2021-07-01 08:34:02 <AppAraat[m]> Q: So if you copy `/etc/apk/repositories` and `/etc/apk/world` to a fresh new install, and then do `apk fix`, all the packages which were on the old machine will get installed onto the new machine?
2021-07-01 08:45:17 <clandmeter> A: yes
2021-07-01 09:00:49 <AppAraat[m]> O: That's pretty sweet!
2021-07-01 11:07:56 <nescafe> Ah, [m] is matrix, not gender :P
2021-07-01 11:08:12 <ikke> yes
2021-07-01 11:11:40 <clandmeter> lol
2021-07-01 11:18:36 <valerius> the matrix has them
2021-07-01 12:34:38 <AppAraat[m]> sorry, Matrix bridge just went bork.
2021-07-01 12:35:21 <MY-R> how is it with firefox in stable? will be updated sooner or later? now I see 89.0 and there is already 89.0.2
2021-07-01 13:55:46 <yyp> Why apk install/upgrade hooks don't propagate environment variables from the parent apk process?
2021-07-01 14:09:06 <lopid> do you mean why don't they inherit? have you exported the variable?
2021-07-01 14:11:29 <yyp> I did
2021-07-01 14:11:41 <yyp> But the hook didn't get it
2021-07-01 14:12:19 <lopid> unfortunately i don't know about apk hooks
2021-07-01 14:33:08 <ansimita> Hope it's ok to ask questions about busybox ash here: What is the best way to disable the creation of ~/.ash_history?
2021-07-01 14:33:11 <ansimita> My current solution is setting HISTFILE to /dev/null based on https://git.busybox.net/busybox/tree/shell/ash.c?h=1_33_stable#n14504
2021-07-01 14:36:38 <ansimita> Coming from bash, I thought unset-ing HISTFILE would do the trick but nevertheless, the code says otherwise.
2021-07-01 15:22:28 <przemoc> hi, was anyone recently updating alpine over many releases, like 3.11 -> 3.14?  is there anything worth paying extra attention for such move?
2021-07-01 15:37:06 <MY-R> przemoc: I would start from reading "upgrade/release notes" https://tpaste.us/EMLR
2021-07-01 15:41:23 <ansimita> Just came across this in the SUSv3 spec: "If the HISTFILE variable is not set, the shell may attempt to access or create a file .sh_history in the directory referred to by the HOME environment variable."
2021-07-01 15:42:09 <ansimita> "Implementations may choose to disable the history list mechanism for users with appropriate privileges who do not set HISTFILE ; the specific circumstances under which this occurs are implementation-defined."
2021-07-01 15:47:21 <minimal> MY-R: re hwclock/osclock with RPI diskless, it is a part of the initramfs shellscript that enables the hwclock init.d
2021-07-01 15:52:25 <MY-R> minimal: so running mkinitfs after remove hwclock will help or it always detect diskless setup and put hwclock in services?
2021-07-01 15:53:29 <minimal> MY-R: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#734
2021-07-01 15:53:54 <minimal> oops, https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L734
2021-07-01 15:54:25 <MY-R> ye Im reading that now
2021-07-01 15:56:12 <MY-R> so ye, running mkinitramfs is enough BUT it isnt even installed by default and doubt that people doing it on diskless mode
2021-07-01 15:56:21 <minimal> well I'm not sure how the behaviour differs between diskless and sys-mode - that code should enable hwclock in both cases but when I'm using sys with osclock enabled then hwclock isn't being run
2021-07-01 15:56:32 <MY-R> and need to mount fat aprtition as writable an dlink /boot to /media* stuff
2021-07-01 15:57:44 <MY-R> oh ye like I said, need do few steps with link and mount and install mkinitfs package to regenerate initramfs, but good to know :)
2021-07-01 15:59:13 <minimal> not familiar with diskless but I assume any enabled init.d scripts are stored in the LBU file rather than the initramfs
2021-07-01 15:59:49 <ikke> the symlink is
2021-07-01 16:00:22 <ikke> complete /etc by default
2021-07-01 16:01:26 <MY-R> minimal: in diskless mode initramfs image is provided in .tar.gz image so is pregenerated and exist in read only fat partition
2021-07-01 16:01:48 <minimal> ok, some day must sit down and try diskless mode :-)
2021-07-01 16:08:36 <MY-R> minimal: so you running this if wanna generate initramfs :) https://tpaste.us/0wN6
2021-07-01 16:09:48 <minimal> MY-R: thanks, I'd like to add diskless alpine images generation to my script in the future
2021-07-01 16:11:32 <MY-R> :)
2021-07-01 16:14:02 <minimal> MY-R: haven't gotten back to the LUKS2 stuff yet - stuck investigating a LUKS-related nlplug-findfs issue
2021-07-01 16:35:18 <xordspar0> ansimita: I think your reading of the busybox ash code is correct. I don't see a way to explicitly disable the history file (other than compiling the feature out).
