
2019-10-18 12:15:43 <CvH> using native builds just workarounds broken build scripts, an half decent build script is usually also able to crosscompile - prime negative example samba :D  but not even the samba devs understand the samba buildscript
2019-10-18 15:04:41 <twobitsprite> Is there a way to pre-build an alpine OS image that can be installed to a system over a network (PXE, etc)?
2019-10-18 15:05:27 <_ikke_> alpine has netboot images: ilt on the targets ?                                                                                                                      │ bleb
2019-10-18 15:05:29 <_ikke_> 34.  danieli           │13:02:43         Cogitri │ Yes, we have native builders                                                                                                                                              │ blueness_
2019-10-18 15:05:32 <_ikke_> https://boot.alpinelinux.org
2019-10-18 15:05:34 <_ikke_> sorry
2019-10-18 15:06:10 <AinNero> twobitsprite: you can pre-configure an alpine system with giving an apkovl to the initramfs while booting
2019-10-18 15:07:16 <AinNero> its also possible to setup a system using a apkovl, so its not necessary to pre-generate a whole image
2019-10-18 15:07:50 <AinNero> definitely possible, but i dont think there is well-documented prior art
2019-10-18 15:08:25 <twobitsprite> AinNero: my use case requires pre-generating the whole image. this will be for an IoT device
2019-10-18 15:09:08 <twobitsprite> we're currently using CentOS Atomic as the base OS, but we've had a lot of problems with it and are investigating alternatives
2019-10-18 15:09:41 <AinNero> i'd hilight strfry since they presented an alpine-diskless managed system at toxcon last weekend
2019-10-18 15:09:49 <AinNero> *IoT system
2019-10-18 15:10:37 <AinNero> but it seems they are offline
2019-10-18 15:11:02 <twobitsprite> googling strfry just finds the glibc function
2019-10-18 15:11:12 <AinNero> he named himself after that
2019-10-18 15:11:47 <AinNero> i was supposed to present it with him, but i got sick unfortunately
2019-10-18 15:11:51 <AinNero> so you may ask me as well
2019-10-18 15:12:25 <twobitsprite> I am asking you :)
2019-10-18 15:13:17 <AinNero> Alpine diskless is putting a kernel, initramfs, kernel modules, repository and tarball with /etc contents onto one partition
2019-10-18 15:13:43 <AinNero> and it builds the system in-ram from these files, all of them could also be supplied via PXE+HTTP or other means
2019-10-18 15:15:29 <AinNero> whats you application running on the devices? docker?
2019-10-18 15:15:41 <twobitsprite> yes
2019-10-18 15:16:07 <twobitsprite> I'm reading the setup-bootable docs now, seems like that's what I'd use to pre-generate an image?
2019-10-18 15:16:39 <AinNero> uhm, take not its not the root file system that is generated, its /boot, and the rootfs will be in ram
2019-10-18 15:16:43 <AinNero> *take note
2019-10-18 15:17:04 <AinNero> will PXE be permanently available to the devices?
2019-10-18 15:17:09 <twobitsprite> but it has to be stored somewhere in order to be loaded in to ram, right?
2019-10-18 15:17:27 <twobitsprite> no, the PXE will be used to image the device, but then the device must be able to boot from local storage
2019-10-18 15:17:54 <twobitsprite> (probably flash storage, but hardware specs haven't been solidified yet)
2019-10-18 15:19:58 <twobitsprite> oh speaking of which, how well is arm64 supported in alpine? are all/most packages provided for arm?
2019-10-18 15:20:20 <AinNero> afaik most, but there aren't as many packages as debian to begin with
2019-10-18 15:20:57 <twobitsprite> that's fine, don't need full desktop OS level of packages, just making sure arm isn't the red-headed step child in the alpine family :P
2019-10-18 15:21:27 <AinNero> raspberri pi is officially supported
2019-10-18 15:22:10 <AinNero> ... do you want to run a docker cluster on raspberri pi's ?
2019-10-18 15:23:07 <twobitsprite> lol, no, this is an appliance we'll be selling to enterprise customers, just needs to run a few containers locally
2019-10-18 15:23:58 <twobitsprite> and it probably won't be a raspberry pi, it might even end up being an in-house developed board
2019-10-18 15:24:36 <twobitsprite> anyway, I have a meeting to go to, I'll probably be back later. thanks for taking the time to chat
2019-10-18 18:51:17 <mistnim> hi, how can I have wpa_supplicant in the installer?
2019-10-18 19:00:04 <grayhatter> mistnim: i's in the larger iso IIRC
2019-10-18 19:00:53 <grayhatter> it's*
2019-10-18 19:01:00 <mistnim> grayhatter: you mean the extended one?
2019-10-18 19:01:06 <grayhatter> yeah
2019-10-18 19:01:35 <mistnim> good, thanks
2019-10-18 22:05:38 <bbuccianti> do we have an alternative to fsck.vfat?
2019-10-18 22:07:10 <clandmeter> dosfstools?
2019-10-18 22:07:33 <bbuccianti> ouch I was searching for dofstools
2019-10-18 22:07:37 <bbuccianti> it's dosfs
2019-10-18 22:07:40 <bbuccianti> thanks clandmeter
2019-10-18 23:41:54 <Cadey> is it expected for qemu-user to not be present in 32 bit x86?
2019-10-18 23:57:05 <Cadey> ah it's not called qemu-user
2019-10-18 23:57:07 <Cadey> that makes sense
2019-10-19 05:54:32 <dngray[m]> hmm i noticed when i type `iptables --version` it says:
2019-10-19 05:54:32 <dngray[m]> `iptables v1.8.3 (legacy)`
2019-10-19 05:54:58 <dngray[m]> my debian machine says `iptables --version`
2019-10-19 05:54:59 <dngray[m]> `iptables v1.8.2 (nf_tables)`
2019-10-19 05:55:15 <dngray[m]> is there anything i have to do to get alpine-linux to use the nf_tables backend?
2019-10-19 05:55:35 <dngray[m]> or is this going to be a change in a future release of alpine linux
2019-10-19 05:57:17 <dngray[m]> seems both debian https://wiki.debian.org/nftables and rhel 8 https://www.redhat.com/en/blog/using-nftables-red-hat-enterprise-linux-8 now use nftables
2019-10-19 06:00:21 <dngray[m]> https://wiki.nftables.org/wiki-nftables/index.php/Legacy_xtables_tools
2019-10-19 06:01:24 <dngray[m]> does anyone know if there's been any discussion about this on the mailing lists?
2019-10-19 06:02:44 <dngray[m]> https://www.phoronix.com/scan.php?page=news_item&px=Fedora-32-Firewalld-Nftables
2019-10-19 06:04:18 <dngray[m]> fcolista_: maybe you know about this
2019-10-19 06:10:17 <dngray[m]> mm i did find this bug https://gitlab.alpinelinux.org/alpine/aports/issues/8744
2019-10-19 06:14:37 <dngray[m]> grep -r CONFIG_NF_TABLES= /media/mmcblk0p1/boot/config-rpi2
2019-10-19 06:14:38 <dngray[m]> CONFIG_NF_TABLES=m
2019-10-19 06:14:45 <dngray[m]> hmm sems to be a module at the moment
2019-10-19 06:16:16 <dngray[m]> i guess you'd want to make sure awall supports it
2019-10-19 06:17:14 <dngray[m]> i have to admit i've never used awall
2019-10-19 06:23:04 <kaey> any reason you want to keep using iptables instead of native nft tool and nftables openrc script?
2019-10-19 06:51:32 <_ikke_> kaey: maybe awall still relies on iptables, not sure
2019-10-19 08:53:10 <crockwork> WHAT UP GAMERS
2019-10-19 12:26:05 <telmich> moin moin
2019-10-19 12:26:43 <telmich> which is the right project to submit bugs to? The link from https://wiki.alpinelinux.org/wiki/Alpine_Linux:Contribute goes to gitlab general
2019-10-19 12:27:14 <telmich> concrete: there is a bug in the init script of named that makes it impossible to reload bind
2019-10-19 12:27:40 <_ikke_> aports
2019-10-19 12:27:54 <_ikke_> https://gitlab.alpinelinux.org/alpine/aports
2019-10-19 12:27:55 <telmich> thanks!
2019-10-19 12:29:38 <_ikke_> telmich: I've fixed the links on that page
2019-10-19 12:34:01 <mckaygerhard> _ikke_: sorry for that link.. i asked but nobody respond me .. was many weeks ago...
2019-10-19 12:40:35 <telmich> _ikke_: thanks a lot!
2019-10-19 12:44:58 <telmich> I'm trying to make my first package patch now for the bug reported - it should actually be not too hard to fix this
2019-10-19 12:45:27 <_ikke_> success
2019-10-19 12:56:20 <telmich> is this the right channel to ask apkbuild related questions or better -:ev?
2019-10-19 12:56:23 <telmich> -dev
2019-10-19 12:57:21 <_ikke_> -dev would be better
2019-10-19 21:14:08 <strfry> hi, is this `buildlab` thing from abuild still supported?
2019-10-19 21:15:23 <_ikke_> strfry: good question. I suppose it's not
2019-10-19 21:17:51 <strfry> _ikke_: i just googled my error message, and found a chatlog of me asking basically the same thing 2 years ago ;D
2019-10-19 21:18:18 <_ikke_> :D
2019-10-19 21:18:39 <strfry> someone said that i should use abuild rootbld instead, but it doesn't support cross-building
2019-10-19 21:46:45 <ahills> what arguments does apk pass to trigger scripts?
2019-10-19 21:46:50 <PureTryOut[m]> Why are locale environment variables like LC_ALL not set by default on Alpine?
2019-10-19 21:49:13 <mps> PureTryOut[m]: musl doesn't support locales
2019-10-19 21:49:42 <mps> default is C.UTF-8 iirc
2019-10-19 21:54:56 <mps> I played with perl to set locale and had some results
2019-10-19 21:56:06 <mps> also I rebuilt about two years ago coreutils and enabled locale, and got nearly working everything except LC_TIME
2019-10-19 21:57:08 <PureTryOut[m]> mps: hmm. It's a bit annoying cause it even makes some Flatpak stuff not run out of the box
2019-10-19 21:57:09 <PureTryOut[m]> LC_ALL=C would fix most things it seems
2019-10-19 21:58:35 <mps> PureTryOut[m]: look here for details, https://wiki.musl-libc.org/functional-differences-from-glibc.html
2019-10-19 21:58:54 <dalias> "doesn't support locales" isn't exactly accurate
2019-10-19 21:59:02 <dalias> what is the problem you're hitting?
2019-10-19 21:59:46 <mps> dalias: well, yes
2019-10-19 21:59:49 <dalias> what's breaking in flatpak?
2019-10-19 22:00:22 <mps> I wanted to say locales doesn't work as in glibc
2019-10-19 22:00:49 <PureTryOut[m]> dalias: well a Flatpak application requiring LC_ALL, but it isn't set by default on Alpine
2019-10-19 22:04:26 <dalias> "requiring LC_ALL" ?
2019-10-19 22:04:43 <dalias> sounds like a buggy application
2019-10-19 22:05:14 <PureTryOut[m]> Well, maybe it's fine with a specific LC_*, but LC_ALL is the easiest
2019-10-19 22:05:14 <dalias> applications aren't supposed to inspect LC_* at all. these variables just control how a locale is selected when you call setlocale(cat,"")
2019-10-19 22:05:39 <dalias> and the application should honor whatever the resulting locale from that ends up being
2019-10-19 22:06:06 <dalias> what's happening instead of that?
2019-10-19 22:06:41 <PureTryOut[m]> I have no clue what's happening internally. It's a proprietary application, which is why I'm running it in Flatpak in the first place
2019-10-19 22:07:05 <dalias> :-P
2019-10-19 22:07:33 <dalias> i'm not asking what's happening internally but what happens different from what you expect
2019-10-19 22:09:01 <dalias> btw it would be great if someone could write some content for the musl wiki about flatpak, appimage, etc. stuff and how well or poorly they work on a musl-based host
2019-10-19 22:09:27 <dalias> all these things tend to promise independence from host library ecosystem but i don't think any of them really deliver that
2019-10-19 22:09:45 <dalias> and documenting how/why they fail would be nice (and possibly lead to them getting fixed!)
2019-10-19 22:10:26 <PureTryOut[m]> Well, without LC_ALL, the application just crashes. With it set, it runs fine 😉 I made an issue for it on the specific Flathub packaging a few days ago https://github.com/flathub/com.mojang.Minecraft/issues/25
2019-10-19 22:11:27 <dalias> ok, it must just be doing something idiotic
2019-10-19 22:12:08 <PureTryOut[m]> I don't know 🤷‍♂️
2019-10-19 22:12:10 <dalias> you can set LC_ALL to anything you want except "C" or "POSIX" and all should be well
2019-10-19 22:12:25 <dalias> LC_ALL=C.UTF-8 is the most obvious/clean choice
2019-10-19 22:12:36 <dalias> but LC_ALL=en_US or fr_FR or anything would work
2019-10-19 22:12:43 <dalias> (all locale names are accepted)
2019-10-19 22:13:26 <PureTryOut[m]> Well LC_ALL=C works fine in this case too
2019-10-19 22:13:55 <PureTryOut[m]> It doesn't seem to like C.UTF-8.
2019-10-19 22:14:14 <PureTryOut[m]> 
ACTION sent a long message:  < https://matrix.org/_matrix/media/r0/download/matrix.org/wXIEDmuMEbXiSlWLcGnPHFpf >

2019-10-19 22:15:29 <dalias> oh right the app inside is a glibc app and wants a locale that its glibc knows about...
2019-10-19 22:15:36 <dalias> *sigh*
2019-10-19 22:15:43 <PureTryOut[m]> Exactly
2019-10-19 22:15:53 <dalias> and i guess they don't ship the glibc locale db
2019-10-19 22:16:09 <dalias> despite pretending flatpak is self-contained
2019-10-19 22:16:47 <dalias> so C is probably an ok choice for it, but don't put that in your environment for other apps or it will break their support for non-ascii text
2019-10-19 22:18:31 <PureTryOut[m]> I'm not 😉
2019-10-19 22:18:54 <PureTryOut[m]> Any clue where Flatpak's .desktop files are installed?
2019-10-19 22:21:20 <dalias> nope
2019-10-19 22:21:26 <dalias> i'd really love to have all this stuff documented
2019-10-19 22:21:36 <dalias> the technical side -- what works, what's broken, and why
2019-10-19 22:21:49 <dalias> rather than the marketing side which is the only thing you get from official sources
2019-10-19 22:23:24 <ahills> should it be possible to update the kernel on an external usb installation with mkinitfs?
2019-10-19 22:25:56 <mps> ahills: update kernel with mkinitfs?
2019-10-19 22:26:20 <mps> did you meant to say update initramfs
2019-10-19 22:26:30 <ahills> sorry, update initramfs
2019-10-19 22:26:55 <mps> mkinitfs -o outputfile
2019-10-19 22:27:17 <ahills> what normally puts new kernels in /boot?
2019-10-19 22:27:23 <mps> but not sure from where it will take modules
2019-10-19 22:27:41 <ahills> which repo has the ISO creation script(s)?
2019-10-19 22:27:48 <mps> apk add -u linux-vanilla, for example
2019-10-19 22:28:19 <mps> aports/scripts
2019-10-19 22:28:27 <ahills> presumably from boot/modloop-$variant
2019-10-19 22:28:30 <ahills> thanks
2019-10-19 23:03:40 <ahills> mps: it looks like the magic happens in the update-
2019-10-19 23:03:44 <ahills> kernel script from alpine-conf
2019-10-19 23:49:10 <ahills> wow, simply "upgrade-kernel" actually took care of the initramfs, kernel, modloop squashfs, etc
2019-10-19 23:49:29 <ahills> time to update https://wiki.alpinelinux.org/wiki/Upgrading_Alpine
2019-10-20 04:52:20 <bovis> Getting to the step here -- https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Grub_with_UEFI -- is installation with APK correct in this tutorial? Should it be apk -p /mnt/ grub ?
2019-10-20 04:52:30 <bovis> I ask because, when using the command in the tutorial, I don't see /mnt/etc/default/grub ever written, but the tutorial says to modify it.
2019-10-20 04:54:01 <bovis> Trying apk -p /mnt/ grub produces the error "failed to get canonical path of '/dev/vg0/root'" and exits the script
2019-10-20 06:24:42 <mps> ahills: did you read last line of the 'apk -h' => 'This apk has coffee making abilities.' :)
2019-10-20 06:25:58 <mps> I didn't understood your question or question was ambiguous
2019-10-20 06:28:00 <mps> alpine have 'hidden' tools which are invoked by apk (and other programs) and these tools can be invoked by admin/user
2019-10-20 06:29:43 <mps> apk can work with separate installation (i.e. mounted on usb, etc) by '--root' option
2019-10-20 11:12:11 <crockwork> Do we have Input as a font in alpine?
2019-10-20 11:13:04 <crockwork> Guess not. Hmm. https://input.fontbureau.com/info/
2019-10-20 11:36:23 <mps> crockwork: what is license for this fonts
2019-10-20 11:47:29 <jn__> mps: https://input.fontbureau.com/license/ esp. no redistribution
2019-10-20 11:51:21 <mps> jn__: thanks, so not free and can't be packaged for Alpine
2019-10-20 11:51:32 <jn__> correct
2019-10-20 12:09:59 <AinNero> uhm, one could make a APKBUILD in non-free
2019-10-20 12:10:02 <AinNero> but i think thats it
2019-10-20 12:10:13 <AinNero> non-free tends to be really badly maintained
2019-10-20 12:11:48 <Cogitri> For a font you could just copy it in your $HOME/.fonts and be done with it
2019-10-20 13:01:16 <mps> Cogitri: right, and I do that even for free fonts and themes, although I do not have much of them
2019-10-20 16:29:19 <strfry> where do the modloop files come from, how are they generated?
2019-10-20 16:44:17 <AinNero> /sbin/update-kernel
2019-10-20 17:14:33 <bovis> I'm trying to install grub while chrooted, and I get the error message "EFI variables are not supported on this system"
2019-10-20 17:14:53 <_ikke_> Did you mount the efi fs?
2019-10-20 17:16:09 <bovis> _ikke_: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Grub_with_UEFI I've done everything here, including the four mount commands in that step
2019-10-20 17:17:46 <_ikke_> bovis: This is archlinux, but looks relevant: https://bbs.archlinux.org/viewtopic.php?id=249546
2019-10-20 17:20:34 <TBB> at some point efivarfs was mounted ro by default
2019-10-20 17:22:11 <bovis> _ikke_ TBB: I guess I'll have a look at the boot of the USB. Maybe it booted automatically as legacy
2019-10-20 17:22:24 <_ikke_> yes, after someones pc got bricked because something did a recursive delete, including efivars
2019-10-20 17:23:16 <_ikke_> (and apparently deleting files from efivars can brick your pc)
2019-10-20 17:23:33 <TBB> ah, I remember this one faintly
2019-10-20 17:24:59 <maxice8> _ikke_: Some faulty firmware will brick the PC if efivars are deleted
2019-10-20 17:27:14 <MY-R> huh, can rly brick pc? there is no recovery mode or something, dunno... dual bios etc?
2019-10-20 17:29:11 <Cogitri> MY-R: Some mainboards don't have that feature
2019-10-20 17:29:50 <Cogitri> Effectively bricking them when you deleted efivars (unless you were in the mood of soldering a new BIOS chip on there :P)
2019-10-20 17:30:28 <MY-R> oh damn that suck, especially if can do it in such a easy way
2019-10-20 17:31:16 <MY-R> 20 years ago I was afraid to clear MBR sector on disk because thought that will break my computer for good!
2019-10-20 17:41:27 <bovis> _ikke_: SeaBIOS lets me select boot device, but not settings. I don't know if USB for sure will boot by default with UEFI or legacy. I know I haven't customized the coreboot install in any way.
2019-10-20 19:30:15 <bovis> Initial installation of Alpine is working. I can boot and decrypt LVM on LUKS volumes. But I get the warning: "Warning: locking directory /run/cryptsetup is missing". What is that, and how to fix?
2019-10-20 20:43:51 <darkfader> bovis: i'm not super certain but i would try to see if there is a /var/run/cryptsetup as opposted to /run
2019-10-20 20:45:43 <bovis> darkfader: If that's the case, can/should I just point cryptsetup or crypt* to that location? Looking at my arch linux install, this directory does exist, but it is empty and simply root protected
2019-10-20 20:46:32 <bovis> To be clear, on arch the /run..., not /var/run..., does exist
2019-10-20 20:47:06 <darkfader> bovis: i am not certain. i noticed that on systemd systems they attempt to use a per-daemon /run instead of the central /var/run
2019-10-20 20:47:20 <darkfader> if it
2019-10-20 20:47:33 <darkfader> is like that then you can try symlinking
2019-10-20 20:47:59 <darkfader> but i have not used luks for years, that's why i just wanted to recommend you look in that direction but can't advise more
2019-10-20 20:48:14 <bovis> darkfader: Thanks. I'll check it out
2019-10-20 21:01:07 <crockwork> mps: Ah, that explains it.
2019-10-20 21:01:27 <crockwork> Can we get a piracy community repo set up? I'll maintain it.
2019-10-20 21:02:40 <mps> crockwork: you are kidding :)
2019-10-20 21:22:11 <crockwork> mps: I mean, only partially kidding.
2019-10-20 22:13:42 <MY-R> crockwork, so what is your alpine torrent tracker address!? ;)
2019-10-20 22:15:59 <crockwork> lmao
2019-10-20 22:16:11 <crockwork> I don't have the drives set up, yet! ;D
2019-10-20 22:18:18 <MY-R> :P
2019-10-20 22:19:48 <crockwork> Do we package gnucash?
2019-10-21 08:14:02 <gspe> Hi guys, I'm new to alpine, I'm trying to change the default sshd port, I've edited /etc/ssh/ssh_config vwith Port portnumber but it doesn't work
2019-10-21 08:14:18 <gspe> seams to alway listen to 22
2019-10-21 08:14:30 <gspe> Am I missing something?
2019-10-21 08:16:32 <mps> gspe: sshd_config not ssh_config
2019-10-21 08:35:42 <gspe> mps: thanks, silly mistake
2019-10-21 08:36:53 <mps> gspe: np, we all make such mistakes
2019-10-21 08:59:08 <Kanibal> Hi! Is anybody else experiencing problems with alpine:edge and the gnutls package? It currently breaks every application that links to it: https://termbin.com/3j1m
2019-10-21 08:59:34 <Kanibal> musl apparently does not offer the secure_getenv function?
2019-10-21 08:59:52 <_ikke_> Kanibal: It's a new function
2019-10-21 09:00:03 <_ikke_> https://git.musl-libc.org/cgit/musl/tree/WHATSNEW
2019-10-21 09:00:42 <Kanibal> Oh, interesting! Perhaps the docker image isn't using the most recent musl?
2019-10-21 09:01:21 <_ikke_> You need musl 1.1.24
2019-10-21 09:01:52 <_ikke_> maxice8: maybe we should add gnutls to https://wiki.alpinelinux.org/wiki/Release_Notes_for_Alpine_3.11.0_(ideas)#musl_.3E.3D_1.1.24-r0_supports_GLOB_TILDE_and_secure_getenv ?
2019-10-21 09:03:17 <maxice8> Don't ask, just add it. It is an idea board
2019-10-21 09:03:25 <maxice8> Add anything you think can make into the release notes
2019-10-21 09:03:26 <maxice8> it is just there so when it comes time to write the release notes there is a place where all the important stuff is written down
2019-10-21 09:03:48 <_ikke_> Sure, was just verifying if it would make sense
2019-10-21 09:03:54 <Kanibal> https://termbin.com/8rnp - indeed. Docker image is using old musl
2019-10-21 09:04:24 <_ikke_> run apk upgrade
2019-10-21 09:04:29 <Kanibal> But I can add step to update the musl library in my Dockerfile as a temporary fix, yes
2019-10-21 09:04:35 <Kanibal> *a step
2019-10-21 09:12:21 <Kanibal> OK. $ apk upgrade fixes the issue. Thanks!
2019-10-21 14:26:36 <jailbox> are there any config settings to make "apk add $x" automagically add "$x"-doc packages  ?
2019-10-21 14:45:11 <KatolaZ> jailbox: apk add docs ?
2019-10-21 14:49:13 <_ikke_> indeed
2019-10-21 14:49:28 <_ikke_> the docs package automatically pulls in all -doc subpackages for installaed packages
2019-10-21 14:50:07 <jailbox> ahh thanks :)
2019-10-21 15:27:29 <fluffiebun> uhm
2019-10-21 15:27:29 <fluffiebun> test
2019-10-21 15:38:31 <fluffiebun> so i've been tryıng to install alpine with luks
2019-10-21 15:40:00 <fluffiebun> i've followed instructions on the wiki and nothing abnormal happened but the grub boots intro rescue by not being able to find the decrypted root by its UUID
2019-10-21 15:51:28 <master> please help. I want install or wine32bit to my Alpine edge 64bit or Steam. How do this?
2019-10-21 15:55:30 <master> holy. Slow channel, really ;)
2019-10-21 17:20:29 <strfry> what is the actual build trigger that listens to the changes of alpinelinux/edge master and pushes the build results to the edge repository?
2019-10-21 17:22:30 <_ikke_> strfry: we have an mqtt server where messages are published
2019-10-21 17:23:04 <_ikke_> so on push, a message is published which the builders use as a trigger
2019-10-21 17:33:45 <PICCORO> hey! wiki page https://wiki.alpinelinux.org/wiki/KDE does not specify that KDE plasma are only in edge.. (not stable) and only by community repos
2019-10-21 17:34:15 <_ikke_> PICCORO: The next alpine release it would be in stable
2019-10-21 17:41:28 <PICCORO> so then that wiki page must be marked as "ork in progress" .. why the wiki are not so taking in care! that's a reason why alpine linux are not so used as desktop.. _ikke_
2019-10-21 17:42:19 <_ikke_> For a month or 2
2019-10-21 19:22:20 <consus> Hi folks
2019-10-21 19:23:17 <consus> Does anybody have a mediawiki aport?
2019-10-22 00:21:23 <jroh> I'm running into an issue with chromium/chromium-chromedriver isn't working for me with the latest packages. The error seems to be that /usr/lib/chromium/swiftshader/libGLESv2.so is missing, but I no longer have the last revision to compare with. Normally the scripts use capybara (ruby lib) to talk to chromedriver, so I'm trying to validate that message. But, it feels like a regression of some sort.
2019-10-22 03:22:20 <KH405> Any plan on packaging podman?
2019-10-22 11:03:40 <shrizza> What does an error like this mean:
2019-10-22 11:03:41 <shrizza> ERROR: shotcut-19.10.20-r0: package mentioned in index not found (try 'apk update')
2019-10-22 11:03:52 <shrizza> I can't install it even after an apk update.