2021-07-01 16:47:09 <invoked> it'd be nice if it was a commandline switch
2021-07-01 16:47:36 <invoked> but, bloat
2021-07-01 16:50:08 <MY-R> minimal: what issue!? all else working fine! :P
2021-07-01 16:50:54 <minimal> MY-R: its to do with the remote LUKS unlock stuff I'm working on
2021-07-01 16:55:51 <MY-R> minimal: ah, so unencrypted /boot and some ssh module? I remember dracut was working fine in other distributions
2021-07-01 16:58:00 <minimal> MY-R: yeah dropbear providing ssh in initramfs. The SSH part is working fine but then when nlplug-findfs is called to deal with luks its (currently) not working for me.....doing fiddly debugging currently. Once its sorted I'll be raising MRs for mkinitfs and dropbear to add it
2021-07-01 17:27:50 <AppAraat[m]> Any special instructions on how to access the Pi ZW via USB TTL (serial) cable? Under Raspbian / Raspberry Pi OS, usually that is done via raspi-config in order to prepare the OS. What is the equivalent of that for Alpine?
2021-07-01 17:29:05 <MY-R> minimal: oh nice, good luck! :)
2021-07-01 18:18:31 <Hello71> AppAraat[m]: https://www.raspberrypi.org/documentation/configuration/uart.md
2021-07-01 18:19:05 <Hello71> you can also use g_serial or g_ether
2021-07-01 18:24:01 <AppAraat[m]> Hello71: yeah the `enable_uart=1` I'm familiar with (AFAIK it's a setting for the SoC), but I've found out that the OS also has to be prepared in a way. For example on Raspbian you have to let systemd know that you're assigning a tty or something. That's what the raspi-config does.
2021-07-01 18:24:18 <AppAraat[m]> But if Alpine doesn't need all of that and works just like that out of the box then that's even better.
2021-07-01 19:32:22 <Hello71> can't remember
2021-07-01 19:32:51 <Hello71> if it does need configuration then systemctl enable serial-getty@ttyS0 is equivalent to echo mumblemumble >> /etc/inittab or console=ttyS0
2021-07-01 19:33:02 <Hello71> actually console=ttyS0 is systemd too
2021-07-01 19:33:24 <Hello71> although mkinitfs does it differently from systemd
2021-07-01 19:45:48 <minimal> I'd expect you'd need to add the serial device to /etc/inittab for a getty, and add console=<device> to the RPI command line config in config.txt. Aren't the RPI serial devices /dev/AMA0 etc
2021-07-01 19:47:47 <minimal> AppAraat: ^^^^^^
2021-07-01 19:50:51 <AppAraat[m]> minimal: I'm not sure about the `console=` part, as I can't find that anywhere here: https://www.raspberrypi.org/documentation/configuration/device-tree.md
2021-07-01 19:51:26 <AppAraat[m]> but I'll try. This is all new territory for me. I don't think I've even connected to a Pi over UART running Raspbian.
2021-07-01 19:54:42 <minimal> "console=" is a general Linux cmdline parameter to tell it what device the console is on, typically on PCs it would be "console=tty1" by default, so on a PC if you want a serial console you either change that to "console=ttyS0,115200" to prepend that before an existing "console=tty1". For RPI the serial device naming is different, have a look in /dev at the names of the tty files
2021-07-01 19:56:38 <ansimita> xordspar0: yeah. Setting HISTFILE to '/' (a directory) seems to work too as it causes open() to fail.
2021-07-01 20:48:59 <figbert> Does anybody know where the synching configuration is located when running from OpenRC?
2021-07-01 20:49:05 <figbert> I can't seem to find it
2021-07-01 21:01:33 <xordspar0> figbert: I believe it's in /var/lib/syncthing. Normally it's in the user's directory under .config/syncthing, and the openrc service sets the user's directory as /var/lib/syncthing.
2021-07-01 21:03:15 <figbert> xordspar0: thank you – it's there!
2021-07-01 21:05:50 <xordspar0> Maybe the service should be updated so that the config file is at /etc/syncthing. Syncthing takes a -config X option.
2021-07-01 21:10:24 <AppAraat[m]> ugh... spent half an hour troubleshooting why my Pi wouldn't show any output on the display and it turns out the display cable wasn't connected properly -_-
2021-07-01 21:11:27 <MY-R> AppAraat[m]: that is how Im charging phone :P
2021-07-01 21:12:11 <c705> in 3.14, i've lost the ability to passthrough usb devices though libvird. I can add the devices and start the machine, OK, no errors, but I can't use the passed though devices in the VM. Anyone know offhand if there's another package I need to add now while I keep troubleshooting?
2021-07-01 21:12:57 <AppAraat[m]> MY-R: huh? By connecting it to your display? :p
2021-07-01 21:14:28 <MY-R> AppAraat[m]: connecting cable to phone but forgeting connect charger to socket :P
2021-07-01 21:14:40 <AppAraat[m]> lmao yeah been there done that :p
2021-07-01 21:14:52 <c705> that's like a weekly occurance for me
2021-07-02 01:18:34 <lmcdo> Perhaps not the right question, but do any of the current ARM64 builds include Hyper-V patches to run as a Hyper-V guest?
2021-07-02 08:59:03 <Ristovski> Is there a way to slim down alpine-base even more? I'm trying to make an ultra-light VM optimized for very fast boot times. One idea I had was to make openrc run in parallel, but I would like to slim down the base system
2021-07-02 09:02:07 <przemoc> MY-R: thanks, somehow over all these years I haven't noticed (or maybe forgot?) there were release notes on wiki
2021-07-02 09:02:07 <lopid> you've no doubt reduced the extlinux timeout already
2021-07-02 09:04:01 <lopid> you could probably put everything in the initrd if you don't need persistence 
2021-07-02 09:22:51 <Ristovski> lopid: Yes, I suppose I could do that, since all the persistent data will be on a separate disk
2021-07-02 09:24:40 <Sheila> only really need alpine-baselayout, busybox, alpine-keys, and apk-tools. the rest you could do an overlayfs for.
2021-07-02 09:27:51 <Ristovski> Sheila: Is there something like https://github.com/alpinelinux/alpine-make-vm-image but more customizable? Perhaps something that bootstraps via apk.static
2021-07-02 09:28:08 <Ristovski> That would be ideal I think
2021-07-02 09:29:39 <Sheila> not that I'm aware of, but perhaps someone else would know.
2021-07-02 09:30:27 <Sheila> well. horizon, maybe, but I'm not sure it does HDD images. disc images and rootfs tarballs, sure, but…
2021-07-02 09:31:55 <Ristovski> Oh hmm, will look into it
2021-07-02 09:35:05 <Ristovski> Sheila: wait nvm, alpine-make-vm-image already bootstraps, I suppose I could just customize it further
2021-07-02 12:12:49 <MY-R> przemoc: not only you who didnt notice that "wiki" link which imho should be better exposed and not "hiden" in middle of release page between many other links
2021-07-02 15:01:33 <markand> are pipeline posting packages somewhere?