2019-10-22 11:05:23 <mps> shrizza: do you have testing repo pinned
2019-10-22 11:06:50 <shrizza> I have it enabled.
2019-10-22 11:08:21 <mps> then, probably 'apk add shotcut@testing'
2019-10-22 11:09:49 <_ikke_> shrizza: If you can give feedback if shotcut is working properly, I could move it to community
2019-10-22 11:10:06 <shrizza> Hm, still getting the same thing. Maybe my mirror is out of date or something?
2019-10-22 11:10:33 <shrizza> Going to try other mirrors.
2019-10-22 11:10:50 <_ikke_> I just updated it this morning
2019-10-22 11:11:04 <_ikke_> *last night
2019-10-22 11:12:48 <mps> what 'apk policy shotcut' says
2019-10-22 11:13:55 <shrizza> There we go. Changing the mirror worked.
2019-10-22 11:14:09 <shrizza> mps: Shall I switch back and try that command?
2019-10-22 11:14:18 <mps> no
2019-10-22 11:14:48 <mps> it just say if you pinned repo and in which repo is package
2019-10-22 11:15:20 <mps> you can invoke without changing /etc/apk/repositories
2019-10-22 11:16:18 <shrizza> _ikke_: Hm, it just exits with this fatal error: [Fatal  ] <> Cannot mix incompatible Qt library (version 0x50c05) with this library (version 0x50c04)
2019-10-22 11:16:20 <mps> in my case it is '@testing https://nl.alpinelinux.org/alpine/edge/testing'
2019-10-22 11:16:54 <shrizza> mps: I don't have it pinned like that, it appears.
2019-10-22 11:17:18 <shrizza> For the record, I was using: [Fatal  ] <> Cannot mix incompatible Qt library (version 0x50c05) with this library (version 0x50c04)
2019-10-22 11:17:20 <shrizza> Er.
2019-10-22 11:17:29 <shrizza> mirror.xtom.com.hk
2019-10-22 11:17:45 <shrizza> I'm now using dl-cdn.alpinelinux.org
2019-10-22 11:18:02 <_ikke_> edge is 11h behind on xtom.com.hk
2019-10-22 11:18:06 <_ikke_> https://mirrors.alpinelinux.org/#mirror27
2019-10-22 11:18:12 <mps> probably should be rebuilt
2019-10-22 11:18:42 <_ikke_> strange, I just pushed a new version last night
2019-10-22 11:18:52 <_ikke_> And I don't see any Qt related upgrades since then
2019-10-22 11:19:42 <mps> _ikke_: maybe builder is not upgraded
2019-10-22 11:26:07 <hcosta> hello
2019-10-22 11:26:26 <hcosta> anybody running sway on alpine ?
2019-10-22 11:28:10 <hcosta> After following the steps on the wiki, i get a error when trying to start sway. "XDG_RUNTIME_DIR is not set in the environment. Aborting"
2019-10-22 11:28:45 <hcosta> does anybody knows how to fix this ? Thnks
2019-10-22 11:33:05 <_ikke_> One option is by setting that environment variable
2019-10-22 11:39:16 <hcosta> should i create a dir under /run/user and give it 0700 and point it there?
2019-10-22 11:39:27 <hcosta> is that the common practice ?
2019-10-22 11:43:02 <_ikke_> Usually you have some kind of session manager that arranges that
2019-10-22 11:43:46 <Cogitri> Common practice would be to use elogind or pam_rundir which do this for you
2019-10-22 11:44:04 <hcosta> have installed elogind.
2019-10-22 11:49:05 <Cogitri> Restart or login out and in again it it should work then
2019-10-22 12:58:31 <Piraty> ARG_MAX is 16 times bigger on debian:buster than on alpine:3.10
2019-10-22 16:07:36 <grayhatter> How can I start services AFTER the TTY come up?
2019-10-22 16:14:47 <Kanibal> .ashrc
2019-10-22 16:15:21 <_ikke_> That requires you to login first
2019-10-22 16:15:38 <grayhatter> so, I'm currently waiting for wpa
2019-10-22 16:15:46 <grayhatter> I'd like to login while I wait
2019-10-22 16:15:54 <grayhatter> wpa_supplicant
2019-10-22 16:18:17 <_ikke_> K d
2019-10-22 16:18:24 <_ikke_> I don't think that openrc is that flexible
2019-10-22 16:19:31 <dalias> find what's waiting on wpa and make it not wait
2019-10-22 16:19:38 <grayhatter> shouldn't I be able to set wpa_supplicant to depend on agetty?
2019-10-22 16:19:53 <grayhatter> dalias: broken wifi drivers that take forever to come online
2019-10-22 16:19:59 <dalias> no
2019-10-22 16:20:05 <dalias> i don't mean what's causing it to come up slow
2019-10-22 16:20:16 <dalias> i mean what service has wpa_supplicant/network as its dependency
2019-10-22 16:20:18 <grayhatter> it hangs for a while trying to find a wireless interface
2019-10-22 16:20:21 <qman__> the system waits for wpa because something depends on it
2019-10-22 16:20:37 <qman__> you would have to alter that dependency
2019-10-22 16:21:37 <qman__> Which might or might not be possible while also making everything work as expected
2019-10-22 16:22:11 <dalias> there's no good semantic reason for anything to depend on network being up
2019-10-22 16:22:16 <grayhatter> agetty depends on local, local depends on *
2019-10-22 16:22:43 <dalias> ah that might be it
2019-10-22 16:22:56 <dalias> replace that dep on * with dep on just stuff that matters
2019-10-22 16:23:10 <grayhatter> lol > stuff that matters
2019-10-22 16:23:17 <dalias> :)
2019-10-22 16:23:26 <grayhatter> I miss one service... and I'm rebooting to a recovery shell :D
2019-10-22 16:23:34 <dalias> ?
2019-10-22 16:23:57 <grayhatter> if agetty depends on some magic to start up, and I miss that magic, agetty might refuse to start
2019-10-22 16:24:05 <grayhatter> then I'd have to boot from USB to fix that
2019-10-22 16:24:15 <dalias> but agetty shouldn't depend on anything
2019-10-22 16:24:24 <_ikke_> local just starts things in /etc/local.d
2019-10-22 17:07:55 <ddevault> can I edit /etc/apk/world manually and then run some command to make reality reflect my changes?
2019-10-22 17:08:09 <_ikke_> apk fix
2019-10-22 17:08:13 <ddevault> thanks
2019-10-22 20:04:13 <uderv> Hello, I am wondering what the ideal way to set up an embedded firewall with Alpine is? I have an x86 embedded machine, and I want to boot Alpine in diskless mode from a liveUSB with a second partition for the apkovl.
2019-10-22 20:05:19 <uderv> The problem is, I can't create a second partition on the same USB that I am booting it off of, because the x86 image is El Torito/iso9660.
2019-10-22 20:06:23 <uderv> I could also technically use a second USB for this, but this box has limited ports, so ideally I would want to have it on the same USB.
2019-10-22 20:06:51 <uderv> So, I guess two questions:
2019-10-22 20:08:43 <uderv> 1. Is there a relatively easy way to create a second partition for the apkovl on the same USB that the iso9660 is on, without manually creating a partition on the USB and extracting the filesystem fro the ISO?
2019-10-22 20:08:58 <uderv> from*
2019-10-22 20:09:01 <_ikke_> Is that even possible?
2019-10-22 20:09:27 <_ikke_> a bootable isofs partition alongside another partition?
2019-10-22 20:10:22 <uderv> 2. By default, Alpine does not mount any other partitions or block devices. Even if I did use a second USB, how would I point it to the APKOVL on that second USB?
2019-10-22 20:10:44 <uderv> @_ikke_ I have no idea.
2019-10-22 20:10:58 <uderv> Some searches for this on stackoverflow seem to suggest it is not.
2019-10-22 20:11:13 <_ikke_> Do you need 2 partitions / media devices?
2019-10-22 20:11:35 <uderv> Yes. I need one for the root FS, one for the APKOVL.
2019-10-22 20:12:07 <_ikke_> diskless mode means the rootfs is in ram
2019-10-22 20:12:23 <_ikke_> The usb drive is only used as bootmedia
2019-10-22 20:12:34 <_ikke_> (and modloop)
2019-10-22 20:12:48 <uderv> The specific reason for this: I want to be able to checksum the root FS and the APKOVL, such that if the box were ever compromised I could verify that the root FS and config had not been modified.
2019-10-22 20:13:13 <uderv> I know it loads it into RAM, but the source media it is loaded from is still mounted.
2019-10-22 20:13:23 <_ikke_> in /media/usb
2019-10-22 20:13:25 <_ikke_> not /
2019-10-22 20:13:56 <uderv> Ok, that is the rootfs. I thought that was where it might mount any additional devices. This clears this up a bit.
2019-10-22 20:14:30 <_ikke_> / lives in RAM, and any changes will need to be persisted in the APKOVL
2019-10-22 20:15:09 <uderv> In theory. In practice, it might be possible for an attacker to theoretically modify the root FS mounted at /media/usb.
2019-10-22 20:15:29 <_ikke_> That's not a / the rootfs
2019-10-22 20:15:37 <_ikke_> So yes, they can change the bootmedia
2019-10-22 20:15:46 <uderv> Yeah.
2019-10-22 20:15:54 <_ikke_> But not sure how having 2 media devices would solve that or make it auditable
2019-10-22 20:16:05 <_ikke_> because both can be altered
2019-10-22 20:16:34 <uderv> Yes, they can both be altered. But if they reside on a USB, and there are two partitions, then the procedure is as follows:
2019-10-22 20:17:00 <uderv> 1. Checksum the rootfs partition, compare checksum to official image.
2019-10-22 20:17:14 <uderv> 2. Checksum apkovl, compare to checksum of known good state.
2019-10-22 20:17:43 <_ikke_> You can do that with a single media device as well
2019-10-22 20:18:23 <uderv> With two partitions, yes.
2019-10-22 20:18:41 <_ikke_> even a single partition
2019-10-22 20:19:05 <uderv> If the APKOVL is stored on the same partition as the rootfs, then it changes the checksum of the rootfs partition, this I can't verify it against the checksum of the official image.
2019-10-22 20:19:48 <uderv> Hence my dilemma.
2019-10-22 20:20:27 <uderv> However, even if the checksum part was not a concern, I still can't save anything to the rootfs partition because iso9660 is a read only filesystem that doesn't like to be modified.
2019-10-22 20:21:25 <uderv> I have tried mount -o r,w as well, but it doesn't want to play nice.
2019-10-22 20:22:38 <_ikke_> I think the initramfs tries to find the APKOVL in the first partition of any mounted device
2019-10-22 20:23:08 <uderv> Ok, that helps. I know TinyCore does something similar, was not sure if Alpine worked the same way.
2019-10-22 20:23:19 <_ikke_> It uses nlplug-findfs
2019-10-22 20:23:39 <_ikke_> https://github.com/alpinelinux/mkinitfs/blob/master/nlplug-findfs.c
2019-10-22 20:24:12 <uderv> Ok, so then I could just save the APKOVL to an ext4 partition on a second USB key, and it will load the config from that on boot?
2019-10-22 20:24:41 <uderv> And it will automatically mount it and everything?
2019-10-22 20:24:47 <_ikke_> I think so, not sure
2019-10-22 20:25:43 <uderv> Ok, I will try that and report back. If this is what it ends up being, then I really overthought it.
2019-10-22 20:26:30 <_ikke_> man nlplug-findfs if you have mkinitfs-doc installed
2019-10-22 20:27:07 <_ikke_> Ok, does not contain that much info
2019-10-22 20:40:37 <uderv> Ok, that was it.
2019-10-22 20:40:46 <uderv> I feel like I must have missed this in the wiki somewhere.
2019-10-22 20:41:51 <uderv> Either way: there is still the minor annoyance of having to use two USB drives. Ideally, the x86_64 image would be separated out into a USB and a CD version.
2019-10-22 20:43:07 <uderv> Like, have the rootfs be something standard like ext4, and when you dd it, it just takes up 112MB of the USB media. Then you can expand the filesystem if you want, create other partitions, etc.
2019-10-22 20:43:19 <uderv> Time to get in the dev channel.
2019-10-22 20:43:46 <_ikke_> there is the setup-bootable script
2019-10-22 20:44:07 <uderv> Where could I find info about that?
2019-10-22 20:45:31 <uderv> Hmm... https://wiki.alpinelinux.org/wiki/Alpine_setup_scripts
2019-10-22 20:45:59 <uderv> I am gonna have to take a look at that. The only real requirement for my use case is that it must be deterministic and reproducible.
2019-10-22 20:46:08 <uderv> So, this might work.
2019-10-22 20:48:27 <_ikke_> there is setup-bootable lsdkfjas;ldkfjasldfkjas;ldkfjalskdfjkjdfasldkfjadslkfjalkdssdfsdfsdfsdfdfsdf
2019-10-22 20:48:31 <_ikke_> sorry about that
2019-10-22 21:28:29 <danieli> cat walk on the keyboard? :)
2019-10-22 21:32:42 <mps> danieli: strange cat, walk only on one row :)
2019-10-22 21:33:00 <mps> danieli: \o
2019-10-22 21:33:00 <danieli> good point, i didn't actually read into it
2019-10-22 21:33:21 <danieli> \o
2019-10-22 21:33:23 <danieli> oops
2019-10-22 21:33:23 <danieli> o/
2019-10-22 21:34:19 <mps> what's a difference between '\o' and 'o/'?
2019-10-22 21:35:41 <consus> Are there any guidelines for web apps in alpine?
2019-10-22 21:36:05 <consus> E.g. I want to package a mediawiki, what user should own the data?
2019-10-22 21:36:12 <consus> www-data?
2019-10-22 21:36:29 <consus> Or newly created "mediawiki" user?
2019-10-22 21:37:22 <mps> danieli: haven't seen your msgs here for some time. welcome back :)
2019-10-22 21:39:09 <mps> consus: I usually use www-data but can depend on use case
2019-10-22 21:39:39 <consus> nothing fancy, just cache and uploads
2019-10-22 21:40:21 <mps> it is good to know what are they by user and group name
2019-10-22 21:40:32 <consus> yap
2019-10-22 21:44:05 <danieli> mps: just realized I hadn't checked IRC in a while and decided to poke my head in
2019-10-22 21:44:07 <danieli> and one winks back :)
2019-10-22 21:44:12 <danieli> waves back*
2019-10-22 21:44:12 <danieli> sorry, tired
2019-10-22 21:58:25 <mps> np, take time to rest, good night
2019-10-23 00:33:10 <ndowens> for abuild, I did the abuild-keygen and steps it said, inside the aport I am working on, I do abuild -r but it gives builddeps failed
2019-10-23 00:34:10 <ndowens> tried another one to make sure it wasnt the one I was working on, and other gave same error
2019-10-23 00:42:49 <ndowens> nvm got it working
2019-10-23 00:54:13 <cim209> ndowens: you got node version manager to work?
2019-10-23 00:54:21 <cim209> nvm = node version manager
2019-10-23 00:54:44 <ndowens> oh lol, nvm for me, means nevermind :)
2019-10-23 00:54:55 <cim209> yes i know, i'm just teasing
2019-10-23 00:54:59 <ndowens> :p
2019-10-23 00:55:21 <cim209> if this was ##linux the mods there would correct you in a heartbeat
2019-10-23 00:58:29 <ndowens> I have not been there in years
2019-10-23 01:15:17 <ndowens> abuild looks so much like arch's pkgbuild file
2019-10-23 01:19:17 <ndowens> bbl
2019-10-23 01:57:35 <ndowens> what does 'abuild patch' do? i run it and it doesnt seem to do anything
2019-10-23 03:06:08 <shrizza> _ikke_: Please let me know if you want me to test anything else with shotcut.
2019-10-23 06:45:30 <sofandr> hi all
2019-10-23 08:55:55 <midasi> I have in /etc/apk/world several .makedepends entries from failed compilations. Is there a command to clean all of them at once, such as 'apk clean'?
2019-10-23 09:01:15 <mps> sudo apk del ".make*"
2019-10-23 09:03:31 <midasi> mps: thanks, I know that already. I thought there might be a general option to clean the environment
2019-10-23 09:04:58 <maxice8> midasi: apk del will automatically remove orphan packages
2019-10-23 09:05:08 <maxice8> but those packages are explicitly added (hence they are in /etc/apk/world
2019-10-23 09:05:32 <maxice8> if you do a lot of compilations you might want to use dabuild which uses docker to build each package in a minimal image of the system
2019-10-23 09:07:37 <mps> midasi: did you tried 'sudo apk del "*"'? :D . please do not try that
2019-10-23 09:09:12 <midasi> maxice8: thanks for the info. I'll take a look at dabuild then
2019-10-23 09:09:29 <midasi> mps: are you really using sudo? :p
2019-10-23 09:09:49 <mps> but, that is power of unix, you can do whatever you want. even 'cat /dev/zero > /dev/sda' is useful command sometimes
2019-10-23 09:10:30 <mps> midasi: yes, but thinking to switch to 'doas' for admin tasks
2019-10-23 09:14:42 <midasi> mps: yes, we use as well 'doas' on our servers. For some reason, it's only a community package in alpine
2019-10-23 09:17:15 <_ikke_> it got only added recently, and it also depends on how long a specific version of doas is supported
2019-10-23 09:59:00 <consus> doas does not have doasedit
2019-10-23 09:59:05 <consus> And that kinda sucks
2019-10-23 10:07:52 <detha> doas has -C, so creating a doasedit would be a few lines shell script
2019-10-23 11:55:25 <midasi> consus: that's right... for this case (and a few others) we still setup sudo and use sudoedit as a supplement to doas
2019-10-23 13:07:37 <markand> is that so hard to write a file with an editor?
2019-10-23 13:14:11 <Cogitri> You can build your own doasedit with inotify-tools and doas
2019-10-23 13:14:39 <Cogitri> Just copy to /tmp with doas, chown for your user, edit it, check for changes with inotify and if it changed mv with doas
2019-10-23 13:19:36 <maxice8> http://ix.io/1xRq enjoy
2019-10-23 13:39:27 <PICCORO> hey guys there's some work or almost a issue feature request for https://www.trinitydesktop.org/ ?
2019-10-23 13:54:55 <mps> do we need yet another desktop
2019-10-23 14:31:19 <TBB> as such it's not a bad one, it's basically KDE before it went to version 4 and became the monster it is today
2019-10-23 14:38:38 <maxice8> KDE 5 is a pretty good monster
2019-10-23 15:23:29 <mps> I prefer simple and small (and more secure as bonus of these two)
2019-10-23 19:50:37 <kaey> why do you need inotify? why not sync only when $EDITOR exits?
2019-10-23 20:16:52 <Cogitri> kaey: What has $EDITOR existing to do with it?
2019-10-23 20:17:04 <Cogitri> You use inotify to see if the file has changed and copy it over when it has been modified
2019-10-23 20:17:20 <Cogitri> Oops, read exits as `exists`
2019-10-23 20:17:45 <Cogitri> Anyway, with inotify you can sync even when the editor is still running, when you just press save but leave the editor running
2019-10-23 20:21:32 <kaey> does this feature warrants added complexity? and what happens when you type :w in vi?
2019-10-23 20:23:52 <kaey> though looks like script from maxice does not check if config is valid
2019-10-23 20:33:21 <Cogitri> How would it check if the config is valid?
2019-10-23 20:34:01 <Cogitri> You can edit any file with that script after all...
2019-10-23 20:35:23 <Cogitri> If you don't want to use inotify because that's too "bloated" for you can always just replace the inotify stuff with something ala `$EDITOR && mv $TEMP_FILE $ORIG_LOCATION`
2019-10-23 20:53:44 <detha> $EDITOR && doas -C $TEMPFILE && mv $TEMPFILE $ORIG
2019-10-23 20:54:28 <detha> otherwise you can still lock yourself out with syntax errors
2019-10-23 21:14:05 <maxice8> no clue what checking the syntax of sudo has anything to do with a replacement of sudoedit
2019-10-23 21:14:07 <maxice8> maybe confused sudoedit and visudo ?
2019-10-23 21:14:20 <kaey> yeah
2019-10-23 21:42:53 <ndowens> #freebsd
2019-10-23 21:43:24 <maxice8> ?
2019-10-23 21:43:52 <ndowens> was doing /join #freebsd but it cut off /join
2019-10-23 21:44:22 <ndowens> so how you been maxice8
2019-10-23 23:48:33 <maldridge> both logrotate and rsyslogd both claim ownership of /var/log/messages.  This causes an error in logrotate when it tries to read a duplicate rule for an already known log path
2019-10-23 23:53:13 <nicolaus> how can i list all the packages i installed manually
2019-10-23 23:53:15 <nicolaus> via apk add
2019-10-24 00:08:59 <cim209> is there a guide to setup multiple ip address?
2019-10-24 00:09:01 <cim209> for a vps
2019-10-24 00:09:24 <cim209> the wiki shows local IPs, or is this setup the same for public IPs?
2019-10-24 01:11:04 <agb> nicolaus: `apk info` https://wiki.alpinelinux.org/wiki/Alpine_Linux_package_management#Listing_installed_packages
2019-10-24 01:11:20 <nicolaus> thanks agb !!
2019-10-24 01:13:57 <cim209> so this is what i added https://kopy.io/K1QJt
2019-10-24 01:14:11 <cim209> the new ip doesn't show up when ifconfig or netstat
2019-10-24 02:45:53 <afontain[m]> hi
2019-10-24 02:50:58 <afontain[m]> I've got trouble to generate a custom iso, I'm following the instructions here `https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage` but with no success
2019-10-24 02:51:14 <afontain[m]> I've double checked the prerequisites
2019-10-24 02:53:15 <afontain[m]> wait, it works
2019-10-24 02:53:42 <afontain[m]> for some reason, when removing the workdir parameter, it suddenly works
2019-10-24 03:15:57 <wowaname> wasnt deadbeef in aports at some point? just making sure im not hallucinating
2019-10-24 03:46:55 <ndowens> seems gitlab may be having a issue, click New merge and get a 404
2019-10-24 04:38:10 <_ikke_> wowaname: yes, but it was moved to unmaintained
2019-10-24 04:38:18 <_ikke_> 199296e63294be05bd25c20e8e648e8817fc0272
2019-10-24 04:38:34 <wowaname> gotcha thanks
2019-10-24 04:39:05 <_ikke_> ndowens: where? If I click on 'new merge request' it works here
2019-10-24 08:29:50 <ndowens> _ikke_: i can do it on the merge requesr bar or by using the black menu bar New->New merge requesr
2019-10-24 08:30:02 <ndowens> request
2019-10-24 09:35:11 <fredrigu> Hi! I've created my own apk's (not using aports) but while installing them I get an error that the package conflicts on itself. Do anyone know why what this strange error could be?
2019-10-24 09:36:19 <fredrigu> https://paste.ubuntu.com/p/TntT5D9Zf2/
2019-10-24 09:48:58 <clandmeter> can you tpaste your apkbuild?
2019-10-24 09:57:01 <fredrigu> clandmeter: I'm not using apkbuild, but I guess the error is in my .PKGINFO files https://paste.ubuntu.com/p/93wwW9Wc6z/
2019-10-24 09:57:36 <clandmeter> fredrigu: how do you build packages without it?
2019-10-24 09:59:31 <fredrigu> clandmeter: using openembedded/yocto
2019-10-24 10:00:01 <fredrigu> I got some help here a few days/weeks ago on how to hand build apk-packages with abuild-tar
2019-10-24 10:00:44 <fredrigu> the data for .PKGINFO is generated by bitbake from it's recipes (bb-files)
2019-10-24 10:02:11 <fredrigu> the reason for this is that the only way to build apk-packages I've found is with abuild and abuild requires a apkbuild file. But all I need is "take this fakeroot/directory and make a package"
2019-10-24 10:03:01 <clandmeter> sorry i dont think i have time available to debug this.
2019-10-24 10:04:19 <fredrigu> clandmeter: alright, thanks for listening :)
2019-10-24 11:09:59 <ndowens> wonder why i cant submit PRs but others can?
2019-10-24 11:15:36 <Cogitri> There's no restriction on making PRs
2019-10-24 11:16:35 <ndowens> asking in gitlab  i dont know what it is  also cleared cache and such but no help
2019-10-24 11:17:34 <ndowens> trird using git pull-request clinas well  it looked liked it worked but it didnt show up so idk, it is weird
2019-10-24 11:39:00 <clandmeter> ndowens: what is your username on gitlab?
2019-10-24 11:50:02 <ndowens> clandmeter: ndowens
2019-10-24 11:55:12 <kmmndr> Hi all :-)
2019-10-24 11:55:34 <kmmndr> I'm trying to create my own package repository
2019-10-24 11:56:02 <kmmndr> I've followed the following links:
2019-10-24 11:56:07 <kmmndr> https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package
2019-10-24 11:56:24 <kmmndr> to create my packages
2019-10-24 11:56:49 <kmmndr> I've an apk file after build
2019-10-24 11:57:28 <kmmndr> And the APKINDEX.tar.gz seems to be signed with my own key, with is perfect for my tests
2019-10-24 11:57:51 <nicola_pav> hi. I am trying to run crond with -L /dev/stdout. When I run it, I get a lot of "1" (ones) in the beginning
2019-10-24 11:58:25 <nicola_pav> any idea how to solve it?
2019-10-24 11:58:39 <kmmndr> But I must be missing something as APKINDEX seems to be untrusted
2019-10-24 11:58:42 <kmmndr> Any idea ?
2019-10-24 11:59:05 <kmmndr> nicola_pav: Is it in a docker container ?
2019-10-24 12:01:49 <mps> kmmndr: did you put you key in /etc/apk/keys
2019-10-24 12:04:13 <kmmndr> Yes I did
2019-10-24 12:04:47 <kmmndr> apk update seems to try fetching .../testing/x86_64/APKINDEX, then .../testing/APKINDEX
2019-10-24 12:05:01 <kmmndr> the later is non-existant
2019-10-24 12:05:18 <kmmndr> Do I need to create an index in that folder too ?
2019-10-24 12:05:47 <mps> why do you '...' in repo line
2019-10-24 12:06:25 <mps> iirc, it must be absolute paths
2019-10-24 12:06:41 <kmmndr> I've cutted here only
2019-10-24 12:06:47 <kmmndr> The real path is:
2019-10-24 12:06:51 <kmmndr> http://mirror.adhoc-gti.com/packages/testing/x86_64/
2019-10-24 12:06:57 <mps> ah, ok
2019-10-24 12:07:47 <mps> did you build pkg with 'abuild -r' or something else
2019-10-24 12:08:12 <_ikke_> ndowens: You disabled merge requests in your fork
2019-10-24 12:08:24 <_ikke_> ndowens: that's why you couldn't make merge requests
2019-10-24 12:08:38 <nicola_pav> kmmndr: yes, docker container alpine
2019-10-24 12:09:28 <kmmndr> mps: abuild (without parameters, as I've installed dependencies first using `abuild deps`)
2019-10-24 12:09:47 <kmmndr> nicola_pav: The package I'm trying to build is supercronic :-)
2019-10-24 12:09:57 <kmmndr> https://github.com/aptible/supercronic
2019-10-24 12:09:59 <mps> then you have to run 'abuild index'
2019-10-24 12:10:10 <kmmndr> mps: in which folder ?
2019-10-24 12:10:14 <mps> to reindex repo
2019-10-24 12:10:28 <mps> in pkg build folder
2019-10-24 12:10:31 <nicola_pav> kmmndr: here is the output I am getting https://pastebin.com/cvNGwSq7
2019-10-24 12:16:47 <kmmndr> mps: The last lines displayed by `abuild` are mentionning an update of `testing/x86_64` folder and signature of index files without error
2019-10-24 12:17:15 <kmmndr> Running `abuild index` raise an error: `UNTRUSTED signature`
2019-10-24 12:17:29 <kmmndr> nicola_pav: strange
2019-10-24 12:20:59 <nicola_pav> kmmndr: indeed. it happens with -l 2. It does not occur with level 8 (default), 7 and 6
2019-10-24 12:21:54 <nicola_pav> kmmndr: it happens with level 5. I am testing now all levels
2019-10-24 12:24:00 <nicola_pav> it happens with levels 0 till 5
2019-10-24 12:24:57 <_ikke_> ndowens: I've enabled merge requests and pipelines again for your fork
2019-10-24 12:26:51 <ndowens> hmm what happened?
2019-10-24 12:27:43 <_ikke_> Not sure, but in your repo settings, some modules were disabled
2019-10-24 12:28:09 <ndowens> weird but thank you
2019-10-24 12:29:04 <_ikke_> https://gitlab.alpinelinux.org/ndowens/aports/edit
2019-10-24 12:57:08 <jusdepatate> Hello ! i tried to send an email to alpine-mirrors@alpinelinux.org and protonmail told me that the adress was invalid, is it a problem from me typing wrong address or from protonmail or from alpine ?
2019-10-24 12:58:44 <_ikke_> jusdepatate: The address has changed: https://lists.alpinelinux.org/~alpine/mirrors
2019-10-24 12:59:46 <jusdepatate> aah ok
2019-10-24 13:00:19 <jusdepatate> i tried the one wrote here https://mirrors.alpinelinux.org/
2019-10-24 13:00:33 <_ikke_> We need to update it
2019-10-24 13:01:02 <jusdepatate> thanks for info
2019-10-24 14:26:22 <kmmndr> mps: I've found, the problem was due my RSA key's filename (containing %40 instead of `@` char)
2019-10-24 14:28:59 <mps> kmmndr: good. did that happened as a result of abuild-keygen or something else
2019-10-24 14:29:37 <kmmndr> No it is only my fault, I wget it into a vanilla alpine docker image to test my repository
2019-10-24 14:30:19 <mps> ok, it is good you found the error
2019-10-24 14:30:31 <kmmndr> But it might be good to document this case. What do you think ?
2019-10-24 14:32:01 <mps> don't know, really
2019-10-24 14:33:02 <mps> how it happened to get %40 in key filename
2019-10-24 14:34:45 <kmmndr> mps: download a file containing `@` using wget
2019-10-24 14:35:29 <kmmndr> The `@` might be mandatory, or `%` forbidden
2019-10-24 14:47:18 <mps> ah, wget, url rewriting. now is everything clear
2019-10-24 17:51:06 <mps> "include/env.h:159:1: error: unknown type name 'ulong'; did you mean 'long'?", what we are missing from glibc based distros
2019-10-24 17:54:45 <Cogitri> https://github.com/lattera/glibc/blob/master/posix/sys/types.h#L148
2019-10-24 17:54:46 <Cogitri> Just a typedef
2019-10-24 17:57:42 <mps> looked there, so we missing it, iiuc
2019-10-24 17:58:20 <Cogitri> We're not missing it, upstream just assumes it's there :)
2019-10-24 17:58:37 <Cogitri> Defining it with an `ifnotdef` and upstreaming that sounds best
2019-10-24 17:59:16 <mps> but, is it posix or not?
2019-10-24 17:59:31 <mps> or glibc-ism
2019-10-24 18:00:21 <Cogitri> Don't think it's POSIX, posix is `unsigned long` or something like that
2019-10-24 18:01:51 <mps> yes, typedef unsigned long u_long, ulong;
2019-10-24 18:08:39 <dalias> https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_types.h.html
2019-10-24 18:09:25 <dalias> neither is posix but they're exposed under bsd/gnu source
2019-10-24 18:09:59 <dalias> https://git.musl-libc.org/cgit/musl/tree/include/sys/types.h#n67
2019-10-24 18:10:44 <mps> dalias: yes, I have sys/types.h open in vim and cursor on this line
2019-10-24 18:11:38 <mps> my solution is to redefine it in source (uboot-tools in this case)
2019-10-24 18:11:44 <dalias> ?
2019-10-24 18:12:02 <mps> 'typedef unsigned long ulong;'
2019-10-24 18:12:05 <dalias> you mean the application here wants to define its own type by that name?
2019-10-24 18:12:16 <mps> no, I redefine it
2019-10-24 18:12:18 <dalias> i don't follow what the problem is
2019-10-24 18:12:46 <mps> give me a minute
2019-10-24 18:14:18 <mps> here it is http://tpaste.us/Rz0b
2019-10-24 18:15:21 <mps> and adding 'typedef unsigned long ulong;' to include/env.h solves build
2019-10-24 18:15:45 <dalias> can you build with V=1 on make command line to see the cflags?
2019-10-24 18:16:20 <mps> a minute
2019-10-24 18:16:42 <dalias> hmm
2019-10-24 18:16:49 <dalias> looks like this file is just missing #include <sys/types.h>
2019-10-24 18:17:16 <_ikke_> kmmndr: the keyname is encoded in the package. So if the keyname changes, it's no longer considered
2019-10-24 18:17:43 <mps> dalias: http://tpaste.us/9NDV
2019-10-24 18:18:14 <mps> I tried to put it but then build says it cannot find sys/types.h
2019-10-24 18:19:00 <dalias> that's weird
2019-10-24 18:19:28 <mps> here is the source file http://tpaste.us/na1p
2019-10-24 18:21:17 <mps> agree, tried different tweaks but nothing helps
2019-10-24 18:21:36 <mps> except typedef in this source file
2019-10-24 18:22:13 <mps> tried build on debian and it builds without problem
2019-10-24 19:00:55 <oliv3r[m]> Hey Leo, you around?
2019-10-24 19:01:11 <maxice8> yes
2019-10-24 19:01:21 <maxice8> Leo@gitlab.a.o ?
2019-10-24 19:02:22 <oliv3r[m]> so confusing :)
2019-10-24 19:02:51 <maxice8> it is a transition period from maxice8 to Leo
2019-10-24 19:04:29 <oliv3r[m]> maxice8: Your removing the v prefix from the version tag actually makes sense, but we don't have a codestyle rule for this; which makes sense to add. I'll push your (now deleted) change myself, and see how we can word this in the codestyle
2019-10-24 19:04:37 <oliv3r[m]> maxice8: /nick leo :p
2019-10-24 19:05:15 <oliv3r[m]> /msg maxice8 this this work too?
2019-10-24 19:05:48 <maxice8> I'm using a bridge from IRC to riot(matrix) so problably not
2019-10-24 19:05:50 <maxice8> you want a pm ?
2019-10-24 19:06:04 <oliv3r[m]> me too
2019-10-24 19:06:35 <oliv3r[m]> I use fractal normally; but trying it in the web UI
2019-10-24 19:14:50 <oliv3r[m]> maxice8: So with regards to the versioning, right now, we have some rules in place with regards to versioning; abuild enforces some of these rules, but they are not clearly described are they, or more importantly, what should we do in case of a conflict of the rules?
2019-10-24 19:15:39 <oliv3r[m]> i think it is clear as day, as you said, that when using semver, tags/releases often (sensibly) prefix the version number with the 'v'; and I do agree in the case of semver, we should remove the v from the version field
2019-10-24 19:16:33 <oliv3r[m]> the difficulty is then of course, but what about the other cases? For example, some packages use the 'r' versioning scheme (I think it revers to svn revision); what if some author gets the wrong idea, and starts doing r123.12.1
2019-10-24 19:18:18 <oliv3r[m]> I want to start drafting some codestyle rules for pkgver;
2019-10-24 19:20:23 <maxice8> I think this is more suited for #alpine-devel
2019-10-24 19:20:47 <oliv3r[m]> if I knew that was there, I would have asked it there :D
2019-10-24 19:42:20 <oliv3r[m]> maxice8: !792
2019-10-24 20:21:40 <ndowens> looking at cve stuff, adobe has got some decent amount, but the one i kept seeing was "Easy Digital Downloads" and FusionPBX many times in new posts from mitre's twitter
2019-10-24 21:41:44 <Kremator> hello folks, im trying to install neofetch but it seems it isnt ont he repo
2019-10-24 21:41:49 <Kremator> i already did apk update
2019-10-24 21:42:13 <Kremator> https://prnt.sc/pnrbq4
2019-10-24 21:42:22 <maxice8> Added community repo ?
2019-10-24 21:42:58 <Kremator> i guess no? where can i read about that?
2019-10-24 21:43:59 <maxice8> https://wiki.alpinelinux.org/wiki/Enable_Community_Repository
2019-10-24 21:44:02 <Kremator> also, in which tool is apk based on? 1st party own developed or is from another tool like apt?
2019-10-24 21:44:30 <maxice8> it is from scratch
2019-10-24 21:46:31 <Kremator> also, in my repositorie's file i see there is a couple fo extra lines referring to "edge" version
2019-10-24 21:46:48 <Kremator> whats that? like an usntable branch same as debian has SID and unstable?
2019-10-24 21:47:12 <maxice8> edge is the development branch
2019-10-24 21:47:57 <Kremator> so more like a nigthly branch
2019-10-24 21:51:49 <iggy> I would say it's like sid
2019-10-25 00:57:21 <saintdev> I'm getting `Error relocating /usr/bin/mergerfs: _ZNSt7__cxx1119basic_ostringstreamIcSt11char_traitsIcESaIcEEC1Ev: symbol not found` When I try to run mergerfs
2019-10-25 01:05:01 <saintdev> my libstdc++ seems to be up to date, so I'm not sure why. (I think that's where that symbol comes from)
2019-10-25 01:13:30 <saintdev> It's fine when I install it in a fresh docker container, but not on my run-from-ram install
2019-10-25 04:11:56 <Herve-M> Hello all, I got a little problem on my little alpine today. I found out that after any major upgrade (and reboot), it start on the old `linux-hardened` (AL:v3.6) kernel.. I tried to remove it with `apk del` (AL:v3.10) but nothing change.. Do someone know how to remove it correctly?
2019-10-25 04:24:22 <_ikke_> if something still depends on it, it won't be removied
2019-10-25 04:24:25 <_ikke_> Herve-M:
2019-10-25 04:24:40 <_ikke_> Herve-M: did you already install linux-vanilla?
2019-10-25 04:27:00 <Herve-M> Yes for long
2019-10-25 04:27:25 <Herve-M> Was even running on it before the reboot
2019-10-25 04:27:37 <_ikke_> what does apk version linux-hardened return?
2019-10-25 04:28:19 <Herve-M> `Installed:                                Available:
2019-10-25 04:28:35 <_ikke_> ok, so linux-hardened is not installed
2019-10-25 04:28:42 <_ikke_> So i guess what you see is just a label
2019-10-25 04:28:56 <_ikke_> where do you see linux-hardenen?
2019-10-25 04:29:11 <Herve-M> Even if `uname -r` return `4.9.32-0-hardened`
2019-10-25 04:29:19 <Herve-M> ?*
2019-10-25 04:30:00 <_ikke_> can you check /etc/mkinitfs/mkinitfs.conf?
2019-10-25 04:30:19 <Herve-M> In `/boot/` if have all the old files, like `initramfs-grsec` and `initramfs-hardened` and `vmlinuz-hardened`
2019-10-25 04:30:59 <Herve-M> ``mkinitds.conf` has `features="ata base ide scsi usb virtio ext4"`
2019-10-25 04:31:32 <_ikke_> apk info -W /boot/vmlinux-hardened
2019-10-25 04:32:42 <Herve-M> Got an `ERROR: /boot/vmlinux-hardened: Could not find owner package`
2019-10-25 04:33:02 <_ikke_> so it looks like it got left behind
2019-10-25 04:33:24 <_ikke_> sorry, /boot/vmlinuz-hardenened
2019-10-25 04:33:29 <_ikke_> hardened
2019-10-25 04:34:08 <Herve-M> My bad too, got `/boot/vmlinuz-hardened is owned by linux-hardened-4.9.32-r0`
2019-10-25 04:34:28 <_ikke_> ok
2019-10-25 04:35:00 <_ikke_> does /boot/vmlinuz-vanilla exist as well?
2019-10-25 04:35:05 <Herve-M> Yes
2019-10-25 04:35:09 <_ikke_> ok
2019-10-25 04:36:23 <_ikke_> apk info -R linux-hardened
2019-10-25 04:37:54 <Herve-M> Return: `linux-hardened-4.9.32-r0 depends on:
2019-10-25 04:38:24 <_ikke_> grep linux-hardened /etc/apk/world
2019-10-25 04:38:45 <_ikke_> sorr
2019-10-25 04:38:47 <_ikke_> sorry
2019-10-25 04:38:58 <_ikke_> You should use small -r in the earlier command
2019-10-25 04:39:46 <Herve-M> I used cat, return nothing :/
2019-10-25 04:40:09 <Herve-M> For `linux-` => `linux-firmware-none
2019-10-25 04:40:54 <_ikke_> is this a vm or bare metal?
2019-10-25 04:41:31 <Herve-M> Seem my old setup got a problem
2019-10-25 04:41:37 <Herve-M> It's a bare metal
2019-10-25 04:42:08 <Herve-M> Installed at 3.3, upgraded at each minor version
2019-10-25 04:42:31 <_ikke_> Do you recall installing linux-firmware-none?
2019-10-25 04:43:09 <Herve-M> Yep, was one of my try today to break dependencies for `apk del` (who was returning nothing btw)
2019-10-25 04:43:39 <Herve-M> `apk del linux-hardened`*
2019-10-25 04:43:54 <_ikke_> ok, so lets start by removing that
2019-10-25 04:44:12 <_ikke_> apk del linux-firmware-none
2019-10-25 04:44:30 <Herve-M> done
2019-10-25 04:45:09 <_ikke_> and now apk info -r linux-hardened, to be sure
2019-10-25 04:46:15 <Herve-M> Got: `linux-hardened-4.9.32-r0 is required by:`
2019-10-25 04:47:13 <_ikke_> Not sure if it does anything, but can you try to run apk fix?
2019-10-25 04:47:50 <Herve-M> Got: `sudo apk fix --verbose
2019-10-25 04:48:08 <Herve-M> So nothing :/
2019-10-25 04:48:39 <_ikke_> It should at least return something
2019-10-25 04:48:47 <_ikke_> OK: 144 packages, 739 dirs, 9723 files, 327 MiB
2019-10-25 04:50:00 <Herve-M> It returned `OK: 398 packages, 3439 dirs, 32713 files, 2215 MiB` but it didn't perform any fix action
2019-10-25 04:50:34 <_ikke_> ok
2019-10-25 04:52:35 <_ikke_> what about apk del -rs linux-hardened
2019-10-25 04:53:21 <Herve-M> ` sudo apk del -rs linux-hardened` => `OK: 2215 MiB in 398 packages`
2019-10-25 05:08:04 <Herve-M> Got it: `sudo apk del -rs linux-grsec` > `(1/1) Purging linux-hardened (4.9.32-r0)
2019-10-25 05:11:31 <Herve-M> But still booting on `linux-hardened` and `linux-grsec` is still listed in `apk/world`
2019-10-25 05:59:45 <_ikke_> just remove them from world and run apk fix
2019-10-25 06:35:59 <Herve-M> The trick has done/fix it :-)  Thanks _ikke_ (y)
2019-10-25 06:36:16 <Herve-M> One beer / coffee on me.
2019-10-25 06:37:35 <_ikke_> :-)
2019-10-25 06:37:43 <_ikke_> I just had a coffee, so consider that done
2019-10-25 07:14:52 <terra> Guys, I recently upgraded my my cpu and motherboard (Ryzen 3600). After that, SDL2 apps, mostly games started behave incorrectly.
2019-10-25 07:14:56 <terra> https://discourse.libsdl.org/t/intense-stuttering-after-cpu-upgrade-on-alpine-linux/26762
2019-10-25 08:10:08 <fredrigu> clandmeter: it seems as my problem yesterday was that I'd a provides that was the same as pkgname
2019-10-25 08:20:17 <Cogitri> fredrigu: Huh, that's weird. It really only happens for SDL2 apps?
2019-10-25 08:29:58 <fredrigu> Cogitri: SDL2 apps?
2019-10-25 08:30:14 <fredrigu> Cogitri: are you confusing me with terra?
2019-10-25 08:30:22 <Cogitri> Derp, right
2019-10-25 08:30:45 <Cogitri> Guess it's coffee time again... :P
2019-10-25 08:32:08 <terra> Cotgiri: Yes. Only SDL2. SDL1 has no problem either.
2019-10-25 08:33:36 <Cogitri> Well, that's super weird
2019-10-25 13:38:06 <Fieldy> hi! i'm trying to install 3.10.3 into a VM on freebsd (freenas). three times now (all of the times i've tried), it hangs at setup-alpine right away. any ideas? https://imgur.com/a/Ud1Q8Mo
2019-10-25 14:44:42 <ndowens> Fieldy: how much memory are you allowing the vm
2019-10-25 14:46:23 <Fieldy> ndowens: 512MB
2019-10-25 14:46:57 <Fieldy> i tried some other distros too, and keep facing various other severe issues at install boot. something seems funky with freebsd vm stuff (bhyve)
2019-10-25 14:47:21 <Fieldy> i'm able to set them up fine within virtualbox with the same resources
2019-10-25 14:47:36 <Fieldy> so. i guess there are larger issues at play :)
2019-10-25 14:47:59 <ndowens> hmm when i did it in bhyve and qemu i generally did 2g  but yea probaby is
2019-10-25 14:48:48 <Fieldy> i could try that. i'd just rather not dedicate that much to it as the system has 16GB and all i'm really trying to do is something minimal for autossh stuff
2019-10-25 14:49:42 <Fieldy> dowwwn the rabbit hole :P
2019-10-25 14:49:47 <ndowens> i only have 8g in my system, didnt notice it slow my pc down, though i didnt use x in it
2019-10-25 14:50:07 <Fieldy> true. though it's a freenas system, RAM is very important to zfs stuff (filesystem)
2019-10-25 14:50:18 <ndowens> if i had left it as default i could not boot into thr vm
2019-10-25 14:50:19 <Fieldy> but i wil ltry it with 2G to see if anything changes
2019-10-25 14:51:02 <ndowens> i dont knoe that it uses what you set it to, probably just allows it to go that high if it has to
2019-10-25 14:52:21 <Fieldy> good point. and it will be almost nil on disk activity so cache wouldn't fill up or anything
2019-10-25 14:52:45 <Fieldy> seeing if i can get what i want in a fbsd jail first (ot i know)
2019-10-25 14:52:46 <ndowens> even if it doesnt work, helps eliminate min ram setting
2019-10-25 14:52:50 <Fieldy> but def prefer alpine!
2019-10-25 14:52:53 <Fieldy> yep
2019-10-25 14:53:12 <Fieldy> i mean i've installed alpine on 128MB ram, kinda figured 512 would be ok :)
2019-10-25 14:53:59 <ndowens> yea imagine it should work but idk  i tried booting alpine with def menory and it would not go past grub
2019-10-25 14:54:12 <ndowens> in qemu i think it was
2019-10-25 14:54:30 <Fieldy> yeah. vm stuff is handy but sometimes introduces unintended complexities :)
2019-10-25 14:56:52 <ndowens> i have freebsd and alpine and windows mostly for kid, took me a bit to get zfs setup on alpine because some issues i ran into
2019-10-25 14:58:04 <Fieldy> hehe i love me some alpine. but once in a while it's like *doh* *headsmack* :P
2019-10-25 14:58:38 <ndowens> first time for me to use it
2019-10-25 14:59:57 <Fieldy> i love minimal things so i was thrilled to see how little disk space and especially ram it uses from default install
2019-10-25 15:00:54 <ndowens> i like the packaging here, reminds me of pkgbuild and makes it easier to me
2019-10-25 15:01:42 <mps> alpine base runs fine in 256MB RAM
2019-10-25 15:02:11 <ndowens> so light, rolling(using edge) and friendly community, arch , when i last used it, people was getting rude
2019-10-25 15:04:09 <Fieldy> nod. i saw alpine uses about 64MB of ram at start with no cruft
2019-10-25 15:04:55 <mps> I didn't tried with less than 128MB
2019-10-25 15:05:48 <ndowens> with tmux,bspwm,chromium and some other services going, using less than 4g ram
2019-10-25 15:06:45 <ndowens> 3.2g is being used
2019-10-25 15:08:03 <mps> with firefox 'free -m' => Mem:           3873         837        1092    231        1943        2766
2019-10-25 15:09:09 <ndowens> and ofc zfs uses ram too, so oretty good
2019-10-25 15:10:04 <ndowens> i normally dont mis-type this much, on my iphone
2019-10-25 15:15:12 <Fieldy> suuuure :P
2019-10-25 15:15:26 <ndowens> :)
2019-10-25 17:57:08 <DEac-> hi.  what would be needed, that kea-dhcp-server would be a part of the next stable alpine linux?  it is already in edge for more than one year (different verstions).  i didn't tried the edge-version, i do not know, if it is stable enough.  i would test it, if it is needed.
2019-10-25 18:14:01 <maxice8> Anyone wants to test mpv 30 before i update it ?
2019-10-25 18:14:01 <maxice8> https://gitlab.alpinelinux.org/alpine/aports/merge_requests/827
2019-10-25 18:14:48 <mps> maxice8: I will
2019-10-25 18:16:27 <mps> do I need to rebuild libplacebo
2019-10-25 18:17:10 <mps> obviously I need
2019-10-25 18:17:28 <maxice8> to build it yourself not required
2019-10-25 18:17:43 <maxice8> but it needs to be in community/ so mpv finds it with the builders
2019-10-25 18:18:42 <mps> buiding it and if succeeds will build mpv
2019-10-25 18:19:32 <mps> building mpv now
2019-10-25 18:20:04 <maxice8> I watched episode 4 of Initial D final stage with forced vulkan, seems to be enough
2019-10-25 18:20:28 <mps> I'm building on aarch64
2019-10-25 18:26:06 <mps> it works on aarch64
2019-10-25 18:27:45 <mps> and if I can find how to use  panfrost gpu driver on aarch64 would be great
2019-10-25 19:01:59 <MY-R> maxice8, looks like mpv working fine but again they changed something in config settings pff :P
2019-10-25 22:32:19 <KennethDodrill> hi there. I seem to have run into a constraint issue with php7, nextcloud, and nextcloud-pgsql. I cannot update any of these packages for some reason, eventually I get down to the base constraint issue (nextcloud-pgsql) and it says that it `breaks: world[nextcloud-pgsql=17.0.0-r0]
2019-10-25 22:32:24 <KennethDodrill> hi there. I seem to have run into a constraint issue with php7, nextcloud, and nextcloud-pgsql. I cannot update any of these packages for some reason, eventually I get down to the base constraint issue (nextcloud-pgsql) and it says that it `breaks: world[nextcloud-pgsql=17.0.0-r0]`
2019-10-25 22:32:32 <KennethDodrill> Sorry. I thought that would edit.
2019-10-25 22:33:33 <cim209> this isn't discord or slack
2019-10-25 23:01:30 <KennethDodrill> @cim209 just habit I guess
2019-10-25 23:01:54 <cim209> i know, i'm on discord too
2019-10-25 23:06:02 <KennethDodrill> oops, just realized that the packages are in edge and I am not on edge
2019-10-25 23:07:23 <cim209> KennethDodrill: if you were on docker, you wouldn't have these limitations
2019-10-25 23:07:31 <cim209> i run 2 nextcloud instances in docker
2019-10-25 23:07:38 <cim209> but using the stable branch
2019-10-25 23:07:56 <KennethDodrill> That may be true, I will look into that
2019-10-25 23:07:59 <KennethDodrill> thank you
2019-10-25 23:08:15 <cim209> good luck
2019-10-26 02:59:26 <chunkypuffs> Hey guys I'm having a really odd issue
2019-10-26 03:00:28 <chunkypuffs> I'm trying to launch weston, but I get nothing but "Error relocating /usr/lib/libxkbcommon.so.0: secure_getenv: symbol not found"
2019-10-26 03:03:35 <chunkypuffs> any idea?
2019-10-26 03:03:47 <chunkypuffs> This is on a raspberry pi.
2019-10-26 03:04:22 <dalias> chunkypuffs, update musl package
2019-10-26 03:04:48 <dalias> alpine has lousy version deps tracking
2019-10-26 03:05:14 <chunkypuffs> dalias, This isn't a musl system, it's a glibc system, afaik
2019-10-26 03:05:32 <chunkypuffs> If I use Alpine in a container via Docker, the host system is using Glibc, what does this imply?
2019-10-26 03:07:43 <dalias> alpine is not a glibc system
2019-10-26 03:08:09 <dalias> if this is happening inside the container, the host outside of the container is not relevant
2019-10-26 03:11:23 <chunkypuffs> Well okay. I did not realise you could use both musl and glibc on the same system
2019-10-26 03:20:54 <chunkypuffs> dalias, What is the package I'm wanting to update?
2019-10-26 03:21:02 <chunkypuffs> just musl?
2019-10-26 03:21:16 <chunkypuffs> Why, if I just pulled the latest alpine base from dockerhub would this occur?
2019-10-26 03:23:07 <chunkypuffs> because updating the musl package didn't change anything whatsoever dalias
2019-10-26 03:26:40 <chunkypuffs> this is just a genuine bug I guess
2019-10-26 03:26:53 <chunkypuffs> dalias, libxkbcommon is supplied by the weston package on edge, yet throws this error on armv7
2019-10-26 03:27:10 <dalias> apk policy musl
2019-10-26 03:27:53 <dalias> will show what's installed vs available
2019-10-26 03:28:13 <chunkypuffs> since this is a dockerfile, I don't really know how to do what you just requested that I do
2019-10-26 03:28:31 <chunkypuffs> dalias, the only thing I'm installing is weston and musl, that's it.
2019-10-26 03:28:40 <chunkypuffs> and weston has a dependency of libxkbcommon on edge
2019-10-26 03:28:46 <chunkypuffs> https://pkgs.alpinelinux.org/package/edge/community/armv7/weston
2019-10-26 03:28:50 <chunkypuffs> I don't need to use the cli to figure that out
2019-10-26 03:28:57 <chunkypuffs> https://pkgs.alpinelinux.org/package/edge/main/armv7/libxkbcommon
2019-10-26 03:29:05 <chunkypuffs> libxkbcommon is available on edge, it's installing it as a dependency of weston
2019-10-26 03:29:08 <dalias> it looks like libxkbcommon was built against musl 1.1.24 but only 1.1.23 is installed
2019-10-26 03:29:09 <chunkypuffs> how is this anything less than a bug?
2019-10-26 03:29:26 <chunkypuffs> dalias, installed on my system?
2019-10-26 03:29:29 <chunkypuffs> okay, well maybe that's true..
2019-10-26 03:29:44 <chunkypuffs> is there a difference between `apk update && apk install musl` and `apk update musl`
2019-10-26 03:29:45 <chunkypuffs> ?
2019-10-26 03:29:45 <dalias> the system = the docker container, but yes
2019-10-26 03:29:59 <dalias> apk update musl is meaningless
2019-10-26 03:30:07 <dalias> apk update just means pull updated package lists
2019-10-26 03:30:16 <chunkypuffs> I'm making assumptions since I've never had to update a package in alpine before :)
2019-10-26 03:30:20 <dalias> apk add -u musl is how to update it
2019-10-26 03:30:26 <chunkypuffs> right, thanks
2019-10-26 03:32:57 <chunkypuffs> Okay I'm further along now dalias :D
2019-10-26 03:32:58 <chunkypuffs>  xkbcommon: ERROR: failed to add default include path /usr/share/X11/xkb
2019-10-26 03:37:36 <chunkypuffs> fixed that
2019-10-26 03:45:21 <chunkypuffs> Failed to load module: Error loading shared library /usr/lib/libweston-7/drm-backend.so: No such file or directory
2019-10-26 03:45:27 <chunkypuffs> anyone know why I'm getting this on a pi with the weston package?
2019-10-26 03:45:55 <chunkypuffs> I've got raspberrypi raspberrypi-libs raspberrypi-dev installed
2019-10-26 04:09:56 <grover1911> Hello! I was wondering if somebody could help me with an issue with VLC, just updated to 3.0.8 and for some reason it no longer launches with the GUI. It states that 'skins2 interface error: cannot instantiate dialogs provider'.  I have tried running VLC by executing; /usr/bin/vlc --reset-config --reset-plugins-cache
2019-10-26 04:10:15 <grover1911> No matter what I do, it launches into the CLI version instead.
2019-10-26 04:30:03 <ndowens> try apk fix vlc-qt
2019-10-26 04:32:39 <grover1911> That did not work unfortunately, but thank you for helping!
2019-10-26 04:34:04 <ndowens> dont take this as jerkish, i am about to go to bed, but found some hints that might help you https://lmgtfy.com/?q=skins2+interface+error+cannot+instantiate+dialogs+provider+vlc&s=g
2019-10-26 04:35:17 <grover1911> Haha, you have no idea how much I have used google/duckduckgo to search for answers, I event searched for answers on slackware forums, ubuntu forums, random blogs that have come across the issue.
2019-10-26 04:35:35 <grover1911> Most of the things that I have tried that people have said "Oh thanks that worked", did not work for me.
2019-10-26 04:35:56 <ndowens> ah ok, one said to remove some qt4 library from ldpath fixed theirs, i dont remember where i seen it
2019-10-26 04:35:59 <grover1911> Most of the solutions go back to using qt5, but I have reinstalled qt5-qtbase also
2019-10-26 04:36:24 <grover1911> I don't see any qt4 libraries for alpine?
2019-10-26 04:36:27 <ndowens> https://bbs.archlinux.org/viewtopic.php?id=110032
2019-10-26 04:36:57 <ndowens> i dont know, dont use qt/gtk stuff myself, atleast not on purpose, i use a WM
2019-10-26 05:19:01 <chunkypuffs> dalias, do you know if there's a way to exclude dependencies with alpine?
2019-10-26 05:19:10 <chunkypuffs> for example xfce4 depends on thunar, but I don't want thunar
2019-10-26 05:19:17 <chunkypuffs> is there a way to not even download thunar?
2019-10-26 08:09:05 <mps> chunkypuffs: xfce4 is metapackage
2019-10-26 08:11:04 <mps> remove it and all pkg on which it depends will be removed (if they don't depend on some other pkgs)
2019-10-26 18:06:55 <crosbymichael_> anyone run their root as zfs? i'm have it all setup but the initrd won't load the zfs module, says spl.ko is invalid format
2019-10-26 18:07:09 <crosbymichael_> the iso loads it just fine
2019-10-26 18:09:02 <grobe0ba> i really should've set my current file server box up like that, but i couldn't get alpine to boot ANY root with GPT instead of MBR.
2019-10-26 18:09:15 <grobe0ba> and MBR is not exactly bueno on a 3T drive...
2019-10-26 18:10:02 <crosbymichael_> after many hours i got it working with GPT
2019-10-26 18:10:24 <crosbymichael_> you just need to use grub instead of syslinux
2019-10-26 18:10:57 <grobe0ba> tried that, and couldn't get grub working either. may just have been having a bad day, plus it was >1 yr. ago, so maybe things are better now.
2019-10-26 18:11:15 <crosbymichael_> you just have to mount a lot of things for grub to be happy in a chroot
2019-10-26 18:11:37 <grobe0ba> ah.
2019-10-26 18:12:30 <grobe0ba> with that zfs root, are you using a seperate boot partition, or grub's zfs support to load from there?
2019-10-26 18:13:23 <crosbymichael_> hand crafted boot and efi partitions
2019-10-26 18:13:48 <Cogitri> crosbymichael_: Huh, weird, root on ZFS works just fine for me
2019-10-26 18:13:58 <Cogitri> You've installed linux-vanilla and zfs-vanilla, right?
2019-10-26 18:14:21 <crosbymichael_> as far as i know, i only have linux-vanilla
2019-10-26 18:14:36 <crosbymichael_> then the zfs modules
2019-10-26 18:14:51 <grobe0ba> the zfs modules are part of zfs-vanilla
2019-10-26 18:16:16 <crosbymichael_> i mean, everything is in the initrd, i just modprobe zfs and it says "spl.ko invalid module format"
2019-10-26 18:18:09 <grobe0ba> to quote Spock, 'Fascinating.'
2019-10-26 18:18:49 <crosbymichael_> Cogitri, did you have a guide that you followed for setting it up?
2019-10-26 18:19:19 <crosbymichael_> i actually got it setup with syslinux on a VM and it worked fine.  trying to get it setup on a laptop ( nvme drive, etc)
2019-10-26 18:19:36 <crosbymichael_> maybe the file just got corrupted
2019-10-26 18:19:42 <Cogitri> I wrote https://wiki.alpinelinux.org/wiki/Setting_up_ZFS_with_native_encryption which works for me
2019-10-26 18:19:53 <crosbymichael_> ok, i'm using the luks one
2019-10-26 18:21:45 <Cogitri> Ah, didn't see much reason to use LUKS now that ZFS natively supports encryptiot
2019-10-26 18:22:49 <crosbymichael_> i can give it a go
2019-10-26 18:30:21 <ndowens> crosbymichael_: Yea Mine is zfs on root
2019-10-26 18:31:14 <ndowens> mine didn't give that error, what I did was install alpine on one usb flash and booted into it and installed zfs and modprobe and then did the install on HD
2019-10-26 18:35:36 <Cogitri> I just used Ubuntu 19.04 to install it since that comes with ZFS
2019-10-26 18:38:11 <crosbymichael_> so just do a 19.04/19.10 install then rewrite root with alpine files? :)
2019-10-26 18:38:46 <ndowens> nah, dont install ubuntu :p, just do alpine install from the ubunto live cd/usb
2019-10-26 18:39:17 <Cogitri> Yup, that's what I did
2019-10-26 18:39:33 <ndowens> or could do it how i did, either way
2019-10-26 18:42:31 <crosbymichael_> i'll try to ubuntu replace first, seems easier
2019-10-26 18:43:09 <Cogitri> But why?
2019-10-26 18:43:30 <ndowens> Started to look up gitlab-runner and such; started my own local repo of abuilds I've done
2019-10-26 20:57:49 <wilywizard> new to alpine, following the "disk to old computer" method for installing. Is there a mechanism to skip the networking setup if I've already got it running? From a fresh boot the setup script doesn't seem to play nicely with wpa_supplicant and my old wifi_card (it wants -Dwext rather than nl80211) I was going to hack the setup script but I was wondering if there was a "cleaner" way to deal w/
2019-10-26 20:57:55 <wilywizard> issues like this
2019-10-27 00:38:36 <grayhatter> wilywizard: you mean setup, or init?
2019-10-27 01:59:37 <mckaygerhard> hey guys https://www.xfce.org/about/news/?post=1565568000 now xfce 4.14+ 4.16+ are pure gtk3 only.. alpine desktop are not more a "light flavor"
2019-10-27 05:30:44 <russkel> is there a working script to create nice APKBUILDs from pypi sources? there's something in abuild repo but it doesn't seem to be included in the abuild apk package
2019-10-27 05:46:21 <siesta> pls help im having problem connecting to x11vnc server on termux alpine-linux
2019-10-27 05:46:42 <siesta> im on android nougat
2019-10-27 05:47:30 <siesta> how to set this up
2019-10-27 08:52:20 <nepugia> termux does not support X11 nor vnc
2019-10-27 08:52:58 <nepugia> You will have to find a vnc client, or an X11 server application for android if you want to use either of these protocols to stream video from linux
2019-10-27 08:54:48 <nepugia> Is there some kernel module i have to load in order to use android debug bridge on alpine?
2019-10-27 09:03:02 <Cogitri> russkel: There's not really a nice way to generate an APKBUILD since Python dep specifying is all over the place, but you can use `newapkbuild` to get standard compliant templates
2019-10-27 09:03:51 <Cogitri> I don't think you need to load some kernel module, but I'm not sure if we have the udev rules to run adb without root, nepugia
2019-10-27 09:06:38 <nepugia> Cogitri: the adb server doesn't start for me, with or without root, so i doubt that the problem is udev
2019-10-27 09:07:28 <Cogitri> Hm, does it say something useful in its logs? I haven't used adb on Alpine yet
2019-10-27 09:17:04 <nepugia> 'could not install *smartsocket* listener: cannot assign requested adress'
2019-10-27 09:17:10 <nepugia> Not sure what that means :)
2019-10-27 09:39:45 <russkel> okay thanks Cogitri. If I have a bunch more of those to do I will hack something together that will do a bulk of the running around
2019-10-27 09:43:17 <nepugia> Figure it out ugh
2019-10-27 09:43:22 <nepugia> i didn't have lo up!
2019-10-27 09:54:27 <Cogitri> russkel: 👍
2019-10-27 09:57:52 <russkel> Cogitri: hmm that doesn't show in irssi/my term
2019-10-27 09:58:34 <russkel> guess I shouldn't be surprised with xfce4-terminal
2019-10-27 10:01:17 <nepugia> what does not show up?
2019-10-27 10:01:28 <russkel> the thumbs up emoji
2019-10-27 10:01:33 <Cogitri> The thumbs up emoji
2019-10-27 10:01:47 <nepugia> ah.. not for me either, wierd
2019-10-27 10:02:01 <russkel> I get the box with the hex numbers in it
2019-10-27 10:03:29 <russkel> this is a long shot, but do you know anything about the proj aport Cogitri ?
2019-10-27 10:03:53 <KatolaZ> russkel: does your terminal support UTF-8?
2019-10-27 10:03:59 <KatolaZ> are you using a UTF-8 locale?
2019-10-27 10:04:03 <Cogitri> Not much, I did review the PR for it though.
2019-10-27 10:06:01 <russkel> KatolaZ: yes I am using a utf8 locale, and xfce term does report to support it
2019-10-27 10:07:11 <russkel> Cogitri: I think too much black magic is going on in it. I use it for some projects and it seems it isn't set to compile for 32b platforms nor can I get the test suite to pass even on x86_64
2019-10-27 10:07:35 <russkel> seeing as it's used for geographic projections I can't help but think it be a good idea if their own test suite passes haha
2019-10-27 10:08:17 <Cogitri> Well, you're welcome to fix it up :P
2019-10-27 10:08:26 <russkel> Yeah I'll give it a shot
2019-10-27 10:08:46 <russkel> after opencv <groan utf8 emoji>
2019-10-27 10:10:45 <Cogitri> Heh, you're touching all the fun projects :P
2019-10-27 10:11:32 <russkel> Cogitri: what we do with them is fun, for sure haha
2019-10-27 11:34:47 <mps> I see these emoji's in irssi under st term but I think they shouldn't be used in IRC because some people can't see them
2019-10-27 11:40:22 <maxice8> They appear as boxes for the people that can't see in emoji ?
2019-10-27 11:41:20 <mps> depends how the client translate them, but usually boxes
2019-10-27 11:41:44 <mps> and terminal if client runs in terminal
2019-10-27 11:44:23 <_ikke_> I see boxes as well, urxvt + weechat
2019-10-27 11:46:07 <mps> just tried, also I see boxes with urxvt+irssi, in st+irssi I see emoji characters
2019-10-27 11:47:07 <_ikke_> What font is being used in each?
2019-10-27 11:47:34 <mps> in st I use cousine and monaco
2019-10-27 11:47:52 <mps> in urxvt monaco
2019-10-27 11:49:20 <mps> just tried cousine with urxvt but again only boxes
2019-10-27 11:50:59 <mps> oh, and I have symbola fonts in ~/.fonts
2019-10-27 11:51:26 <mps> emoji's I see are from this fonts
2019-10-27 11:52:16 <mps> and don't understand why urxvt can't use them
2019-10-27 11:57:28 <_ikke_> Me neither
2019-10-27 16:13:55 <grayhatter> I can't get wifi to come up on boot. the card works, but it takes some time for the driver to push the firmware to the card, I tried to recompile a kernel with my card built-in rather than as a module, but then the card straight up didn't work... any thoughts/advice?
2019-10-27 18:38:41 <siesta> hello
2019-10-27 18:38:46 <siesta> anyone
2019-10-27 18:39:28 <mps> siesta: just ask
2019-10-27 18:42:56 <siesta> can someone help me setup x11vnc i cant seem to make it work
2019-10-27 18:43:14 <_ikke_> siesta: did you read what nepugia replied to you when asked it here earlier?
2019-10-27 18:43:43 <siesta> sorry ikke i got dc
2019-10-27 18:44:19 <siesta> no _ikke_
2019-10-27 18:44:44 <siesta> what does he said?
2019-10-27 18:44:47 <_ikke_> http://tpaste.us/YpnM
2019-10-27 18:46:29 <siesta> but i can vnc using termux ubuntu
2019-10-27 18:47:40 <siesta> i can connect to it using realvnc viewer
2019-10-27 18:48:44 <siesta> please help i really like alpine linux
2019-10-27 18:49:18 <_ikke_> I have not experience with it myself, so I cannot really help with that
2019-10-27 18:49:59 <siesta> i think its possible
2019-10-27 18:50:24 <siesta> i install xorg server and lightdm with jwm
2019-10-27 18:51:07 <siesta> in termux ubuntu it works
2019-10-27 18:51:32 <KatolaZ> o_O
2019-10-27 18:51:40 <KatolaZ> does termux support VNC?
2019-10-27 18:51:48 <siesta> yes
2019-10-27 18:52:09 <siesta> i can connect o it using realvnc and tigervnc server
2019-10-27 18:52:20 <KatolaZ> then you know where the problem is
2019-10-27 18:52:51 <siesta> alpine has no tigervncserver
2019-10-27 18:52:59 <siesta> only x11vnc
2019-10-27 18:53:25 <siesta> i cant seem to make it work
2019-10-27 18:54:42 <siesta> i dont know how to setup x11vnc
2019-10-27 18:55:13 <siesta> i follow the tutorials in the net but i cant seem to make it work
2019-10-27 18:56:22 <siesta> can you point me to something
2019-10-27 18:56:34 <siesta> may be a tutorial
2019-10-27 18:56:58 <KatolaZ> siesta: can you connect to the x11vnc server from another client?
2019-10-27 18:57:08 <KatolaZ> just to understand if the problem is on the vnc server or client side
2019-10-27 18:57:45 <siesta> may be the problem is my x11 setup
2019-10-27 18:58:41 <KatolaZ> siesta: what error do you get?
2019-10-27 18:58:43 <siesta> i canti only have realvnc client on my android
2019-10-27 18:58:58 <siesta> no display:0
2019-10-27 19:00:40 <siesta> may be the problem is it cant execute my jwm window manager
2019-10-27 19:01:17 <siesta> can i use .xinitrc
2019-10-27 19:01:24 <siesta> with x11vnc
2019-10-27 19:01:39 <siesta> hmmm this is hard
2019-10-27 19:04:53 <siesta> 11vnc -localhost -display :0 -findauth -forever -loop -noxd
2019-10-27 19:05:12 <siesta> amage -repeat -rfbauth /home/harry/.vnc/passwd -rfbport 5901 -shared
2019-10-27 19:05:37 <siesta> x11vnc
2019-10-27 19:06:05 <siesta> that is my command for x11vnc
2019-10-27 19:09:35 <siesta> -auth guess: failed for display=':0'
2019-10-27 19:09:40 <siesta> got this error
2019-10-27 19:10:13 <siesta> when i chage -findauth to -auth guess
2019-10-27 19:27:24 <siesta> oww never mind i got tuts and i can connect no
2019-10-27 19:27:30 <siesta> now
2019-10-27 19:27:35 <siesta> yehey
2019-10-27 19:27:40 <siesta> thanks people
2019-10-27 19:29:52 <_ikke_> :)
2019-10-27 19:30:01 <_ikke_> What did you use?
2019-10-27 19:50:53 <siesta> Xvfb
2019-10-27 19:51:03 <siesta> to create a display
2019-10-27 19:51:21 <siesta> i can connect now
2019-10-27 20:15:25 <siesta> got it running now
2019-10-27 20:15:50 <siesta> thanks again guys alpine-linux rocks!
2019-10-27 20:36:17 <siesta> here it is https://ibb.co/nbVBzZx
2019-10-27 20:36:24 <siesta> now working
2019-10-27 20:55:56 <siesta> now i want mono csharp and vb
2019-10-27 20:56:03 <siesta> huhuhu
2019-10-27 20:56:15 <_ikke_> good luck
2019-10-28 01:45:17 <pixelherodev> I set up a system using a chroot, and everything works fine within it; I installed linux-vanilla and set up the configuration, but when I try booting I get an error while mounting the root device fails "No such file or directory." Given that the device node and target folder exist, did I forget to install a needed package?
2019-10-28 01:46:20 <pixelherodev> Actually, same error occurs if I attempt to mount *any* device
2019-10-28 02:02:24 <siesta> i got another problem it says GLX not supported
2019-10-28 02:03:01 <siesta> how to fix this
2019-10-28 02:12:28 <siesta> hello anyone
2019-10-28 02:12:37 <siesta> please help
2019-10-28 02:38:25 <grayhatter> siesta: might wanna ask your whole question again, and then likely wait
2019-10-28 02:38:51 <grayhatter> how can I allow [normal user] to write to /sys/class/backlight/brightness?
2019-10-28 02:49:34 <siesta> my opengl program wont execute it says GLX not supported
2019-10-28 02:50:46 <siesta> im running it in termux alpine Xvfb x11vnc real vnc client
2019-10-28 02:51:50 <siesta> all other GL xscreensavers not running GLX issue again
2019-10-28 02:54:46 <KH405> Anyone is able to run podman on alpine ?
2019-10-28 02:57:10 <siesta> what is podman?
2019-10-28 03:05:58 <siesta> wow podman is great
2019-10-28 03:06:30 <siesta> but i havent tried docker yet
2019-10-28 03:07:09 <siesta> having problem runing opengl programs in termux alpine
2019-10-28 03:07:57 <siesta> yesterday i cant connect to vnc but that was solved
2019-10-28 03:08:07 <siesta> now GLX problem
2019-10-28 03:16:06 <siesta> exit
2019-10-28 03:19:30 <KH405> Sorry siesta, yeah podman is great! Would like it to be packaged with alpine
2019-10-28 03:34:09 <siesta> im out of luck
2019-10-28 03:34:21 <siesta> cant find any solution
2019-10-28 04:43:56 <larbob> Hello, if packages are required to be licensed under FSF/OSI licenses how are packages like linux-firmware-nvidia allowed? Is it just because they’re blobs?
2019-10-28 04:44:00 <larbob> genuinely curious
2019-10-28 05:51:43 <grayhatter> larbob: what make you say packages have to be OSI licenses?
2019-10-28 06:07:58 <larbob> https://wiki.alpinelinux.org/wiki/Package_policies
2019-10-28 07:13:43 <grayhatter> interesting, it says should, not that it's required. Perhaps the firmware bins are released MIT? that's OSI approved, but doesnt' required release of source
2019-10-28 07:44:44 <crockwork> Should I be able to run kismet without packaging it?
2019-10-28 07:46:14 <_ikke_> Sure
2019-10-28 07:46:23 <_ikke_> Packaging is not required to run anything on Alpine
2019-10-28 07:47:54 <markand> are you happy with busybox in Alpine? I thought it was a good idea to use it until: http://lists.busybox.net/pipermail/busybox/2019-October/087554.html
2019-10-28 07:55:44 <clandmeter> markand: we are happy
2019-10-28 07:56:03 <_ikke_> markand: I'm not following? A segfault when trying to build with a LLVM9 means that busybox is not a good idea?
2019-10-28 07:56:21 <_ikke_> s/ a //
2019-10-28 07:58:26 <markand> no but I've asked on the musl channel as well and with llvm, some people said that busybox is "insert your badword here"
2019-10-28 08:00:04 <markand> I've discovered busybox while using alpine a long time ago and decided to create an experimental distribution using a pure llvm toolchain + musl and busybox too but now seeing these comments + this problem I was wondering if other people felt the same
2019-10-28 08:11:49 <_ikke_> Like clandmeter said, busybox has been workig fine for us so far
2019-10-28 08:12:19 <clandmeter> we are using gcc, so we dont have issues.
2019-10-28 08:13:08 <clandmeter> well not those issues :)
2019-10-28 08:13:40 <_ikke_> There is also toybox, but not sure how it compares to busybox
2019-10-28 08:30:16 <markand> it is cleaner but it is quite limited in term of components
2019-10-28 08:30:20 <markand> but quite promising :)
2019-10-28 08:30:28 <markand> much more consistent than busybox
2019-10-28 08:43:46 <winny> it looks like toybox has active development too
2019-10-28 08:44:37 <winny> a lot of novel homegrown components too
2019-10-28 08:46:25 <markand> active... yes and no
2019-10-28 08:46:38 <markand> there are trivial issues pending in the github for several years
2019-10-28 08:50:46 <manjaro-user_> Hello, on Alpine 3.9 I get ERROR: unsatisfiable constraints:libheif (missing):required by: world[libheif]
2019-10-28 08:54:48 <_ikke_> Yeah, 3.9 does not include libheif
2019-10-28 08:54:55 <_ikke_> 3.10 is the first release that has it
2019-10-28 09:04:27 <siesta> mysql segfaults damn
2019-10-28 09:05:34 <_ikke_> on edge?
2019-10-28 09:06:28 <_ikke_> https://gitlab.alpinelinux.org/alpine/aports/issues/10901
2019-10-28 09:06:52 <siesta> yes on edge
2019-10-28 09:07:01 <_ikke_> That ticket refers to a patch
2019-10-28 09:07:10 <_ikke_> not sure if it's related to the segfault
2019-10-28 09:07:59 <siesta> its installing mariadb
2019-10-28 09:08:41 <_ikke_> What is?
2019-10-28 09:09:12 <siesta> instead of mysql
2019-10-28 09:09:33 <_ikke_> Alpine currently only has mariadb
2019-10-28 09:33:22 <siesta> i dont know how to patch code
2019-10-28 09:33:38 <_ikke_> abuild has support for it :)
2019-10-28 09:35:49 <siesta> is there a mirror for working mariadb
2019-10-28 09:39:57 <siesta> damn too many bugs
2019-10-28 09:40:19 <siesta> i need an insecticide
2019-10-28 09:40:26 <_ikke_> siesta: try a stable branch :)
2019-10-28 09:40:42 <siesta> what version to try
2019-10-28 09:40:49 <_ikke_> start with 3.10
2019-10-28 09:40:54 <siesta> ok
2019-10-28 09:57:36 <siesta> i got error in apk cache clean it says package cache is not enabled
2019-10-28 09:59:32 <_ikke_> You need to create /etc/apk/cache to cache packages (or create a symlink)
2019-10-28 10:00:05 <jn__> on the other hand, when there's no cache, there's no need to clean it
2019-10-28 10:00:17 <_ikke_> indeed
2019-10-28 10:00:41 <_ikke_> And my suspicion is that siesta is trying to solve an issue that that command is not meant to solve
2019-10-28 10:11:20 <siesta> i change my mirror to v3.10 del mysql and add it again it still segfaults
2019-10-28 10:11:56 <siesta> i update and upgrade before del add
2019-10-28 10:12:10 <_ikke_> did you add --available to the upgrade command?
2019-10-28 10:12:19 <siesta> no
2019-10-28 10:12:30 <_ikke_> Do that first then
2019-10-28 10:12:33 <siesta> ok
2019-10-28 10:14:03 <siesta> its now downloading packages
2019-10-28 10:16:39 <siesta> net connection here is slow
2019-10-28 10:17:45 <siesta> its downgrading
2019-10-28 10:18:05 <_ikke_> yes, that's the idea
2019-10-28 10:18:17 <siesta> from edge to v3.10
2019-10-28 10:18:17 <_ikke_> Otherwise you get a mixed set of packages from edge and stable
2019-10-28 10:19:29 <siesta> oh ok
2019-10-28 10:20:56 <siesta> musl is nice
2019-10-28 10:21:17 <siesta> so minimalist
2019-10-28 10:28:34 <siesta> thanks _ikke_ i learn so much from you
2019-10-28 10:28:49 <_ikke_> siesta: you're welcome :)
2019-10-28 10:28:54 <_ikke_> siesta: I appreciate that
2019-10-28 10:33:07 <crockwork> Is running a snap package on alpine just as simple as downloading and running it? Why would it rely on anything if it's just a full package?
2019-10-28 10:34:42 <Cogitri> It needs AppArmor to work
2019-10-28 10:41:32 <crockwork> What's apparmor, cogitri?
2019-10-28 10:42:07 <crockwork> Also, kismet requires libmicrohttpd--do I install libmicrohttpd-static, libmicrohttpd, or libmicrohttpd-dev?
2019-10-28 10:42:21 <_ikke_> https://en.wikipedia.org/wiki/AppArmor
2019-10-28 10:45:33 <crockwork> Is this something not included in alpine?
2019-10-28 10:49:40 <crockwork> Also, figured out it was libmicrohttpd-dev. Now I'm looking for libsqlite3.
2019-10-28 10:50:15 <_ikke_> sqlite-dev
2019-10-28 10:52:00 <crockwork> Ah, sweet. Found that out just before you had said it. I assume that I need to append -dev when trying to build software, now? What's the difference between actually getting sqlite and getting sqlite-dev?
2019-10-28 10:52:21 <_ikke_> alpine splits up packages
2019-10-28 10:52:23 <crockwork> I can see why people started packaging things instead of just running ./configure a billion times.
2019-10-28 10:52:47 <_ikke_> so the -dev package containers the headers and lib that is necessary to build against it
2019-10-28 10:52:50 <_ikke_> https://pkgs.alpinelinux.org/contents?branch=edge&name=sqlite-dev&arch=x86_64&repo=main
2019-10-28 10:53:10 <Cogitri> crockwork: A security mechanism that snap uses for sandboxing
2019-10-28 10:55:15 <crockwork> Actually, now that I thinking about it, I don't know how dynamic linking usually works. Does it only matter for loading the code into memory? Does the entire lib have to be loaded into memory when this happens?
2019-10-28 10:57:29 <Cogitri> Yes
2019-10-28 10:57:59 <Cogitri> The entire lib is loaded into memory. If you did static linking you could optimize parts that aren't needed out of the lib with LTO though
2019-10-28 11:12:15 <mps> Cogitri: entire lib or just needed entries?
2019-10-28 11:13:01 <Cogitri> Without LTO the entire thing will be loaded IIRC
2019-10-28 11:15:08 <mps> not sure nowadays with modern linker (memory manufacturer friendly) but when I learned about linkers they pull just needed entries
2019-10-28 11:18:27 <mps> s/just/only/
2019-10-28 11:19:59 <fredrigu> So I'm trying to use apk to install some packages and get this output:
2019-10-28 11:20:04 <fredrigu> The following NEW packages will be installed:
2019-10-28 11:20:12 <fredrigu>   libc6 libz1
2019-10-28 11:20:23 <fredrigu> After this operation, 0 KiB of additional disk space will be used.
2019-10-28 11:20:37 <fredrigu> how is it possible that I install 2 new packages but not adding any files?
2019-10-28 11:21:38 <crockwork> Oh boy. After a great length of time compiling, it couldn't compile python.h. Hmm.
2019-10-28 11:21:41 <mps> LTO is something else
2019-10-28 11:23:31 <mps> fredrigu: libc6 on alpine?
2019-10-28 11:23:52 <fredrigu> mps: no, using apk to install a custom set of apk-files generated by yocto
2019-10-28 11:24:51 <mps> probably they are already installed or doesn't exists at all
2019-10-28 11:26:23 <fredrigu> they do exists, they are not installed previously (I've emptied root / and run a fresh apk add --initdb) and after apk add libz they are added to the apk database successfully and can later on be removed as well
2019-10-28 11:28:49 <mps> have no idea, don't know how your system is built and what it contains
2019-10-28 11:35:43 <fredrigu> mps: nah, it's kind of a special usecase. I hope that this will give some back to alpine in the end
2019-10-28 11:39:12 <mps> anyway, apk message you got says that it didn't installed anything new, but reasons for that can different
2019-10-28 11:39:58 <mps> couldn't find files to install, conflict with something, missing dependencies or something else
2019-10-28 11:41:19 <fredrigu> mps: I think I've found it, turns out that if the .PKGINFO file does not contain a size parameter the package is assumed to be empty and thus not installed but just added to the database. Not an error, just malformed packages.
2019-10-28 11:42:00 <fredrigu> my next issue is now that installing zlib is not looking for zlib.apk but for $ARCH/libz1-1.2.11-r0.0.apk
2019-10-28 11:43:06 <fredrigu> so I need to rename my file names (should be easy once I understand which parts is which) and then understand how alpine is handling a multiarch system
2019-10-28 11:43:18 <fredrigu> the last part can't be unique for my usecase
2019-10-28 11:43:43 <mps> alpine doesn't support multiarch
2019-10-28 11:45:56 <fredrigu> mps: thanks, then at least I don't have to look before inventing it myself :)
2019-10-28 11:47:10 <mps> yes :) and when you invent it inform us please
2019-10-28 11:50:55 <fredrigu> mps: of course
2019-10-28 17:32:45 <DCiple> Hi Everyone,
2019-10-28 17:32:45 <DCiple> Has anyone added a 2nd route table for the purpose of configuring a separate default gateway? I want all app specific traffic to use the second interface exclusively. I've googled around a bit and haven't come up with anything besides adding a static route in the Alpine docs. From what I've been reading, adding a gateway static route won't work e.g. 0.0.0.0/0 via [next hop int address]. This is why I'm exploring a secondary route t
2019-10-28 17:32:45 <DCiple> able to be applied to the interface.
2019-10-28 17:35:18 <KatolaZ> DCiple: what do you need to do exactly?
2019-10-28 17:35:47 <KatolaZ> if you have a point-to-point that manages connection to other nets, you have to declare a point-to-point route for that
2019-10-28 17:35:50 <KatolaZ> not a default gw
2019-10-28 17:36:01 <KatolaZ> but i might have misinterpreted your question
2019-10-28 17:38:13 <known-lizard> Hello, a brief question on the LBU/apkvol system. I'm trying to find the method to do this: "Restoring a system to a previously configured state (e.g. after reboot), including all previously installed packages and locally modified configuration files."
2019-10-28 17:39:13 <known-lizard> The LBU page documents how to create the .apkvol. Can someone help me with finding the documentation to use the created .apkvol? Thank you.
2019-10-28 17:41:00 <DCiple> Basically I have a docker host running various containers running docker networking via the primary interface. I want to add another docker container and use the docker host networking to translate the real ip on the secondary alpine interface directly into the docker container. I can add the 2nd interface and configure the static ip settings in Alpine however gateway setting is ignored. I'm guessing this is because you can only ha
2019-10-28 17:41:01 <DCiple> ve 1 default route per route table etc. Thus I'm exploring mulitple route tables.
2019-10-28 17:42:05 <DCiple> I haven't looked into point to point
2019-10-28 17:44:08 <DCiple> I just want to isolate the eth0 traffic to eth0 and eth1 traffic to eth1
2019-10-28 17:58:14 <kaey> DCiple: dont know how to configure second rt, but ive been using ip netns for a long time with great success
2019-10-28 17:58:16 <kaey> http://man7.org/linux/man-pages/man8/ip-netns.8.html
2019-10-28 18:01:08 <DCiple> That seems exaclty what I need. I'll check it out and pop back if I have any questions. Thank you all!
2019-10-28 18:06:32 <mps> youtube-dl in edge is broken. anyone with python knowledge would fix it
2019-10-28 18:16:28 <clandmeter> Would it could it should it?
2019-10-28 18:17:52 <mps> clandmeter: I don't have dict irssi plugin, yet :)
2019-10-28 18:18:01 <_ikke_> mps: what is broken about it?
2019-10-28 18:18:14 <clandmeter> :)
2019-10-28 18:18:50 <mps> 'ImportError: cannot import name 'load_entry_point' from 'pkg_resources' (unknown location)'
2019-10-28 18:19:13 <clandmeter> Setuptools?
2019-10-28 18:19:21 <_ikke_> yes, that's setuptools
2019-10-28 18:19:25 <mps> I don't know
2019-10-28 18:19:30 <_ikke_> apk add setuptools
2019-10-28 18:19:32 <_ikke_> try that
2019-10-28 18:20:07 <Piraty> can apk search revdepends like xbps-query -RX foo ?
2019-10-28 18:20:17 <mps> py3-setuptools?
2019-10-28 18:20:27 <_ikke_> mps: indeed
2019-10-28 18:20:41 <_ikke_> Piraty: apk info -r foo
2019-10-28 18:21:12 <mps> _ikke_: yes, that solves problem
2019-10-28 18:21:36 <_ikke_> Ok, so it should be added as a dep
2019-10-28 18:21:39 <mps> shouldn't it be in depends field
2019-10-28 18:21:45 <mps> ahm, ...
2019-10-28 18:22:35 <clandmeter> It should be listed in deps
2019-10-28 18:22:41 <clandmeter> In the project
2019-10-28 18:22:47 <_ikke_> Working on it
2019-10-28 18:23:04 <clandmeter> It's easy to verify
2019-10-28 18:23:21 <_ikke_> clandmeter: what do you mean?
2019-10-28 18:23:29 <_ikke_> clandmeter: isn't it already clear that it's required?
2019-10-28 18:23:42 <clandmeter> Not to mps
2019-10-28 18:23:50 <clandmeter> Only after we tell him
2019-10-28 18:24:40 <clandmeter> I mean it's easy to find out if you are missing drps
2019-10-28 18:24:48 <Piraty> _ikke_: thanks
2019-10-28 18:25:10 <clandmeter> Everyone who does py packaging should check it
2019-10-28 18:25:30 <_ikke_> But it's hard to verify without installing the package
2019-10-28 18:25:39 <clandmeter> But a lot just bump pkgver
2019-10-28 18:25:44 <_ikke_> nod
2019-10-28 18:26:14 <mps> clandmeter: why do you think it is not clear to me?
2019-10-28 18:26:28 <_ikke_> http://tpaste.us/xkWQ
2019-10-28 18:27:44 <clandmeter> mps: sorry i should not use you as an example
2019-10-28 18:28:02 <_ikke_> pushed
2019-10-28 18:28:06 <clandmeter> I guess you are the user not the maintainer
2019-10-28 18:28:19 <_ikke_> mps: can you verify later that it fixed it for you?
2019-10-28 18:28:47 <mps> ofc, I will
2019-10-28 18:29:11 <_ikke_> locally it works at least
2019-10-28 18:29:40 <Piraty> I'm a bit lost, i require some fonts obviously, after installing xorg-server. is there a minimal xorg font package?
2019-10-28 18:35:39 <_ikke_> afaik there is no meta package for fonts
2019-10-28 18:36:17 <Piraty> is there a meta package for full blown xorg that i could bisect?
2019-10-28 18:36:36 <Piraty> revdepends on xorg-server don't show me anything though
2019-10-28 18:36:43 <_ikke_> we have a setup-xorg-base script, but I don't see it installing any fonts
2019-10-28 18:37:18 <_ikke_> https://pkgs.alpinelinux.org/packages?name=*font*&branch=edge&arch=x86_64
2019-10-28 18:37:21 <Piraty> maybe fonts are part of desktop envs ?
2019-10-28 18:38:22 <Piraty> yeah i got the list by "apk search font", but i thought there's a default-ish set of fonts most graphical applications can use without much configuration
2019-10-28 18:38:59 <_ikke_> When I follow this: https://wiki.alpinelinux.org/wiki/XFCE_Setup
2019-10-28 18:39:05 <_ikke_> I don't have to install any fonts
2019-10-28 18:39:56 <Piraty> weird
2019-10-28 18:40:13 <_ikke_> So something is pulling in fonts
2019-10-28 18:41:14 <Piraty> xfce4 pulls ttf-dejavu at least
2019-10-28 18:41:18 <Piraty> thanks, i'll try
2019-10-28 18:41:37 <_ikke_> Ah, missed that
2019-10-28 18:42:48 <Piraty> ah much better looking :)
2019-10-28 18:42:50 <Piraty> i'm preparing a docker container based on alpine that should just jump to an application (firefox for example)
2019-10-28 18:42:57 <Piraty> thanks ikke for holding my hand
2019-10-28 18:43:14 <_ikke_> np
2019-10-28 18:43:24 <artok> xorg-server pulls font-misc-misc
2019-10-28 18:43:52 <_ikke_> Those aren't any fonts
2019-10-28 18:43:58 <_ikke_> https://pkgs.alpinelinux.org/package/edge/main/x86_64/font-misc-misc
2019-10-28 18:44:00 <Piraty> i got many artifacts nevertheless, ttf-dejavu works alright
2019-10-28 18:44:24 <_ikke_> fonts-misc-misc are just font tools/libs
2019-10-28 18:44:49 <artok> ah yeah
2019-10-28 18:46:17 <_ikke_> it contains a lot of pcf files though
2019-10-28 18:46:27 <_ikke_> But I guess those aren't suited for a graphical environment
2019-10-28 19:16:12 <laser`> hi all, had a search but couldn't immediately find anything, I was wondering if there was a way to create an Alpine ISO image (for a USB stick) that would automatically kick it into a fully unattended install (so I guess either something that automatically runs a script after booting or an automatic way of providing an answerfile)?
2019-10-28 19:16:52 <Piraty> lol. /home/buildozer
2019-10-28 19:17:41 <_ikke_> Piraty: ?
2019-10-28 19:18:20 <Piraty> firefox in my container crashes and shows the path to the code file where it crashed
2019-10-28 19:18:30 <Piraty> the builder's homedir i assume
2019-10-28 19:18:32 <_ikke_> aha
2019-10-28 19:18:34 <_ikke_> yes
2019-10-28 19:19:36 <known-lizard> I had the same type of question :laser`, but haven't found a solution.
2019-10-28 19:19:49 <laser`> ah :(
2019-10-28 19:20:50 <_ikke_> setup-alpine does allow an answer file, but not sure you can configure an unattended setup
2019-10-28 19:23:33 <laser`> I was wondering if an apkovl could be used to do it, drop one on the USB stick that contains a script and get init to run it after boot, but hadn't gone down that route
2019-10-28 19:27:34 <laser`> ah, https://github.com/alpinelinux/alpine-make-vm-image looks promising :)
2019-10-28 19:32:28 <free_mind> Can someone provide me a recommendation for a rpi3 or rpi4; install the aarch64, armhf or armv7? Is hardware floating point instructions important? Both rpi3 and rpi4 have hardware support for 64bit instructions. I don't know what is wisdom is... :)
2019-10-28 19:34:50 <grayhatter> https://wiki.alpinelinux.org/wiki/Raspberry_Pi
2019-10-28 19:34:54 <_ikke_> free_mind: There is no support for rpi4 as of yet
2019-10-28 19:40:12 <free_mind> Thank you for the link grayhatter. Is there a difference is the amount of packages support between aarch64 and armv7?
2019-10-28 19:40:44 <grayhatter> free_mind: I don't think so, but having spent hours working in the aports tree already, I wouldn't care
2019-10-28 19:40:56 <grayhatter> it's SO easy to build packages with abuild
2019-10-28 19:41:25 <_ikke_> free_mind: Some packages don't have 32 bit support, those would be missing on armv7
2019-10-28 19:41:36 <_ikke_> grep aports for "!armv7" to see what's disabled
2019-10-28 19:41:40 <free_mind> _ikke_: Do you know what the schedule is for rpi3?
2019-10-28 19:41:51 <free_mind> typo: rpi4
2019-10-28 19:41:52 <_ikke_> rpi4 you mean?
2019-10-28 19:41:54 <_ikke_> right
2019-10-28 19:41:56 <_ikke_> not sure
2019-10-28 19:42:20 <grayhatter> what's broken on the rpi4?
2019-10-28 19:42:22 <_ikke_> I think some users are trying to get it o work
2019-10-28 19:42:51 <_ikke_> It might need a dedicated kernel, but I don't know all the details
2019-10-28 19:50:25 <free_mind> Need to look into aports/abuild grayhatter; first time I read about it. I'm reading the wiki page on APKBUILD_Reference and APKBUILD_examples; Do you have a tip for a nice intro, since those wiki pages are technically a bit overwhelming right now.
2019-10-28 19:50:50 <grayhatter> yeah
2019-10-28 19:50:54 <grayhatter> ever use git?
2019-10-28 19:51:17 <free_mind> yes
2019-10-28 19:51:21 <_ikke_> free_mind: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package
2019-10-28 19:51:47 <grayhatter> git clone aports.git -> then
2019-10-28 19:52:00 <grayhatter> cd dir/with/your/pkg
2019-10-28 19:52:02 <grayhatter> abuild
2019-10-28 19:52:21 <_ikke_> That wiki page has some info on how to setup your system
2019-10-28 19:52:31 <grayhatter> you'll have to have the build tools installed first obviously, but as you've used git I assume you'd know that
2019-10-28 19:53:04 <grayhatter> that's the guide I followed
2019-10-28 19:54:23 <grayhatter> free_mind: first read/skim through the whole thing, then try to follow step by step... if you do get stuck somewhere, join IRC, link to the guide your following, explain the package you're trying to build, then list the error you're having with any output (try not to spam the channel)
2019-10-28 19:54:46 <grayhatter> someone will be able to help you, you can even highlight me, but I might not be at my computer
2019-10-28 19:57:53 <free_mind> Will look into the provided information grayhatter _ikke_  Great community here very helpful and friendly as well.
2019-10-28 19:58:03 <free_mind> :)
2019-10-29 00:17:09 <pixelherodev> Figured out why Alpine wasn't booting for me - I was booting through refind, and didn't realize I needed to specify `modules=ext4`. Everything's fine now ;)
2019-10-29 01:00:22 <pixelherodev> ... okay, am I hallucinating, or is qutebrowser working under Alpine now? If so, to whom may I direct my utmost appreciation?
2019-10-29 05:33:42 <_ikke_> pixelherodev: last change was an upgrade to 1.0.1
2019-10-29 07:54:40 <crockwork> pixelherodev: what's qutebrowser?
2019-10-29 07:55:00 <cim209> probably a Qt based browser
2019-10-29 07:56:50 <Cogitri> pixelherodev: qa3Txu0iak0F fixed QtWebEngine and as such fixed QuteBrowser :)
2019-10-29 08:07:28 <siesta> how can i source .profile in starting termux alpine in proot
2019-10-29 08:10:05 <Cogitri> Depends on your $SHELL, I suppose
2019-10-29 08:11:31 <siesta> i echo $SHELL its /bin/sh
2019-10-29 08:11:42 <cim209> what? not /bin/ash?
2019-10-29 08:11:46 <siesta> no
2019-10-29 08:12:07 <cim209> sourcing a file is a bash feature, i believe
2019-10-29 08:12:11 <siesta> it is on proot
2019-10-29 08:12:48 <siesta> proot calls /bin/sh --login
2019-10-29 08:13:16 <cim209> zsh can source too
2019-10-29 08:13:19 <cim209> not sure about ash
2019-10-29 08:14:17 <siesta> sh can source too when i type source ~/.profile
2019-10-29 08:14:30 <siesta> but how can it be automatic
2019-10-29 08:14:32 <cim209> oh it can?
2019-10-29 08:14:36 <siesta> when i login
2019-10-29 08:15:03 <Cogitri> siesta: https://unix.stackexchange.com/questions/176027/ash-profile-configuration-file
2019-10-29 08:23:28 <siesta> oh ok i got it now
2019-10-29 08:23:34 <siesta> thanks
2019-10-29 08:23:49 <siesta> it loads /etc/profile on default
2019-10-29 08:24:02 <siesta> i put my script there
2019-10-29 08:27:58 <siesta> ok it works now
2019-10-29 08:32:20 <Cogitri> 👍
2019-10-29 09:53:35 <crockwork> Does anyone have a clue as to why polybar's "network you're connected to" widget is stuck on the previous network I was associated with?
2019-10-29 09:54:07 <crockwork> It goes away after reboot, but the steps to reproduce are basically to just disconnect normally from a network and then associate to a new network.
2019-10-29 14:12:58 <DeLord> Hi, I am having trouble adding a user to more than 31 groups in an Alpine Linux Docker container. `docker run alpine getconf NGROUPS_MAX` gives back the value '32', while `sysctl kernel.ngroups_max` gives back the value '65536' both on the host and inside the container. `docker run ubuntu getconf NGROUPS_MAX` also returns '65536'.
2019-10-29 14:13:08 <DeLord> why is NGROUPS_MAX set to 32 in the Alpine container?
2019-10-29 14:15:39 <_ikke_> DeLord: just out of curiosity, why do you need to have so many groups for a user?
2019-10-29 14:18:44 <DeLord> _ikke_: I have several users on the host (more than 32...) and I need files (with g+rx permissions) from each of them to be accessible inside the container (via bind mount) for a single user. so I just add that user in the container to all the group-ids of the users group on the host
2019-10-29 14:18:55 <DeLord> don't know whether that was well explained, do you get what I am doing? :D
2019-10-29 14:19:56 <_ikke_> DeLord: would acls work for you?
2019-10-29 14:20:51 <_ikke_> although you would need to add acls on all of the files
2019-10-29 14:21:22 <DeLord> I thought about that, but I always end up in situations in which the ACL gets lost for some files, and I add a nightly cronjob to fix them or something... I thought the group-membership would be the cleaner way
2019-10-29 14:21:56 <_ikke_> DeLord: default acls on directories should fix that
2019-10-29 14:23:29 <DeLord> wasn't there also a problem with ACLs and extracing tarballs etc., they overwrite that?
2019-10-29 14:24:00 <_ikke_> probably if they unlink files and create them again
2019-10-29 14:24:37 <DeLord> I have no good experience with acls at least :D
2019-10-29 14:25:05 <_ikke_> In the cases I did use them, they have worked well for me
2019-10-29 14:25:31 <_ikke_> Just use default acls so that they are automatically inheritted
2019-10-29 14:25:39 <DeLord> but however, shouldn't this be independently from my problem, or do you know that the limit is set intentionally to 32 on Alpine?
2019-10-29 14:25:49 <_ikke_> No, I don't :)
2019-10-29 14:26:15 <DeLord> couldn't find something useful regarding this on google either :(
2019-10-29 14:26:49 <DeLord> guess its kind of a unusual thing to do
2019-10-29 14:27:14 <_ikke_> You set a default acl on a directory, and every new file in that directory will inherit that acl
2019-10-29 14:28:37 <_ikke_> seffacl -dm u:<uid>:rx <directory>
2019-10-29 14:30:12 <DeLord> yeah, thanks, I will check whether switching to ACLs will be the better solution, but would be nice to know if the limit of 32 is intended or not in Alpine :)
2019-10-29 14:31:03 <DeLord> all other major distributions have a limit of 65536 when being used as container, thats what I find pretty curious
2019-10-29 14:34:21 <_ikke_> DeLord: Might also have to do with musl, not sure though
2019-10-29 14:35:11 <DeLord> if I dont get an answer here in the channel till later, can you tell me where I could ask that kind of question?
2019-10-29 14:35:35 <_ikke_> yes, it's musl that sets it
2019-10-29 14:35:47 <_ikke_> musl/src/musl-1.1.19/include/limits.h
2019-10-29 14:35:48 <_ikke_> 51:#define NGROUPS_MAX 32
2019-10-29 14:38:22 <_ikke_> DeLord: so youhave to ask #musl then
2019-10-29 14:40:10 <DeLord> uuuuh, nice finding
2019-10-29 14:49:46 <DeLord> _ikke_: oh, I totally forgot. thank you!
2019-10-29 18:21:26 <KH405> Anyone has podman running on alpine?
2019-10-29 18:28:02 <_ikke_> No one has packaged it yet afaik
2019-10-29 20:07:11 <DEac-> podman would be interesting.
2019-10-29 20:57:16 <mckaygerhard1> ncopa _ikke_ now that "new freedesktop "standard": /etc/machine-id" was announced at  https://www.mail-archive.com/dng@lists.dyne.org/msg24687.html how will be the dbus package implementation at alpine edge?
2019-10-29 20:58:24 <_ikke_> mckaygerhard1: please don't constantly highlight us
2019-10-29 20:58:40 <_ikke_> mckaygerhard1: You can just ask questions and see if someone responds
2019-10-29 21:03:07 <AinNero> previously, i generated it manually via dbus-uuidgen
2019-10-29 21:03:18 <AinNero> dbus requires it at /var/lib/dbus/machine-id for some applications
2019-10-29 21:03:35 <_ikke_> https://www.mail-archive.com/dng@lists.dyne.org/msg24707.html
2019-10-29 21:07:26 <AinNero> thinking about it, a cronjob that automatically regenerates it does make sense
2019-10-29 21:07:48 <_ikke_> Or having one id for everyone
2019-10-29 21:08:37 <AinNero> heh, yesss
2019-10-29 21:08:49 <AinNero> pick the md5 of "paprika" please :D
2019-10-29 21:08:53 <_ikke_> hehe
2019-10-29 21:09:11 <_ikke_> de2c9f8839385b3fcf939952de28bb01 it shall be
2019-10-29 21:09:18 <AinNero> 
ACTION dances

2019-10-29 21:09:39 <_ikke_> 
ACTION updates /etc/machine-id

2019-10-29 21:10:47 <_ikke_> who joins?
2019-10-29 21:11:04 <AinNero> me ofc
2019-10-29 21:12:24 <mps> :D
2019-10-29 21:13:48 <_ikke_> 
ACTION wonders if we could sneak it into alpine-baselayout :D

2019-10-29 21:14:14 <AinNero> i'd be fine with it
2019-10-29 21:14:28 <_ikke_> https://www.mail-archive.com/dng@lists.dyne.org/msg24794.html
2019-10-29 21:14:36 <_ikke_> fawatch
2019-10-29 21:18:01 <AinNero> i get fawatch: error returned by fanotify_init(): Function not implemented
2019-10-29 21:18:57 <_ikke_> Are we missing kernel support?
2019-10-29 21:19:14 <AinNero> yes
2019-10-29 21:19:18 <AinNero> # CONFIG_FANOTIFY is not set
2019-10-29 21:19:32 <AinNero> (checked vanilla so far)
2019-10-29 21:19:35 <_ikke_> The fanotify API is available only if the kernel was built with the CONFIG_FANOTIFY configuration option enabled.
2019-10-29 21:19:40 <_ikke_> ah ok
2019-10-29 21:19:42 <_ikke_> sad
2019-10-29 21:20:58 <TBB> has it already been superseded by yet another notification method just like it replaced an earlier one?
2019-10-29 21:23:34 <_ikke_> it works on my non-alpine machine
2019-10-29 21:23:56 <_ikke_> /etc/machine-id open by "cat" (pid=169716, owner="kevin")
2019-10-29 21:41:19 <mps> just upgraded kernel to 5.4-rc5 and I see a loot of differences in .config (created by 'make defconfig') from current config-vanilla.x86_64
2019-10-29 21:41:55 <mps> and some config options in config-vanilla.x86_64 looks somewhat strange
2019-10-29 22:05:10 <mckaygerhard1> AinNero: -> "pick the md5 of "paprika" please" X-D X-D X-D of course only those tht already watching paprika will understand!
2019-10-29 22:05:55 <mckaygerhard1> ok guys the point of that maybe in the future could that uuid of id machine will must be a constant since boot!
2019-10-29 22:06:11 <mckaygerhard1> so the trick of update it randownly could be a problem ..
2019-10-29 22:06:38 <mckaygerhard1> also kernel 5.X now has terrific differences..
2019-10-29 22:07:49 <AinNero> ok you just outed yourself as a weeb
2019-10-29 22:08:16 <AinNero> i just like peppers.
2019-10-29 22:10:09 <AinNero> don't you guys know the green pepper from GIMP?
2019-10-29 22:11:43 <mps> I remember, and wondered why is it there
2019-10-29 22:12:48 <AinNero> a few years ago i ran true.stallman.faith, which essentially was a picture of RMS with a green pepper in a field of green peppers
2019-10-29 22:12:52 <AinNero> i dont really know either.
2019-10-29 22:16:03 <jn__> i've seen the Green Pepper on the over of an old paper magazine about image processing, once
2019-10-29 22:16:12 <jn__> which gave me the idea that maybe it's older than GIMP
2019-10-29 22:20:23 <dalias> ncopa, python seems to have x86_64-linux-gnu as its idea of the host system type for the purpose of C modules, etc. i wonder if this is a mistake
2019-10-29 22:55:58 <mckaygerhard> AinNero:  weeb ? wa the fuck its that?
2019-10-29 22:56:28 <maxice8> mckaygerhard: Someone that likes anime
2019-10-29 22:56:56 <AinNero> level 2 anime nerd
2019-10-29 23:13:33 <iggy> I thought someone was just Hungarian (they sure are proud of their paprika there)
2019-10-29 23:14:04 <maxice8> AinNero: level 2 anime nerd ?
2019-10-29 23:22:02 <dalias> weeb isn't just anime nerd. it's more like anime nerd who thinks being obsessed with anime makes them honorary-japanese
2019-10-29 23:22:36 <maxice8> This should definitively be in -offtopic
2019-10-29 23:23:04 <maxice8> i'm really interested in AinNero's anime level grading
2019-10-29 23:23:15 <dalias> :-)
2019-10-30 00:12:12 <mckaygerhard1> maxice8: "someone that like anime" that are a stupid those people are named "otakus" not "weeb" puff
2019-10-30 00:12:42 <maxice8> -> -offtopic
2019-10-30 00:12:58 <mckaygerhard1> AinNero: level2 anme nerd are not anime like! those are just normal people!
2019-10-30 00:14:51 <mckaygerhard1> i'm a high fan of anime.. but i never hear of a stupid name for those people.. for sure Some frustrated immature teenager sure invented that little name.
2019-10-30 00:15:04 <mckaygerhard1> the corected name are just otaku!
2019-10-30 00:35:28 <mckaygerhard1> hey guys! let me show the black of future:
2019-10-30 00:35:29 <mckaygerhard1> 1) are you haste of pulseaudio and lenny P projects?
2019-10-30 00:35:29 <mckaygerhard1> 2) are you enought of waste of systemshit? alsa get stalled?
2019-10-30 00:35:29 <mckaygerhard1> i'll give you all kind of hope (oh yeah sure->shit), NOW PRESENT https://pipewire.org/
2019-10-30 00:35:29 <mckaygerhard1> 3) are you dissapointed (praid) of the multimedia linux faules!
2019-10-30 00:35:29 <mckaygerhard1> the fusion of the heavyweith of JACK and the faulies of pulseaudio! YEAHHHH!!!!
2019-10-30 02:54:24 <grayhatter> does apk store old versions anywhere?
2019-10-30 02:54:50 <grayhatter> I updated "something" and now my sound card dosent' work
2019-10-30 02:54:55 <dalias> client-side? i dont think so but maybe you can configure it to
2019-10-30 02:54:59 <dalias> :(
2019-10-30 02:55:10 <maxice8> in /var/cache/apk if you setup-apkcache
2019-10-30 02:55:22 <grayhatter> which I didn't
2019-10-30 02:55:26 <dalias> did you reboot after update? it might just be that kernel modules for your running kernel were rm'd
2019-10-30 02:55:28 <grayhatter> logs anywhere?
2019-10-30 02:55:36 <grayhatter> custom kernel
2019-10-30 02:55:38 <dalias> and you need to boot the new one so new version's modules can get loaded
2019-10-30 02:55:39 <dalias> ah
2019-10-30 02:55:52 <dalias> no sound in any apps?
2019-10-30 02:58:09 <grayhatter> bluezalsa works
2019-10-30 02:58:35 <grayhatter> [    2.985860] skl_hda_dsp_generic skl_hda_dsp_generic: Unsupported HDAudio/iDisp configuration found
2019-10-30 02:58:38 <grayhatter> [    2.985863] skl_hda_dsp_generic: probe of skl_hda_dsp_generic failed with error -22
2019-10-30 02:58:40 <grayhatter> this is new
2019-10-30 03:01:51 <dalias> googled it?
2019-10-30 03:02:24 <grayhatter> nothing helpful
2019-10-30 03:03:15 <grayhatter> my guess is that a new sound driver appeared, and it's borked
2019-10-30 03:05:39 <artok> wheeeee
2019-10-30 03:06:22 <artok> rpi4 and alpine aarch64 running
2019-10-30 03:07:09 <grayhatter> I think I've just gotta figure out what I need to blacklist
2019-10-30 03:07:23 <artok> now just trying to do sd boot only and root on usb media
2019-10-30 05:18:52 <GregP> Has anyone been able to Alpine to boot and run with USB support on a Raspberry Pi 4?
2019-10-30 06:17:04 <slap_happy> it would be great if we could get this on Alpine
2019-10-30 06:17:05 <slap_happy> https://www.reddit.com/r/PFSENSE/comments/douvuz/speedtestnet_just_launched_an_official_c_cli/
2019-10-30 06:20:07 <slap_happy> I have been using this old python one but from the sounds of it this official C++ one will perform much better
2019-10-30 07:00:50 <fredrigu> slap_happy: that sounds awesome! I really like to need to start a browser just for a speed test. You do know about https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package ?
2019-10-30 07:09:53 <mps> slap_happy: I have long time some perl cli script and module to speed test apk
2019-10-30 07:09:54 <slap_happy> Very cool.  I'm actually a Linux newb setting up my first Linux server with Alpine dockers this week.  Hopefully I'll get savvy enough to build packages no prob.
2019-10-30 07:11:35 <fredrigu> slap_happy: welcome to the community :)
2019-10-30 07:12:42 <slap_happy> Thx.  Stellar distro.  I'm stoked how many dockers I can run with it.
2019-10-30 07:13:37 <slap_happy> Headed off to bed.  Cheers all.
2019-10-30 07:14:06 <ncopa> morning
2019-10-30 07:14:31 <ncopa> dalias: can you please join #alpine-devel for the python x86_64-linux-gnu thing
2019-10-30 07:15:27 <ncopa> i am rebuilding all python stuff right now, almost done...
2019-10-30 07:15:35 <ncopa> for python 3.8
2019-10-30 10:51:42 <artok> whee https://hastebin.com/abacumukon.bash
2019-10-30 10:51:57 <artok> rpi4 up and running
2019-10-30 11:02:02 <misterunknown> Hi all. I just recieved a message from the mailing list, that I should remove my "application/x-pkcs7-signature attachments" in order to be able to send mails to this list. I really consider this a bug.
2019-10-30 11:07:42 <misterunknown> I guess at least signatures should be whitelisted, as they are an important criterion of email authenticity.
2019-10-30 11:20:50 <qa3Txu0iak0F> who has your pkcs7 pubkey to actually check your sig, and which clients do this correctly anyway?
2019-10-30 11:27:16 <misterunknown> qa3Txu0iak0F: I use an S/MIME certificate from sectigo. I know that both GMail and mutt verify this signature correctly.
2019-10-30 11:28:46 <misterunknown> Even if there are clients which can't verify this signature, banning signatures is insane. Really.
2019-10-30 11:31:06 <misterunknown> It's hard enough for a private mail server admin to send mails to the big mail providers.
2019-10-30 11:38:00 <qa3Txu0iak0F> hmmm. `fgrep -ri sectigo /etc/ssl/` gives me zero results.
2019-10-30 11:41:09 <qa3Txu0iak0F> btw > both GMail and mutt verify this signature correctly. is the positive case, which is easy to test, the case that is impossible to test is if mutt/google also identify well done fake sigs immitations correctly?
2019-10-30 11:41:27 <qa3Txu0iak0F> see also efail
2019-10-30 11:43:22 <qa3Txu0iak0F> anyway, i'm off-topic sorry.
2019-10-30 11:47:49 <KatolaZ> qa3Txu0iak0F: I guess the burden of proving identity should not me on the ML manager, rather on the recipient
2019-10-30 11:48:23 <KatolaZ> s/ me / be /
2019-10-30 12:12:31 <MY-R> https://gitlab.com/gitlab-com/www-gitlab-com/issues/5672
2019-10-30 12:12:43 <MY-R> ^ read with popcorn :D
2019-10-30 12:12:51 <AinNero> [angry NPC face]
2019-10-30 12:13:30 <AinNero> when i got the first email i realized that i have a gitlab account and recovered access to it
2019-10-30 12:13:38 <AinNero> and enabled mirroring of my repos
2019-10-30 12:13:44 <MY-R> :P
2019-10-30 12:14:38 <artok> heh
2019-10-30 12:14:52 <artok> but the email last night
2019-10-30 12:15:15 <artok> "sorry"
2019-10-30 12:15:44 <MY-R> more like "sorry but we will do it anyway"
2019-10-30 12:17:00 <AinNero> i dont care too much
2019-10-30 12:17:14 <AinNero> after i lost trust in github i started pushing to my own server
2019-10-30 12:17:15 <artok> at least they make it proposal
2019-10-30 12:17:32 <AinNero> and now my own server is mirroring it to github, gitlab, or whatever is currently hip
2019-10-30 12:19:58 <AinNero> dunno. you feel much less rage about a company misbehaving if you dont depend on them.
2019-10-30 13:12:18 <ncopa> misterunknown: can you report that issue on https://gitlab.alpinelinux.org/alpine/infra/infra
2019-10-30 13:21:46 <misterunknown> ncopa: No problem :)
2019-10-30 14:08:49 <misterunknown> qa3Txu0iak0F: I absolutely don't get your point. I really don't care if *you* can verify my signature, as long as GMail can. And this is a real-world problem, as there are at least some gmail users subscribed.
2019-10-30 14:09:32 <misterunknown> qa3Txu0iak0F: BTW. you won't find any results using fgrep on /etc/ssl... Apart from that sectigo uses an intermediate cert, signed by USERTrust, which is part of the ca-certificates.
2019-10-30 14:33:56 <p3rL> how to add hwloc-dev
2019-10-30 14:34:02 <p3rL> i am using 3.9 Alpine
2019-10-30 14:34:20 <_ikke_> 3.9 does not contain it
2019-10-30 14:34:33 <_ikke_> It's still in testing
2019-10-30 14:34:38 <p3rL> then which version ?
2019-10-30 14:34:38 <_ikke_> so no stable branch contains it
2019-10-30 14:34:58 <_ikke_> 3.11 would be the first if someone moves it out of testing
2019-10-30 14:35:58 <_ikke_> https://git.alpinelinux.org/aports/tree/testing/hwloc/APKBUILD#n1
2019-10-30 14:36:00 <_ikke_> That's the maintainer
2019-10-30 14:51:16 <p3rL> i just installed it manually lets see
2019-10-30 14:51:19 <p3rL> its work fine or not
2019-10-30 14:59:32 <mps> Cogitri: does NM work with latest iwd
2019-10-30 15:00:01 <mps> 0.23
2019-10-30 16:09:25 <dalias> does apk info -r only consider installed packages, not all packages known about?
2019-10-30 16:09:40 <dalias> that's what it seems to do
2019-10-30 16:10:00 <dalias> if so, is there any way to get a list of all reverse deps in available package db ?
2019-10-30 16:13:24 <Cogitri> mps: I don't think it breaks but haven't tested
2019-10-30 16:16:31 <mps> ok, thanks, there is discussion on #iwd and someone told that the iwd doesn't work with openrc and NM
2019-10-30 16:47:42 <Greg43198702> Does anyone know if the PI 4 boot and USB issue has been addressed with any recent commits, and if there are, if they require any compatibility testing on pi hardware?
2019-10-30 16:54:34 <_ikke_> Greg43198702: artok_ has been working on running on RPI4
2019-10-30 16:57:41 <Greg43198702> Tnx _ikke_
2019-10-30 16:59:22 <mckaygerhard1> xfce4 are now gtk3+since 1.14...
2019-10-30 16:59:36 <mckaygerhard1> and gtk4 are started to use
2019-10-30 17:10:02 <Cogitri> gtk3 and gtk4 are pretty similar though, so it's not like the effort to move to gtk3 has been wasted
2019-10-30 17:59:28 <mps> Cogitri: sorry again to bother you, iwd developer ask how long you tested iwd with NM
2019-10-30 18:00:41 <Cogitri> For a few weeks, until I had to switch back to wpa_supplicant because uni uses WPA2 enterprise (but I think by now I could switch back to iwd)
2019-10-30 18:01:28 <mps> thanks, looks like iwd 1.0 is problem because of some path changes
2019-10-30 18:02:32 <Cogitri> Oh, okie
2019-10-30 18:03:42 <mps> so, NM will not work with 1.0 till it fixed
2019-10-30 18:05:58 <mps> good news is that NM will need one or few lines change
2019-10-30 18:13:53 <grayhatter> https://wiki.alpinelinux.org/wiki/Alpine_Linux:FAQ#My_cron_jobs_don.27t_run.3F
2019-10-30 18:14:06 <grayhatter> why does the . char stop scripts from running?
2019-10-30 18:14:08 <Cogitri> Will check with NM upstream
2019-10-30 18:14:19 <grayhatter> i.e., who's bug is this?
2019-10-30 18:14:59 <mps> wait till we push iwd 1.0
2019-10-30 18:15:28 <mps> to test, I mean. ofc you can ask right now of the change
2019-10-30 18:16:20 <mps> and I think balrog-k1n (NM developer) is on #iwd, so he probably know already
2019-10-30 18:16:25 <Cogitri> Yup :)
2019-10-30 18:17:00 <mps> I remember he helped me with NM upgrade on alpine some time ago
2019-10-30 19:37:44 <hooogs> I am having issues with mounting a fat32 filesystem in a VM. I have also noticed this behavior inconsistently on bare metal as well.
2019-10-30 19:38:10 <hooogs> I have created a FAT32 filesystem to store the APKOVL. I am booting off the official x86 standard ISO.
2019-10-30 19:39:02 <hooogs> The FAT32 filesystem will not mount, it shows this error: invalid argument.
2019-10-30 19:39:13 <artok> dosfstools installed ?
2019-10-30 19:40:18 <hooogs> No, they are not. But they shouldn't have to be to mount a fat partition, should they? Otherwise, how would Alpine be able to mount the fat32 partition to read the APKOVL during the boot process?
2019-10-30 19:40:35 <hooogs> Which I know it does.
2019-10-30 19:41:06 <hooogs> This isn't anything fancy, it is "mount /dev/sda1 /mnt"
2019-10-30 19:42:16 <artok> try installing and run the same mount command =)
2019-10-30 19:43:55 <hooogs> Nope. Same error.
2019-10-30 19:44:00 <hooogs> It isn't the lack of dosstools.
2019-10-30 19:44:06 <hooogs> dosfstools*
2019-10-30 19:44:19 <artok> but I suggest having that, there is fsck for fat
2019-10-30 19:44:35 <artok> but nevertheless, what does kernel say
2019-10-30 19:44:40 <artok> after mount command
2019-10-30 19:44:43 <artok> dmesg
2019-10-30 19:46:06 <hooogs> https://imgur.com/Ikn5f81
2019-10-30 19:46:18 <hooogs> Sorry, no VM tools installed. Copy and paste is a pain )
2019-10-30 19:46:21 <hooogs> :)*
2019-10-30 19:46:47 <hooogs> So, does it just not like the FS label?
2019-10-30 19:47:23 <hooogs> I created this in parted, so maybe it mixed letter cases when naming the partition or something, and now it won't mount because it expects FAT32 to be all caps?
2019-10-30 19:48:02 <artok> shouldn't be, as I've got many times Boot or something else as label
2019-10-30 19:48:22 <artok> it just finds some utf8 char there
2019-10-30 19:49:18 <dalias> does dmesg show anything ?
2019-10-30 19:49:50 <hooogs> See imgur link for dmesg output.
2019-10-30 19:49:56 <dalias> that's no problem
2019-10-30 19:50:20 <dalias> try strace on the mount command and make sure it's actually the mount syscall that's failing
2019-10-30 19:50:23 <dalias> and not something idiotic
2019-10-30 19:50:38 <dalias> also, do you have -t vfat on the mount command line ?
2019-10-30 19:50:52 <dalias> it's possible autodetect is failing for some reason
2019-10-30 19:50:54 <hooogs> Not for that output, but I have tried that as well.
2019-10-30 19:51:16 <dalias> or that /dev/sda1 is in your /etc/fstab and thus it's using settings from there rather than autodetecting
2019-10-30 19:52:56 <hooogs> Nope. -t vfat has the same error
2019-10-30 19:53:02 <hooogs> Same thing in dmesg as well.
2019-10-30 19:53:41 <hooogs> Nothing in fstab besides the standard usbdisk and cdrom entries.
2019-10-30 19:55:11 <dalias> strace?
2019-10-30 19:56:53 <hooogs> I am trying to get it into pastebin
2019-10-30 20:00:24 <dalias> apk add sprunge, then strace ... 2>&1 | sprunge
2019-10-30 20:00:55 <dalias> and never have to think about these awful pastebins that actually require copy&paste again
2019-10-30 20:02:16 <hooogs> http://sprunge.us/h4s2yz
2019-10-30 20:02:35 <hooogs> Yeah, stackoverflow suggested as much.
2019-10-30 20:03:25 <Cogitri> 
ACTION suggests wgetpaste to paste to loads of services

2019-10-30 20:03:26 <dalias> ok, are you sure the fat32 fs is actually on /dev/sda1 ?
2019-10-30 20:04:13 <dalias> have you checked this from another system, and comparing /proc/partitions with the expected size to make sure sda1 is actually what you think it is?
2019-10-30 20:05:09 <hooogs> http://sprunge.us/gCHR0K
2019-10-30 20:05:14 <dalias> i think this EINVAL from the syscall means the kernel thinks there's no valid fat fs on that device
2019-10-30 20:05:27 <dalias> could it actually be exfat or something awful?
2019-10-30 20:05:30 <hooogs> Well, the way I am doing this is with a VDI disk as sda
2019-10-30 20:05:37 <hooogs> This is in virtualbox.
2019-10-30 20:06:03 <hooogs> But I have seen the same behavior with multiple FS types on a USB drive as well when booted into Alpine on baremetal.
2019-10-30 20:06:09 <dalias> try something like
2019-10-30 20:06:16 <hooogs> At some point, the problem magically went away on baremetal.
2019-10-30 20:06:36 <dalias> dd if=/dev/sda1 of=imagehead bs=1024 count=64 ; file imagehead
2019-10-30 20:06:41 <hooogs> I have a Ubuntu 16.04 VM I can boot it into, I will mount the VDI there and check it in gparted.
2019-10-30 20:06:43 <dalias> and see if file(1) reports that it's a fat filesystem
2019-10-30 20:06:59 <hooogs> Ok.
2019-10-30 20:07:52 <hooogs> It reports it as data.
2019-10-30 20:08:20 <hooogs> Same thing with a BS of 4096.
2019-10-30 20:08:22 <artok> what code did you give it?
2019-10-30 20:08:30 <artok> partition code, that is
2019-10-30 20:09:01 <hooogs> Do you mean the partition number, or the code for that partition or FS type?
2019-10-30 20:09:08 <artok> 0c ?
2019-10-30 20:09:17 <artok> code
2019-10-30 20:09:37 <artok> partition number seems to be 1 =)
2019-10-30 20:10:03 <jailbox> 2nd paste: id=c type=Win95 FAT32 (LBA)
2019-10-30 20:10:12 <hooogs> ID = c.
2019-10-30 20:10:22 <hooogs> See second paste, that is the fdisk -l output.
2019-10-30 20:10:25 <artok> oh missed =)
2019-10-30 20:10:27 <artok> sorry..
2019-10-30 20:10:39 <artok> which alpine version ?
2019-10-30 20:10:56 <hooogs> Latest.
2019-10-30 20:10:59 <hooogs> 3.10
2019-10-30 20:11:23 <hooogs> 3.10.3 Standard x86 to be exact.
2019-10-30 20:11:31 <dalias> the partition type code is meaningless to linux
2019-10-30 20:11:38 <dalias> only the contents of the partition matter
2019-10-30 20:11:59 <dalias> if file is saying "data", there's not a fatfs boot record on the partition
2019-10-30 20:12:25 <dalias> either it's been corrupted or the partition table is wrong about where it is or something like that
2019-10-30 20:12:40 <artok> dosfstools and fsck
2019-10-30 20:12:59 <dalias> you could hexdump imagehead
2019-10-30 20:13:15 <dalias> and see if it looks meaningful for guessing what happened
2019-10-30 20:13:16 <hooogs> Ok, so this is some kind of error during the initial creation of the partition then... Actually, that makes sense. When this went away on baremetal was when I created the config partition from another OS in gparted.
2019-10-30 20:13:45 <hooogs> So, parted is screwing it up in some capacity.
2019-10-30 20:14:47 <hooogs> Let me try creating the filesystem with mkfs.vfat instead...
2019-10-30 20:15:03 <hooogs> quit
2019-10-30 20:15:10 <hooogs> Sorry, wrong window.
2019-10-30 20:22:47 <hooogs> http://sprunge.us/C09Atc
2019-10-30 20:23:12 <hooogs> That is the output from the dd | file series of commands.
2019-10-30 20:23:33 <hooogs> This time, I went and created the partitions in parted, and used mkfs.vfat -f32 for the FS.
2019-10-30 20:24:13 <hooogs> File is saying it is a valid MBR/FAT32 FS, but when I try to mount it I am still getting the same error.
2019-10-30 20:24:47 <hooogs> Nothing in dmesg this time though.
2019-10-30 20:25:36 <artok> um.
2019-10-30 20:25:38 <artok> you have img ?
2019-10-30 20:26:15 <artok> losetup /dev/loop0 image.img
2019-10-30 20:26:19 <hooogs> Here is the strace of the mount command now, but it looks the same to me: http://sprunge.us/fOS7FB
2019-10-30 20:27:13 <hooogs> "you have img ?"
2019-10-30 20:27:17 <hooogs> Not sure I follow.
2019-10-30 20:27:26 <hooogs> What are you talking about?
2019-10-30 20:27:28 <artok> file image, not mounted on virtualbox?
2019-10-30 20:28:09 <hooogs> Yes, I dd'ed sda1 to a file, bs 4096 count 64.
2019-10-30 20:28:19 <hooogs> Then I ran the file command on it.
2019-10-30 20:28:42 <hooogs> This is all within the VM mind you.
2019-10-30 20:29:56 <hooogs> I am gonna mount this VDI in a Ubuntu VM, create the partition in gparted,and see if this is still an issue.
2019-10-30 20:36:33 <MY-R> hooogs, maybe try just witf cfdisk and create new partition table: cfdisk -z /dev/sda and after on sda1 your fat filesystem
2019-10-30 20:47:59 <hooogs> I did create a new partition table with parted.
2019-10-30 20:48:19 <hooogs> I mean, I could do it with cfdisk too, if parted could potentially be the issue...
2019-10-30 21:04:29 <hooogs> Ok, I am back.
2019-10-30 21:05:04 <hooogs> Creating the FAT32 partition in gparted in another OS and then trying to mount it in Alpine does not solve the problem.
2019-10-30 21:06:06 <mps> hooogs: did you made FS on partition
2019-10-30 21:06:20 <hooogs> Yes.
2019-10-30 21:06:42 <mps> what kernel version you use
2019-10-30 21:07:16 <hooogs> I started gparted on a Ubuntu VM with the VDI connected to it, recreated the partition table, and made a FAT32 partition named CONFIG.
2019-10-30 21:07:39 <mps> and what 'cat /proc/filesystems' says about fat fs
2019-10-30 21:07:40 <hooogs> I don't know the kernel version, VM is off. Whatever shipped with 16.04.
2019-10-30 21:07:57 <mps> did you mkfs.vfat for that partition
2019-10-30 21:08:25 <mps> I mean alpine kernel version
2019-10-30 21:08:35 <MY-R> ye wanted ask same cus have no idea what gparted mean by FAT32
2019-10-30 21:09:27 <mps> ime, you can mark partition however you want but you have to create proper FS on it
2019-10-30 21:09:35 <hooogs> 4.19.80
2019-10-30 21:09:55 <hooogs> @mps Yes, I know.
2019-10-30 21:10:11 <mps> hooogs: ok, what ' cat /proc/filesystems | grep fat' says
2019-10-30 21:10:26 <MY-R> hooogs, so "modprobe vfat" on Alpine and cant mount this partition?
2019-10-30 21:10:40 <hooogs> There is no fat entry in /proc/filesystems
2019-10-30 21:10:58 <mps> then what MY-R says
2019-10-30 21:11:04 <_ikke_> Did you do a system upgrade without rebooting?
2019-10-30 21:11:07 <hooogs> Modprobe vfat doesn't show anything either...
2019-10-30 21:11:16 <_ikke_> (just a random guess without context)
2019-10-30 21:11:17 <hooogs> _ikke_ No.
2019-10-30 21:11:34 <mps> 'modinfo vfat' ?
2019-10-30 21:11:44 <MY-R> hooogs, good, that how it looks, now try mount partition
2019-10-30 21:11:48 <hooogs> This is just a stock Alpine standard 3.10.3 x86 image that is booted in virtualbox.
2019-10-30 21:12:29 <_ikke_> so no virtual image or something like that?
2019-10-30 21:12:47 <hooogs> @MY-R Now it works.
2019-10-30 21:13:05 <MY-R> so enjoy :)
2019-10-30 21:13:17 <hooogs> So, there is fat support, but modprobe loads the kernel module for it?
2019-10-30 21:13:26 <hooogs> Why is this not loaded by default?
2019-10-30 21:13:56 <mps> because it is not needed by default
2019-10-30 21:14:12 <MY-R> you can include load module in mount command but dont remember the flag
2019-10-30 21:14:29 <MY-R> -t vfat or -m vfat?
2019-10-30 21:14:43 <hooogs> I tried -t vfat without success earlier.
2019-10-30 21:14:53 <mps> hooogs: if you want it loaded by default put it in /etc/modules
2019-10-30 21:15:12 <_ikke_> That shouldn't be necessary normally though
2019-10-30 21:16:47 <MY-R> _ikke_, loading module?
2019-10-30 21:17:02 <_ikke_> yes
2019-10-30 21:17:05 <MY-R> it is
2019-10-30 21:17:20 <MY-R> many times I forgot about it and had to load module before mount anything
2019-10-30 21:17:25 <MY-R> but -t vfat was doing it for me
2019-10-30 21:17:28 <mps> mount doesn't autoload modules
2019-10-30 21:18:15 <hooogs> _ikke_ I was in here about a week ago asking about something similar. I think under the username uderv. I know for a fact that Alpine can load the APKOVL from a fat32 partition on boot, because you told me.
2019-10-30 21:18:19 <_ikke_> /proc/filesystems includes vfat for me, but no relevan module has been loaded
2019-10-30 21:18:41 <hooogs> It also seemed to automatically mount that partition on boot.
2019-10-30 21:18:48 <hooogs> So, this is really weird behavior.
2019-10-30 21:19:08 <MY-R> I got liek that only if using Alpine installation media
2019-10-30 21:19:29 <hooogs> Because if it can automatically mount the fat32 partition on boot, then it must load the vfat module on boot.
2019-10-30 21:19:45 <_ikke_> this is my /proc/filesystems; https://tpaste.us/kyqb
2019-10-30 21:19:49 <MY-R> hooogs, ye after installation is fine :)
2019-10-30 21:20:04 <_ikke_> lsmod: http://tpaste.us/Obz6
2019-10-30 21:20:17 <hooogs> MY-R I am not installing. I am running from RAM, and loading the config on boot. I am not installing to disk.
2019-10-30 21:20:28 <MY-R> ah
2019-10-30 21:20:37 <MY-R> then weird
2019-10-30 21:20:41 <mps> _ikke_: maybe you have it modules option in syslinux or grub
2019-10-30 21:21:03 <_ikke_> shouldn't it show up in lsmod then?
2019-10-30 21:21:15 <hooogs> I am not seeing any of vfat, or the ext's in /proc/filesystems on mine.
2019-10-30 21:21:37 <_ikke_> let me try to boot a new vm
2019-10-30 21:21:56 <MY-R> well, grub adding entry with rootfstype=
2019-10-30 21:21:59 <mps> _ikke_: linux-vanilla ?
2019-10-30 21:22:35 <hooogs> @MY-R I shouldn't HAVE to make any config changes here. I am running Alpine standard with no modifications, and I know this works on another system.
2019-10-30 21:23:08 <hooogs> I shouldn't have to make changes to the boot parameters or GRUB options for this either.
2019-10-30 21:23:20 <_ikke_> mps: i guess a custom kernel
2019-10-30 21:23:26 <_ikke_> mps: it's a vps
2019-10-30 21:24:22 <mps> ah, that explains
2019-10-30 21:24:34 <MY-R> hooogs, I mean that after grub detect everything then it automaticly add that entry on which you got root partition, so should work without changing anything
2019-10-30 21:24:53 <mps> MY-R: hm, true, 'mount -t vfat ....' autoloads module
2019-10-30 21:25:00 <MY-R> :>
2019-10-30 21:25:41 <MY-R> weird that didnt work before for hooogs but I think now should work fine after he used gparted again I guess
2019-10-30 21:25:48 <_ikke_> still my experience is that I usually don't have to load any kernel modules
2019-10-30 21:25:59 <mps> first time I noticed that, but I'm sure in some of previous versions it didn't worked because I remember that I had to load module manually
2019-10-30 21:26:16 <hooogs> MY-R It is possible that this is some VirtualBox weirdness. Like, it doesn't make the FAT32 volume available to Alpine until later in the boot process or something...
2019-10-30 21:26:51 <MY-R> I always use -t when installing Alpine from Alpine image/iso
2019-10-30 21:26:57 <mps> 
ACTION everyday learn something new

2019-10-30 21:28:33 <hooogs> I didn't install it. I am BOOTED from the ISO. This is literally just live booted from the ISO.
2019-10-30 21:28:52 <hooogs> And I have tried the -t option with mount. I have tried -t auto and -t vfat, neither of which work.
2019-10-30 21:28:56 <MY-R> hooogs, so in that case ye need use -t like I'm using when boot Alpine media :)
2019-10-30 21:29:14 <MY-R> try now again :P
2019-10-30 21:30:01 <hooogs> The only way the -t option works is when I load the kernel module first, which I shouldn't have to do.
2019-10-30 21:30:20 <mps> maybe busybox mount doesn't autoload module
2019-10-30 21:30:39 <MY-R> oh that could be too since Im installing coreutils before...
2019-10-30 21:30:39 <_ikke_> ok, yea, -t vfat is required
2019-10-30 21:30:51 <MY-R> I elarnt from this wiki: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS
2019-10-30 21:30:53 <MY-R> about -t :P
2019-10-30 21:31:31 <hooogs> I don't think this is standard behavior, _ikke_ remarked that he never had to load modules. I have never had to do this up until now either.
2019-10-30 21:31:45 <hooogs> MY-R: Yes, I know how the -t option works.
2019-10-30 21:31:57 <MY-R> and I always had to do it :)
2019-10-30 21:32:02 <_ikke_> but after I run it once with -t, the next time it works without
2019-10-30 21:32:09 <MY-R> ye cus module is loaded
2019-10-30 21:32:14 <_ikke_> what module?
2019-10-30 21:32:17 <MY-R> vfat
2019-10-30 21:32:24 <_ikke_> lsmod doesn't list it for me
2019-10-30 21:32:33 <mps> yes, coretuils mount autoload module while busybox applet doesn't
2019-10-30 21:32:39 <_ikke_> sorry, it does
2019-10-30 21:33:11 <mps> _ikke_: sorry, you are right :)
2019-10-30 21:33:19 <mps> I messed parameter
2019-10-30 21:33:47 <_ikke_> ok, now that's cleared up
2019-10-30 21:33:52 <MY-R> :)
2019-10-30 21:34:37 <MY-R> I never got problem with it, thought it have to be like that, just Alpine way to not load all modules to save ram space
2019-10-30 21:34:55 <_ikke_> it's a modloo
2019-10-30 21:34:57 <_ikke_> modloop
2019-10-30 21:35:10 <_ikke_> though, laoding modules still costs ram I guess
2019-10-30 21:42:48 <hooogs> So, summary of this so far:
2019-10-30 21:43:32 <hooogs> 1. If I create the partition outside of Alpine, (i.e. on Ubuntu), then I can mount the partition with -t vfat.
2019-10-30 21:43:58 <hooogs> 2. I can't just create this partition within Alpine, something goes wrong and it doesn't let me mount it period.
2019-10-30 21:44:46 <mps> hooogs: re 2. I can
2019-10-30 21:45:03 <hooogs> Ok, how are you doing it? What utility?
2019-10-30 21:45:16 <hooogs> I am using parted and mkfs.vfat, and I can't mount it.
2019-10-30 21:45:25 <hooogs> But when I do it in GParted, I can.
2019-10-30 21:45:57 <mps> fdisk or parted
2019-10-30 21:46:31 <hooogs> Ok, that is strange.
2019-10-30 21:47:06 <hooogs> Either way: I have a workaround, therefore I don't care why Alpine screws the partitioning up. I will just use Ubuntu for creating that fat32 volume.
2019-10-30 21:47:30 <hooogs> The other issue: why is it not loading the config from the APKOVL on that partition at boot?
2019-10-30 21:48:20 <hooogs> I mount the partition under /mnt, I run "lbu pkg /mnt" which creates the APKOVL tarball, and then I reboot.
2019-10-30 21:48:32 <hooogs> In theory, it should be loading the stored config from that tarball.
2019-10-30 21:48:38 <hooogs> But it doesn't.
2019-10-30 22:02:45 <hooogs> I know for a fact that this works on baremetal. I just went and replicated this.
2019-10-30 22:02:50 <hooogs> I am using the exact same ISO.
2019-10-30 22:03:14 <hooogs> lsmod | grep vfat shows that the vfat module is loaded on boot.
2019-10-30 22:03:40 <hooogs> Why is it that some modules load on boot when Alpine is running one bare metal, but not when it is running in VirtualBox?
2019-10-30 22:04:13 <hooogs> I can get the Alpine install to mount a fat32 partition on a USB drive on bare metal without the -t option.
2019-10-30 22:04:35 <hooogs> I can get it to load the APKOVL config from that partition on boot, like it should.
2019-10-30 22:04:46 <hooogs> Neither of these things work in virtualbox.
2019-10-30 22:06:10 <hooogs> Maybe this is some kind of hardware detection thing. If it is booted in a VM, maybe the hypervisor blocks Alpine from seeing the FAT32 partition until later in the boot process. If Alpine selectively loads kernel modules based on which devices it can see...
2019-10-30 22:07:28 <hooogs> Nope. Booting it without the FAT32 USB plugged in, still loads the vfat module.
2019-10-30 22:08:20 <hooogs> But it doesn't in virtualbox.
2019-10-30 22:08:37 <MY-R> maybe try with alpine-virt image
2019-10-30 22:08:39 <hooogs> Both of these are stock, running from ram Alpine ISOs.
2019-10-30 22:09:09 <hooogs> MY-R: Good point.
2019-10-30 22:09:33 <hooogs> I did not realize that existed until just now.
2019-10-30 22:09:37 <hooogs> (facepalm)
2019-10-30 22:10:59 <hooogs> Hmm... The alpine-virt image doesn't load vfat either...
2019-10-30 22:19:32 <MY-R> hooogs, https://gitlab.alpinelinux.org/alpine/aports/issues/2269
2019-10-30 22:20:19 <MY-R> hooogs, add to kernel cmd: usbdelay=5
2019-10-30 22:20:38 <hooogs> No, that isn't the issue.
2019-10-30 22:20:43 <hooogs> I am not booting from a CD.
2019-10-30 22:20:51 <hooogs> I did figure it out though.
2019-10-30 22:20:58 <hooogs> The virt image solved the problem.
2019-10-30 22:21:30 <hooogs> It seems like Alpine doesn't load the vfat module unless you boot with some kind of secondary storage device attached.
2019-10-30 22:21:44 <ptx0> howdy, any way i can modify the initramfs trigger to use dracut that i've built?
2019-10-30 22:21:49 <hooogs> So, I added the VDI to the sata controller in virtualbox, and booted it up.
2019-10-30 22:21:55 <MY-R> and iso isnt a CD in virtualbox?
2019-10-30 22:22:21 <hooogs> I guess technically it is, but in this case, I am not having that issue. I solved my problem.
2019-10-30 22:22:42 <hooogs> Anyway, once I added the VDI to the sata controller again, I booted it up. It loaded vfat correctly.
2019-10-30 22:23:11 <hooogs> I have done a run-from-RAM install with the virt image and the APKOVL on the fat32 partition as well.
2019-10-30 22:23:18 <hooogs> It loads the config on boot like it should.
2019-10-30 22:23:43 <hooogs> So, TL;DR: use the virt image.
2019-10-30 22:24:05 <MY-R> so there is no issue with Alpine at all, enjoy!
2019-10-30 22:24:19 <hooogs> I am going to do parted + mkfs from the virt image just to make sure that it works correctly.
2019-10-30 22:24:33 <hooogs> But yeah, it looks like this was a severe case of overthinking the problem :)
2019-10-30 22:24:52 <hooogs> Thanks for all of your help.
2019-10-30 22:25:15 <MY-R> np, have fun with it!
2019-10-30 23:42:18 <velix> Hi there. Just because I'm curious, isn't there "proj" in v3.10.3 ?
2019-10-30 23:44:12 <velix> seems like it is in edge.testing only
2019-10-31 00:08:44 <iggy> velix: only in testing (which implies edge)
2019-10-31 00:08:54 <velix> I see. Thanks.
2019-10-31 00:12:53 <ptx0> any hint toward how alpine's update triggers work would be awesome
2019-10-31 00:13:22 <ptx0> i have been scouring the web but documentation is nil and the source code is mystical
2019-10-31 07:23:25 <xvilka> hi! I am working in a Docker container for setting up the environment for creating Alpine packages. I ran `addgroup -g 1000 -S user && adduser -u 1000 -D -S me -G user && addgroup me abuild`. But if I build and run the container, and try to run `abuild -r` it complains that user is not in the `abuild` group. Indeed, if I ran `groups` from the user-logged shell it shows only `user`, but if I do `sudo
2019-10-31 07:23:31 <xvilka> groups me` it shows the `me: user abuild`. WTF?
2019-10-31 07:23:48 <xvilka> see the dockerfile in question here: https://github.com/XVilka/alpine-dev-env
2019-10-31 07:24:30 <xvilka> and it was working a few days before
2019-10-31 07:24:39 <xvilka> using `FROM alpine:edge`
2019-10-31 07:31:11 <Cogitri> If I were you I'd just use docker-abuild which does what you want and already works
2019-10-31 07:35:06 <xvilka> I love custom things
2019-10-31 07:35:11 <xvilka> but thanks for the pointer
2019-10-31 07:44:01 <xvilka> interestingly, seems happened after the switch from docker to podman
2019-10-31 07:44:08 <xvilka> looks like podman is crap
2019-10-31 07:55:58 <xvilka> so, can you trigger the rebuild of ocaml packages?
2019-10-31 07:56:08 <xvilka> it's such an annoyance it's broken
2019-10-31 08:01:53 <Cogitri> Well, anyone can make the PR for that :^)
2019-10-31 08:03:58 <mps> or create issue on gitlab.a.o to explain why the pkg need to be rebuilt
2019-10-31 08:04:33 <Cogitri> There's already an issue for that
2019-10-31 08:04:39 <xvilka> Cogitri: I would do that, but I don't understand how to trigger
2019-10-31 08:05:06 <xvilka> should just bump a revision?
2019-10-31 08:05:11 <Cogitri> And a PR to rebuild ocaml but that's not ready for merge since it doesn't rebuild the dependants of ocaml
2019-10-31 08:05:35 <Cogitri> Yes, bump the pkgrel of the dependants and make sure they still build
2019-10-31 08:06:18 <xvilka> Cogitri: no, I wan't to fix ocaml-4.08.0 first, it's fine as is, just need to be rebuilt
2019-10-31 08:07:34 <xvilka> *want
2019-10-31 08:10:53 <wildefyr> quick question, does anyone know if plex is supported on alpine? or if not another equiv. media server. thanks
2019-10-31 08:16:25 <xvilka> ah, I know the problem
2019-10-31 08:31:44 <xvilka> it's failure of podman/buildah indeed
2019-10-31 08:32:07 <xvilka> they fail to substitute ARG / ${username} in some cases
2019-10-31 08:43:16 <xvilka> I updated the pull request
2019-10-31 08:43:47 <xvilka> please check https://github.com/alpinelinux/aports/pull/12008
2019-10-31 08:46:17 <xvilka> argh, it's so annoying
2019-10-31 08:47:19 <xvilka> bug in older version of ocaml, so can't use it, bug in the newer alpine release, so can't use alpine, bug in the podman so can't use it and have to switch to the cgroups-v1 in this Fedora linux host.
2019-10-31 08:52:52 <xvilka> everything I try is broken
2019-10-31 09:21:10 <massctrl> Is there an api for https://pkgs.alpinelinux.org or some query param which returns json or similar ?
2019-10-31 09:28:42 <_ikke_> massctrl: no, not atm
2019-10-31 09:31:17 <massctrl> _ikke_: tnx
2019-10-31 09:31:57 <_ikke_> https://github.com/clandmeter/aports-turbo is the code behind it. If you feel like building something to expose an API, feel free
2019-10-31 09:33:38 <clandmeter> https://github.com/alpinelinux/aports-turbo
2019-10-31 09:34:05 <_ikke_> uh, right
2019-10-31 09:34:12 <_ikke_> ddg found that one first
2019-10-31 09:52:00 <xvilka> opened one more issue regarding `abuild` group, see https://gitlab.alpinelinux.org/alpine/aports/issues/10925
2019-10-31 10:30:42 <qa3Txu0iak0F> what about apk-file is that not a kinda-replacement for pkgs.a.o?
2019-10-31 10:35:38 <clandmeter> It uses pkgs
2019-10-31 10:36:45 <qa3Txu0iak0F> oh
2019-10-31 11:50:48 <Cogitri> wildefyr: It needs glibc so no, you'll have to run it in docker or on a chroot
2019-10-31 12:57:35 <Leo_> Hey, i have a server with one network interface (eth0) and now i want to add another ip address to it. what is the preferred way? adding an eth1 or doing this alias thing via eth0:1?
2019-10-31 12:58:01 <_ikke_> Leo_: interfaces can have multiple addresses
2019-10-31 12:58:15 <_ikke_> No need for a separate (sub) interface
2019-10-31 12:58:54 <Leo_> _ikke_, yes, but when i add multiple addresses, can i still say to for example ping use this special address for outgoing traffic? as far as i have seen this only works if i specify and interface
2019-10-31 12:59:17 <AinNero> Leo_: many programs have a 'bind adress' option for that
2019-10-31 12:59:17 <_ikke_> Yes, you can specify a source address
2019-10-31 12:59:57 <AinNero> busybox ping -I also works with IP adresses
2019-10-31 13:00:09 <_ikke_> Leo_: we have a (non AL) setup where we have two addresses, and via the default route we specify what address should be used
2019-10-31 13:00:54 <mps> or 'ip rule'
2019-10-31 13:00:55 <Leo_> oh, thats nice to hear! did not know that!
2019-10-31 13:01:16 <Leo_> how would this look in an interface file? just add two address lines?
2019-10-31 13:02:03 <_ikke_> Leo_: You need to add up commands
2019-10-31 13:02:04 <mps> post-up ip address add .......
2019-10-31 13:02:06 <_ikke_> up ip addr add
2019-10-31 13:02:08 <_ikke_> right
2019-10-31 13:02:30 <mps> and yes, 'up' is also good
2019-10-31 13:04:43 <_ikke_> up ip addr add dev eth0 <address>/<prefixlength>
2019-10-31 13:06:38 <Leo_> works perfectly!
2019-10-31 13:06:43 <Leo_> thanks to both of you!
2019-10-31 13:43:56 <kytart> hi, does anybody know how to completely disable on-screen terminal on RPi? Do I have to recompile the kernel or something or can I do it with the stock Alpine somehow? Thanks!
2019-10-31 13:51:02 <mps> kytart: what you mean by 'on-screen terminal'
2019-10-31 13:51:43 <_ikke_> Do you mean the different consoles?
2019-10-31 13:52:16 <kytart> mps: the basic terminal that you see when you turn on RPi connected to a monitor. I don't know what's the technical term for it lol sry
2019-10-31 13:53:01 <mps> I would disconnect cable, simply :)
2019-10-31 13:53:08 <kytart> I'm building a purely graphical app and need to get rid of the system initialization stuff on the screen
2019-10-31 13:53:23 <mps> ah, you mean console
2019-10-31 13:53:34 <kytart> I guess
2019-10-31 13:54:13 <mps> but you will need boot loader text at least
2019-10-31 13:54:45 <kytart> what's boot loader text?
2019-10-31 13:55:21 <mps> text menu which appears first after you boot computer
2019-10-31 13:55:33 <kytart> I want like complete blank, nothing, until I draw something to framebuffer
2019-10-31 13:56:15 <kytart> I found some info that says that in general you gotta disable some flag during kernel compilation to disable console
2019-10-31 13:56:20 <kytart> but I'm not an expert
2019-10-31 13:56:23 <mps> that could be done with setting boot loader timeout to zero
2019-10-31 13:56:58 <mps> and kernel cmdline to quiet
2019-10-31 13:57:38 <mps> and, no, you don't need to recompile kernel
2019-10-31 13:58:29 <mps> adding image to bootloader screen can help, but I don't know how to do this with RPi's bootloader
2019-10-31 13:58:56 <mps> you want something like phones boot, I presume
2019-10-31 14:00:30 <PureTryOut[m]> You mean uboot?
2019-10-31 14:00:59 <mps> PureTryOut[m]: does RPi can use uboot as bootloader
2019-10-31 14:01:36 <PureTryOut[m]> Probably, I don't see what else it would use
2019-10-31 14:02:07 <mps> afaik, it have some binary blob but not sure how it is used
2019-10-31 14:02:11 <PureTryOut[m]> There is https://pkgs.alpinelinux.org/package/edge/main/armv7/u-boot-raspberrypi
2019-10-31 14:03:12 <mps> heh, I pushed today uboot-tools to aports, and noticed that upstream removed easylogo utility
2019-10-31 14:06:50 <mps> and I think this utility is for 'splash' screen, although never tried it, I like text at boot
2019-10-31 14:13:43 <kytart> how do I use uboot? Sry, this is some advanced shit for me haha
2019-10-31 14:26:19 <kytart> I see that I should use it in place of kernel that rpi firmware boots from config.txt
2019-10-31 14:26:27 <kytart> but then what?
2019-10-31 16:47:24 <ptx0> anyone know how the alpine linux trigger system works? i.e. install new kernel, have initrd created. how does alpine know to use mkinitfs?
2019-10-31 16:49:01 <_ikke_> ptx0: https://git.alpinelinux.org/aports/tree/main/mkinitfs/APKBUILD#n17
2019-10-31 16:49:21 <_ikke_> https://git.alpinelinux.org/aports/tree/main/mkinitfs/mkinitfs.trigger
2019-10-31 16:51:22 <Cogitri> Triggers are added in the APKBUILD
2019-10-31 16:53:46 <_ikke_> anoying, somehow building on my 3.7 containers, a lot of packages fail to build becaue they rely on SHELL (being zsh for me)
2019-10-31 16:54:15 <_ikke_> I just added export SHELL=sh to my abuild.conf
2019-10-31 16:59:05 <ptx0> hmmm
2019-10-31 16:59:27 <ptx0> so if i have dracut packaged i need to add a trigger for its APKBUILD but can i disable the mkinitfs trigger somehow?
2019-10-31 17:00:23 <ptx0> i don't want mkinitfs trigger running but i doubt i can remove that package unless i somehow can tell apk that the dracut package is providing mkinitfs
2019-10-31 17:02:56 <TBB> out of curiosity... how come you've ported dracut to Alpine in the first place?
2019-10-31 17:03:05 <ptx0> iscsi boot from iBFT
2019-10-31 17:03:35 <ptx0> it allows me to share one iSCSI image between bare metal machines and also native VMs that don't netboot
2019-10-31 17:04:10 <ptx0> i do this with centos/debian/gentoo/others and i've gotten dracut running but it won't netboot, i have yet to figure that one out, currently just trying to fix the integration between dracut and alpine
2019-10-31 17:04:19 <_ikke_> ptx0: provides="mkinitfs"
2019-10-31 17:04:33 <ptx0> now my next question is whether this will royally fudge anything up
2019-10-31 17:04:53 <_ikke_> I have no clue
2019-10-31 17:04:57 <ptx0> if i provides="mkinitfs" think a lot of packages updates will fail since the binary isn't present eh
2019-10-31 17:05:05 <ptx0> unless i make a wrapper
2019-10-31 17:05:37 <TBB> that's how I imagine it is
2019-10-31 17:05:37 <ptx0> is there a file i can 'rm' just to destroy the mkinitfs trigger? :)
2019-10-31 17:05:51 <ptx0> chmod -x would be better
2019-10-31 17:06:34 <TBB> you can overpower mkinitfs with your dracut package. the exact magic to do that escapes my mind at the moment, but you can basically tell the system that the trigger file -you- provide is to replace the mkinitfs one
2019-10-31 17:06:48 <TBB> it's ugly but it works
2019-10-31 17:06:59 <_ikke_> I don't think you can replace single files with apk
2019-10-31 17:07:03 <_ikke_> it would complain\
2019-10-31 17:07:24 <TBB> it's been a while but I think it could be as simple as replaces=
2019-10-31 17:07:46 <TBB> the rest of the files from the mkinitfs apk might still remain under the ownership of that package or something
2019-10-31 17:07:48 <_ikke_> That replaces the entire package
2019-10-31 17:08:13 <TBB> we did something like that to take ownership of single files
2019-10-31 17:08:32 <ptx0> /lib/apk/db/triggers has some strings in it
2019-10-31 17:08:34 <ptx0> what are they?
2019-10-31 17:08:46 <ptx0> Q1kyqEnqGziBnv/fq2KNVA+b8MojM= /usr/share/kernel/*
2019-10-31 17:08:52 <ptx0> maybe i just delete this line?
2019-10-31 17:09:06 <TBB> but, all that said... Alpine's been missing something like dracut (not necessarily dracut tho) for quite some time now
2019-10-31 17:09:38 <ptx0> too bad there's no way to override trigger of another package explicitly
2019-10-31 17:09:39 <Cogitri> Just using dracut would be the best IMHO
2019-10-31 17:09:53 <ptx0> and i agree with Cogitri
2019-10-31 17:09:57 <Cogitri> Lots of other distros use it so rolling with what others use would be kinda nice
2019-10-31 17:10:14 <Cogitri> And lots of stuff (e.g. ZFS) provides support for dracut
2019-10-31 17:10:27 <ptx0> it's hosted on kernel.org and so on, i actually developed the zfs support for dracut with some others.
2019-10-31 17:10:37 <TBB> could be I'd be now with the experience to agree with that, but at least during my RHEL days dracut looked like an absolutely horrific mess
2019-10-31 17:11:01 <ptx0> dracut has changed substantially since its introduction to the point that some docs make no sense any longer
2019-10-31 17:11:31 <ptx0> one issue i have with it is that i can not override standard dracut module functionality using third party modules without overwriting their files
2019-10-31 17:11:48 <ptx0> i.e. i want to resume from swap partition held on zfs pool but resume module knows nothing about zfs module
2019-10-31 17:12:05 <TBB> mkinitfs is relatively clean; I'd almost rather add dynamic init script generation to its feature handling...
2019-10-31 17:12:08 <ptx0> have to modify dracut-resume to know that it needs to import zpool before resume.
2019-10-31 17:12:40 <ptx0> i looked at mkinitfs and considered adding iBFT support to it but that is a can of worms
2019-10-31 17:12:41 <TBB> but then, dracut's right there, it exists and does what it's supposed to
2019-10-31 17:13:18 <ptx0> also for anyone who doens't know, ibft is when the system firmware (pxe, etc) tells the OS where to find its iSCSI target, iSCSI Boot Firmware Table.
2019-10-31 17:13:26 <TBB> well, getting the necessary files and modules into the initramfs isn't that difficult; but you'd probably have to add code to initramfs-init as well
2019-10-31 17:14:03 <ptx0> no it's just that ibft itself is complicated
2019-10-31 17:14:14 <ptx0> also open-iscsi sucks
2019-10-31 17:14:34 <ptx0> i'd probably just copy most of the code from dracut directly if i want it to actually.. work vOv
2019-10-31 17:15:20 <TBB> now, if that was its own snippet and mkinitfs actually knew how to deal with snippets... that'd be the clean solution
2019-10-31 17:15:40 <ptx0> features.d/90-ibft.sh
2019-10-31 17:15:47 <ptx0> :P
2019-10-31 17:15:53 <TBB> something like that, yes
2019-10-31 17:16:18 <TBB> now, I remember talking about this on the developer channel and someone said it has been considered
2019-10-31 17:16:58 <ptx0> for now i'm going to just try and remove / override the mkinitfs trigger
2019-10-31 17:17:15 <ptx0> as others said dracut code is used by other distro so when they come out with a new feature i automagically benefit
2019-10-31 17:17:33 <ptx0> booting from bonded NIC over NBD or iSCSI is already there
2019-10-31 17:19:05 <TBB> ah well, soon it'll be systemd-mkinitramfs anyway :)
2019-10-31 17:21:12 <mps> TBB: yes, more and more people tries to drag us to systemd
2019-10-31 17:21:27 <ptx0> well dracut supports systemd and openrc
2019-10-31 17:22:00 <ptx0> my people at redhat are discussing removing systemd anyway
2019-10-31 17:22:09 <Cogitri> Yes
2019-10-31 17:22:25 <TBB> wow. that's surprising.
2019-10-31 17:22:26 <ptx0> they are going to replace it with something even worse so dont get excited
2019-10-31 17:22:47 <MY-R> fork of systemd? :D
2019-10-31 17:22:54 <ptx0> yeah my company has this channel to suggest changes to the distro
2019-10-31 17:23:07 <ptx0> we have been stuck with centos 6.9 and they go EOL in 2020
2019-10-31 17:23:14 <mps> ptx0: oh, I just mentioned few minutes ago on #alpine-devel that the debian discussing systemd again
2019-10-31 17:23:36 <ptx0> we've been trying to get them to listen but our millions of dollars aren't enough apparently so now we're looking at alpine which is why i'm here
2019-10-31 17:23:52 <Cogitri> Ah, neat
2019-10-31 17:24:04 <ptx0> now they're keen on removing systemd but like someone wise once said, 'too little, too late, harry'
2019-10-31 17:24:07 <Cogitri> Well, help is always appreciated :P
2019-10-31 17:24:51 <ptx0> pardon my french but binary logs are retarded
2019-10-31 17:25:08 <mps> on debian-devel alpine is mentioned as successful distro without systemd
2019-10-31 17:25:29 <ptx0> was that the first one they mentioned or was it after gentoo? :P
2019-10-31 17:25:39 <TBB> as a sysadmin, I totally agree. as a dreamer... everything could be so much better
2019-10-31 17:26:13 <ptx0> i have customers sending me logs all the time and journalctl is the last thing i need
2019-10-31 17:26:26 <mps> don't worry, more than a 3 decades I'm reading that the unix will die tomorrow :)
2019-10-31 17:26:32 <Cogitri> tbh, I kinda like systemd if it wasn't for it not working on musl
2019-10-31 17:26:52 <_ikke_> Cogitri: I feel the same
2019-10-31 17:26:55 <Cogitri> ptx0: I personally prefer having journalctl instead of having a mediocre syslog
2019-10-31 17:27:04 <crosbymichael_> i like systemd as an init system but resolved and all the mount stuff has started to get in my way
2019-10-31 17:27:10 <TBB> I just read Lennart's original essay on his new init system, and I've gotta say I agreed with a lot of what was said
2019-10-31 17:27:20 <_ikke_> crosbymichael_: I never use resolved with systemd
2019-10-31 17:27:31 <Cogitri> Hm, systemd-automount is pretty good compared to autofs IMHO
2019-10-31 17:27:44 <crosbymichael_> ubuntu has it by default so I have to disable it each time
2019-10-31 17:27:46 <TBB> it's just that reality took a different direction, and now it's a borg cube
2019-10-31 17:27:50 <_ikke_> crosbymichael_: yeah, indeed
2019-10-31 17:28:49 <mps> I was one of the early adopter of systemd, but when it started to be everything I switched from it
2019-10-31 17:29:28 <crosbymichael_> feel the same way.  as an init, it's nice and clean, easy to use.  everything else that keeps getting piled on is making things harder
2019-10-31 17:31:04 <TBB> some things are described as solutions seeking a problem
2019-10-31 17:31:15 <TBB> in case of systemd, it goes a step further and generates its own
2019-10-31 17:32:06 <MY-R> 
ACTION got alergy on "[    **] A start/stop job is running.........(2min 55s / 10 min)" :D

2019-10-31 17:32:22 <mps> MY-R: right
2019-10-31 17:32:38 <_ikke_> Yeah, that's kind of anoying
2019-10-31 17:32:40 <MY-R> then Im clicking Ctrl + C and.... nothing
2019-10-31 17:32:49 <TBB> me too. besides, I want my [  ]'s with green text to the RIGHT
2019-10-31 17:32:52 <MY-R> doing ctrl alt del
2019-10-31 17:32:53 <MY-R> NOTHING
2019-10-31 17:33:04 <_ikke_> MY-R: try to do it 7 times :P
2019-10-31 17:33:12 <_ikke_> in quick succession
2019-10-31 17:33:14 <MY-R> _ikke_, you joking, right? :D
2019-10-31 17:33:17 <_ikke_> nope
2019-10-31 17:33:23 <MY-R> ahaha no way!
2019-10-31 17:33:40 <MY-R> any other cool "fatality" combination which I should know, like in windows? :D
2019-10-31 17:33:42 <_ikke_> basically a rage function
2019-10-31 17:33:44 <mps> don't ask me to define unix philosophy but systemd is against unix philosophy
2019-10-31 17:33:59 <_ikke_> (1 guess how I found out :P)
2019-10-31 17:34:23 <MY-R> damn...
2019-10-31 17:34:39 <mps> MY-R: reset button :D
2019-10-31 17:34:52 <TBB> we could've just gone up up down down left right left right b a
2019-10-31 17:34:56 <_ikke_> :D
2019-10-31 17:35:15 <MY-R> mps, ye that is what systemd is... my dusted reset button was abused when was using it :D
2019-10-31 17:35:24 <ptx0> the systemd emergency shell won't even let me run a command that takes more than 15 seconds to execute
2019-10-31 17:35:30 <ptx0> what a great shell
2019-10-31 17:35:43 <AinNero> man, guys, what about going to offtopic?
2019-10-31 17:35:51 <AinNero> im not here to listen to systemd stories
2019-10-31 17:36:10 <ptx0> i'm not here to listen to what other people are here to listen to
2019-10-31 17:36:15 <mps> I'm not joking, sometimes I really press reset when waiting for systemd to finish some simple thongs
2019-10-31 17:36:23 <mps> things*
2019-10-31 17:36:26 <ptx0> scroll back up and leave the channel at some point in the past :)
2019-10-31 17:36:29 <crosbymichael_> things...sure
2019-10-31 17:37:35 <ptx0> * mps is now known as Sisqo
2019-10-31 17:38:41 <AinNero> ptx0 today in: "how to train people to stop taking you seriously", part 1
2019-10-31 17:42:29 <ptx0> AinNero: i'm sorry if i offended you. who are you, again?
2019-10-31 17:42:47 <_ikke_> AinNero has a lot more standing in this community than you do :)
2019-10-31 17:42:57 <AinNero> _ikke_: [citation needed]
2019-10-31 17:42:59 <ptx0> i was being sincere. i have no clue who they are.
2019-10-31 17:43:47 <ptx0> don't care much for 'standing', i just want to make the thing work. sometimes we have to discuss ugly things in order to learn.
2019-10-31 17:44:24 <mps> I don't see anything wrong in ptx0 writings here
2019-10-31 17:44:57 <ptx0> i did use the word retarded earlier. i felt bad after i did it.
2019-10-31 17:45:45 <AinNero> 
ACTION scratches head

2019-10-31 17:46:17 <mps> a lot of us are non native English speakers and we sometimes use words with bad connotations but mostly without bad intentions
2019-10-31 17:46:36 <AinNero> like, ptx0 gave a weird useless response to me, so i gave a weird useless response to ptx0
2019-10-31 17:47:17 <mps> AinNero: I don't see any serious issue with both of you
2019-10-31 17:47:56 <ptx0> yeah gotta love how a complaint about being off-topic lead to ... surprise, even more off-topic
2019-10-31 17:48:03 <_ikke_> "who are you again" can be considered "hostile"
2019-10-31 17:48:16 <AinNero> _ikke_: i have enough skin for that
2019-10-31 17:48:17 <_ikke_> (I'm saying this in a non-accusatory way)
2019-10-31 17:48:23 <mps> _ikke_: really?
2019-10-31 17:48:36 <AinNero> like, im somewhat if introvert, im actually quite fine with being some anonymous bystander
2019-10-31 17:48:38 <ptx0> mps: it is true, anything can be considered hostile if you have enough context.
2019-10-31 17:49:11 <_ikke_> mps: something like: "Who are you tell me what to do"
2019-10-31 17:49:19 <_ikke_> to tell me*
2019-10-31 17:49:40 <ptx0> exactly, i had no idea if they're some major project leader who sets rules on the IRC channel
2019-10-31 17:49:57 <ptx0> i'm autistic, i'm not an asshole :P
2019-10-31 17:49:58 <AinNero> ptx0: IMHO even if, that shouldn't make a difference
2019-10-31 17:50:00 <nepugia> ugh, why isnt the wacom driver packaged
2019-10-31 17:50:09 <nepugia> ptx0, those arent mutually exclusive :)
2019-10-31 17:50:17 <mps> If I have thin skin I would long ago disconnect from all free/open IT world
2019-10-31 17:50:37 <ptx0> nepugia: because you haven't come along yet to submit a PR of course
2019-10-31 17:50:52 <nepugia> A "PR"
2019-10-31 17:50:59 <nepugia> not a 'patch'? hmmm
2019-10-31 17:51:09 <MY-R> or MR!?
2019-10-31 17:51:15 <_ikke_> whatever floats your boat
2019-10-31 17:51:23 <ptx0> autobots, assembla
2019-10-31 17:51:31 <ptx0> 
ACTION submits merge request

2019-10-31 17:51:35 <nepugia> something one can, you know, submit per email...
2019-10-31 17:51:47 <mps> ptx0: only those with '@' prefix in their nick are somehow important here
2019-10-31 17:51:52 <TBB> oh if I could just get my eyes off this entertainment, reinstall my system as has been my intention for months, and start packaging...
2019-10-31 17:51:53 <ptx0> nepugia: an email can be considered a pull request. can you pull this from my git repo? thanks
2019-10-31 17:52:00 <_ikke_> not at all
2019-10-31 17:52:04 <nepugia> if you read the irc spec you will realize that nobody has a @ in their nick
2019-10-31 17:52:19 <nepugia> ptx0, emails usually contain patches directly
2019-10-31 17:52:53 <ptx0> html formatted, hopefully
2019-10-31 17:52:55 <nepugia> i doubt that most developer would even consider manually merging a remote branch, since that is a huge pain in comprison
2019-10-31 17:53:18 <nepugia> no, why would you send a patch as text/html?
2019-10-31 17:53:20 <AinNero> nepugia: should be a single git fetch...
2019-10-31 17:53:24 <_ikke_> nepugia: even git.git uses pull requests (in a traditional git request-pull way)
2019-10-31 17:54:13 <_ikke_> imho there are less issues with fetching and merging than applying patches and merging
2019-10-31 17:54:23 <nepugia> AinNero, Having an email already downloaded and applying it requires no additional data, remotely pulling data requires pulling remote data
2019-10-31 17:54:27 <mps> 'git send-email' to aport mailing list
2019-10-31 17:54:38 <nepugia> the just having the patch patch case is easier
2019-10-31 17:55:01 <mps> s/aport/aports/
2019-10-31 17:55:07 <_ikke_> nepugia: pulling remote data equals to downloading the patches
2019-10-31 17:55:14 <ptx0> yeah that just blew my mind
2019-10-31 17:55:28 <ptx0> 6 of one and half a dozen of the other
2019-10-31 17:55:42 <nepugia> _ikke_, cloning a remote repo is certainly much more than /just/ a patch
2019-10-31 17:55:43 <mps> actually, at the bottom these are always patches
2019-10-31 17:55:55 <ptx0> no one needs to clone the repo
2019-10-31 17:56:01 <_ikke_> nepugia: if you want to apply patches, you already need to have repo
2019-10-31 17:56:16 <_ikke_> it's just a matter of fetching the commits they want you to pull
2019-10-31 17:56:16 <ptx0> see man page for 'git-am'
2019-10-31 17:56:41 <_ikke_> which can be quite efficient (
2019-10-31 17:56:50 <nepugia> Its still more work than already having the patches...
2019-10-31 17:56:58 <_ikke_> it's less work for me
2019-10-31 17:57:02 <ptx0> magic is indeed easier than doing work.
2019-10-31 17:57:09 <nepugia> I dont see why i should do that more work if they could have just send the patches in the first place
2019-10-31 17:57:14 <_ikke_> git fetch <url> <refspec>; done
2019-10-31 17:57:32 <ptx0> _ikke_: parse error near 'done'
2019-10-31 17:57:45 <_ikke_> And I would still have the refs :P
2019-10-31 17:57:56 <nepugia> Which requires way more infrastructure than having a text document on your computer
2019-10-31 17:58:04 <ptx0> _ikke_: not on zsh in my testing :P
2019-10-31 17:58:36 <ptx0> nepugia: but your imagined scenario isn't real.
2019-10-31 17:59:00 <_ikke_> nepugia: You need infrastructure for sending e-mails as well
2019-10-31 17:59:09 <nepugia> ptx0, my internet connection fucking up randomly because of shitty wifi drivers isnt real? good riddance
2019-10-31 17:59:32 <ptx0> you don't even need an internet connection if you just travel to the developer's house with a hard copy of the patch. no computer needed, just write pen and paper.
2019-10-31 17:59:41 <ptx0> if you walk you don't even need a car. you can hike straight across wales.
2019-10-31 18:02:16 <AinNero> man wtf, how meta did this even get
2019-10-31 18:02:32 <AinNero> what was with the wacom drivers?
2019-10-31 18:03:14 <_ikke_> xf86-input-wacom exists
2019-10-31 18:03:39 <ptx0> i think there's a UI tool that isn't available
2019-10-31 18:03:43 <nepugia> not in the normal repos
2019-10-31 18:04:01 <nepugia> not sure what you use... :)
2019-10-31 18:04:22 <ptx0> oh yeah it's only in edge
2019-10-31 18:04:27 <ptx0> i guess you can build it though
2019-10-31 18:04:44 <ptx0> do edge apkbuild often work in v3.10 without modification?
2019-10-31 18:04:47 <_ikke_> It's in testing indeed
2019-10-31 18:04:50 <nepugia> I already build it
2019-10-31 18:05:03 <nepugia> not with any alpine stuff though, just normally
2019-10-31 18:05:03 <_ikke_> ptx0: There is no guarantee they work together
2019-10-31 18:05:12 <ptx0> yeah i figured no guarantee
2019-10-31 18:05:12 <nepugia> still wondered why it wasnt available
2019-10-31 18:05:24 <_ikke_> Because it's testing, which is not built for stable branches, only for edge
2019-10-31 18:05:27 <ptx0> just wondered if there's some new non-backwards compatible apk language changes
2019-10-31 18:05:35 <ptx0> how does something get out of testing?
2019-10-31 18:05:44 <_ikke_> up to the maintainer of that package
2019-10-31 18:05:56 <_ikke_> Providing feedback that a package works properly sure helps
2019-10-31 18:06:02 <ptx0> oh so it could have been packaged long before 3.10 but just never considered stable?
2019-10-31 18:06:10 <ptx0> that makes sense.
2019-10-31 18:06:22 <ptx0> i guess nepugia needs to contact its maintainer then and find out why it's not in stable.
2019-10-31 18:06:53 <ptx0> i've been porting Adelie builds to alpine so i can understand 'no guarantee'
2019-10-31 18:06:58 <ptx0> :)
2019-10-31 18:07:27 <_ikke_> dependencies are the problem
2019-10-31 18:07:43 <_ikke_> a package can be built against dependency versions that are only in edge
2019-10-31 18:10:24 <nepugia> ptx0, why? "its not stable" is enough of an explanation
2019-10-31 18:11:27 <_ikke_> nepugia: it's not *verified* stable
2019-10-31 18:12:03 <nepugia> i dont see any difference between those states
2019-10-31 18:12:17 <_ikke_> The package could be perfectly stable but still live in testing
2019-10-31 18:12:18 <nepugia> unless you are saying that "verified stable unstable" stuff is in the normal repos
2019-10-31 18:12:25 <ptx0> it's ok, the difference exists. if you want to resolve it you can test the build in edge and report to the maintainer any issues.
2019-10-31 18:12:42 <_ikke_> nepugia: ie, the fact that it's in testing does not mean it's unstable
2019-10-31 18:12:54 <nepugia> i dont care enough, my wacom tablet works now
2019-10-31 18:13:02 <nepugia> _ikke_, no, it means that it isnt stable
2019-10-31 18:13:09 <_ikke_> nepugia: ..
2019-10-31 18:13:10 <ptx0> so that's the answer then to why it's not available outside of testing repo
2019-10-31 18:13:16 <nepugia> not having been tested enough also implies that
2019-10-31 18:13:20 <ptx0> because you don't care enough and neither do the other wacom users :D
2019-10-31 18:13:45 <_ikke_> testing does not make a package stable
2019-10-31 18:13:52 <_ikke_> it verifies it's stable
2019-10-31 18:14:19 <nepugia> Really? ive never had /no/ issues in testing software
2019-10-31 18:14:23 <nepugia> like, never
2019-10-31 18:14:57 <_ikke_> I mean the act of testing
2019-10-31 18:15:11 <_ikke_> itself
2019-10-31 18:15:24 <_ikke_> A lot of packages in testing are probably just fine
2019-10-31 18:15:38 <_ikke_> and could be moved to community if someone bothered
2019-10-31 18:16:04 <_ikke_> nepugia: And note that the testing repo is mostly fore testing the alpine packaging, not the software itself
2019-10-31 18:16:32 <mps> and software in community or main doesn't guarantes it is stable, bugs arises from time to time
2019-10-31 18:16:38 <_ikke_> indeed
2019-10-31 18:16:59 <nepugia> Uhh, that makes little sense, alpine is a different enviroment to other distros, you obviously want to test whether the software works properly before you distribute it, whether it runs on other distros doesnt really matter
2019-10-31 18:17:19 <_ikke_> yes, but that falls under packaging
2019-10-31 18:17:32 <_ikke_> We don't expect that each feature of a project is tested
2019-10-31 18:17:35 <mps> nepugia: because that we have releases
2019-10-31 18:17:35 <nepugia> no, that falls under bug testing and fixing
2019-10-31 18:18:28 <_ikke_> We *do* run the test suites of each package if it's available and feasible
2019-10-31 18:18:31 <ptx0> nepugia: anyway, glad your issue is resolved now and you know you can use apkbuild from edge in v3.10 if you just modify a few dependencies. this will at least give you a 'native'  package with any patches the maintainer determines were necessary.
2019-10-31 18:19:17 <_ikke_> nepugia: Your premise that because a package is in testing it's unstable is just plain false
2019-10-31 18:19:20 <nepugia> Its native in either case :)
2019-10-31 18:19:28 <nepugia> _ikke_, it was your premise but okay...
2019-10-31 18:19:42 <mps> I build a lot of packages from edge in stable, rarely I need any change
2019-10-31 18:19:53 <_ikke_> nepugia: I said it was not part of a stable branch
2019-10-31 18:20:26 <_ikke_> meaning, that it's built for stable alpine releases
2019-10-31 18:20:45 <ptx0> i just don't understand, nepugia came in here to ask why it's not packaged and then got an answer but say they don't care, and they continue to go on about pedantics
2019-10-31 18:21:01 <ptx0> i think they came in here to argue and not to get anything resolved
2019-10-31 18:21:06 <nepugia> I said that i dont care to package it myself
2019-10-31 18:21:13 <_ikke_> nepugia: You don't have to
2019-10-31 18:21:16 <_ikke_> it is already packaged
2019-10-31 18:21:22 <nepugia> i dont see why you are trying to assign that responsibility to me, i only asked a question
2019-10-31 18:21:42 <ptx0> yeah, just email the maintainer after testing it to let them know if it worked or not. you already said you are familiar with sending patches via email, this is even easier than that.
2019-10-31 18:21:59 <_ikke_> Alpine is a community driven project. If you want to get something done, the fastest way is to contribute
2019-10-31 18:22:13 <_ikke_> If you don't care for that, fine, but don't require someone else to do it for you either
2019-10-31 18:22:29 <nepugia> I did not require anything, i simply asked a question
2019-10-31 18:22:39 <nepugia> If nobody had responded that would have been fine too
2019-10-31 18:22:42 <ptx0> i don't use wacom tablet nor do i own one or i might do this for you, otherwise it is up to you like most things in Linux.
2019-10-31 18:23:07 <_ikke_> nepugia: fyi: https://pkgs.alpinelinux.org/package/edge/testing/x86_64/xf86-input-wacom
2019-10-31 18:23:11 <nepugia> ptx0, the only thing i wanted was an answer to my question, not that anyone does anything for me
2019-10-31 18:23:36 <nepugia> if i do have alpine on another computer with a wacom tablet i might aswell test the edge thing then to test it
2019-10-31 18:23:38 <ptx0> nepugia: the answer was because the maintainer said so, and what do you need now to be satisfied and go on about your day?
2019-10-31 18:23:42 <_ikke_> nepugia: The question you asked was kind of in the sense "why didn't any one to package this for me yet"
2019-10-31 18:23:58 <nepugia> _ikke_, that is literally not what i said
2019-10-31 18:24:29 <ptx0> well _ikke_ has colourful interpretations of text sometimes so give them a break, they are trying to help you
2019-10-31 18:24:54 <nepugia> ptx0, you are asking me stuff, not me, i dont see any point in this discussion anymore, but i do respond out of courtesy
2019-10-31 18:25:02 <_ikke_> Let's stop then
2019-10-31 18:25:05 <nepugia> ptx0, give them a break from what?
2019-10-31 18:27:55 <ptx0> hm, zfs can now use avx512bw for fletcher4 on skylake+ CPUs
2019-10-31 18:28:24 <ptx0> byteswap occurs in a single instruction instead of... many :D
2019-10-31 18:28:42 <ptx0> should help improve performance on large systems with the appropriate hw
2019-10-31 20:30:13 <hooogs> The default behavior of Alpine is to automount the partition containing the APKOVL on boot under /media/sda1 (the partition and disk are happen to be sda1 in this case). In order to commit changes to it, I have to run `mount -o -rw,remount /dev/sda1`, because by default it is mounted read only. I have added sda1 to fstab, but no matter what options
2019-10-31 20:30:14 <hooogs> I specify, the fstab settings seem to be overridden by the boot time auto mounting. How do I get the partition containing the APKOVL to mount as R/W on boot?
2019-10-31 20:30:53 <hooogs> disks happen to be*
2019-10-31 20:36:43 <AinNero> how do you commit changes?
2019-10-31 20:37:26 <AinNero> because 'lbu commit' handles remounting the boot device on its own
2019-10-31 20:55:22 <hooogs> The basic process is: I have an ext4 partition (partition type doesn't matter, that is just what I happen to be using) that I am storing the APKOVL on. On boot, Alpine automatically scans that volume, loads the APKOVL from there, and mounts it readonly.
2019-10-31 20:55:47 <hooogs> To make config changes, I remount that partition as rw, then I use lbu pkg <mountpoint> to update the config.
2019-10-31 20:56:05 <AinNero> use lbu commit
2019-10-31 21:04:16 <hooogs> Yes, that is what I would normally be doing.
2019-10-31 21:05:20 <hooogs> But I need to set a backup directory in /etc/lbu/lbu.conf, and the sda1 partition is mounted read only. Are you saying that lbu will automatically remount it rw if it has to?
2019-10-31 21:05:59 <AinNero> see /sbin/lbu_commit, function mount_once_rw()
2019-10-31 21:06:02 <AinNero> yes
2019-10-31 21:06:39 <hooogs> Another thing: even if it does, I would prefer to specify this by UUID instead, e.g. via fstab. There may be multiple USB devices plugged into this box, and I don't want it committing config changes to the wrong one just because it happens to be sda1 on that particular boot.
2019-10-31 21:08:00 <AinNero> hm, UUID isn't supported at this place yet
2019-10-31 21:08:40 <hooogs> Yes, that is what I assumed. So, how do I override the on-boot auto mounting and make Alpine use the options in fstab instead?
2019-10-31 21:09:04 <hooogs> That way, I can make sure that only that specific UUID is mounted at the path I specify in lbu.conf.
2019-10-31 21:11:22 <AinNero> the initramfs has code to respect the /etc/fstab in the apkovl
2019-10-31 21:12:40 <AinNero> like, automounting is done as usual, but the mounts are relocated
2019-10-31 21:12:48 <AinNero> i haven't ever used this feature before :P
2019-10-31 21:13:16 <AinNero> relocate_mounts() in /usr/share/mkinitfs/initramfs-init
2019-10-31 21:26:06 <hooogs> Just putting this here for my own reference: there are two issues. 1. Does lbu successfully commit changes to a dir mounted as an R/O FS? 2. Does automount respect the settings in fstab?
2019-10-31 21:33:13 <hooogs> Ok, I have tested it. With LBU_BACKUP_DIR set to the mount point /media/sda1, lbu does not create the archive because the partition is mounted readonly.
2019-10-31 21:33:26 <hooogs> This is with the stock fstab.
2019-10-31 21:33:35 <hooogs> After I remount it R/W, it works.
2019-10-31 21:34:16 <hooogs> So, if lbu is supposed to be remounting the partition as rw automatically... it isn't.
2019-10-31 21:34:42 <hooogs> Also: I am using lbu ci, not lbu pkg here.
2019-10-31 21:38:42 <hooogs> Again, I have not touched fstab in the above case. I am simply trying to lbu commit to the read only partition that was automatically mounted.
2019-10-31 21:51:37 <hooogs> I just tested the whether the automount respects fstab. It doesn't.
2019-10-31 21:51:48 <hooogs> Here is my fstab: http://sprunge.us/oN3R7L
2019-10-31 21:52:08 <hooogs> Here is the output of mount | grep sda right after boot: http://sprunge.us/UV7Bkq
2019-10-31 21:53:02 <hooogs> This is in the Alpine Virtual image, release 3.10.3, for x86.
2019-10-31 21:54:24 <hooogs> As you can see, it should be mounting /dev/sda1 as /media/disk1, but the automount seems to override fstab and mount it as /media/sda1 instead.
2019-10-31 23:28:09 <KH405> I can't get podman to work on alpine, If someone could package it that would be great!
2019-10-31 23:33:20 <hooogs> AinNero, any ideas about the auto mount issue?
2019-10-31 23:42:21 <michael__> what runlevel should `networking` be at?