2021-07-02 15:01:50 <markand> because it would help me not burning my raspberry pi when building large packages :D
2021-07-02 16:36:34 <minimal> Ristovski: you might be interested in https://github.com/dermotbradley/create-alpine-disk-image
2021-07-02 16:36:40 <minimal> its a work-in-progress
2021-07-02 17:01:17 <Ristovski> minimal: oh nice, thanks!
2021-07-02 17:02:11 <minimal> Ristovski: must push an update to the git repo later today - have reworked quite a lot of stuff.
2021-07-02 17:04:28 <Ristovski> minimal: oh fancy, it supports encryption?
2021-07-02 17:05:47 <minimal> Ristovski: the encryption stuff is what I've been working on the past 2 weeks - so its not up-to-date in the repo. Let me push a revised version.
2021-07-02 17:06:54 <minimal> it supports full encryption (including /boot) with Grub and all-but-boot encryption with Syslinux and RPI. Having probs with LUKS v2 currently so its all using LUKS v1 at present.
2021-07-02 17:07:05 <Ristovski> minimal: I've been meaning to write my own custom `init` (or maybe I can make it into an openrc service?) that can read a decryption password from a virtio socket (passed by firecracker)
2021-07-02 17:07:59 <minimal> Ristovski: have been working on remote unlocking of LUKS via Dropbear SSH but that's not ready yet
2021-07-02 17:08:04 <Ristovski> oh nice
2021-07-02 17:08:24 <Ristovski> seems like i'm not the only one with these niche needs lol
2021-07-02 17:09:03 <minimal> its based on an Ansible playbook I've been using for about a year for all my RPIs, physical machines and VMs
2021-07-02 17:35:51 <minimal> Ristovski: just pushed latest changes to the repo
2021-07-02 17:37:53 <Ristovski> minimal: Very nice! Will definitely check out
2021-07-02 17:39:11 <minimal> currently I can only vouch for QEMU & Virtualbox VMs working as that's what I routinely test. Need to find time to test the rest
2021-07-02 17:41:22 <Ristovski> I will be quite busy the next few weeks, but I could probably try and get firecracker integration working when I (finally) get free time
2021-07-02 17:42:04 <minimal> Have been meaning to look at firecracker for a while but have too many other things on the go :-)
2021-07-02 17:45:47 <Ristovski> I used to use qemu with microvm (https://github.com/qemu/qemu/blob/master/docs/system/i386/microvm.rst) but then realized firecracker is more lightweight than having a full qemu install (and qemu is annoying to strip out), its literally a 3.8MB static binary and has all the things I need (including nice stuff like virtio-vsock and other virtio goodies)
2021-07-02 17:48:11 <Ristovski> It also manages to boot my minimal kernel a bit faster than qemu, wasn't able to figure that one out, the configuration is /identical/. But I don't hate myself enough to sit down and figure out why that happens :P
2021-07-02 17:49:04 <minimal> in my script, for VMs I tweak the initramfs to (try) and have only the needed drivers for stuff like virtio. I've built a local Alpine package for a more stripped down linux-virt package (removing drivers not relevant for VMs), need to find the time to raise a PR to try and get some or all of the changes included.
2021-07-02 17:49:59 <minimal> Ristovski: doesn't Firecracker have a stripped down BIOS (e.g. does it even have PCI?) so obviously the kernel init would be quicker if the BIOS supports less functionality
2021-07-02 17:51:09 <Ristovski> minimal: it should be nearly identical with `microvm`, I even run qemu with 'pit=off,pic=off,rtc=off', dmesg output is nearly identical (different order of some memory regions etc)
2021-07-02 17:57:47 <minimal> Ristovski: Firecracker apparently doesn't have a BIOS at all (whereas QEMU does) so that would make some sort of difference in startup times
2021-07-02 17:58:07 <Ristovski> Indeed, I suppose that could be it
2021-07-02 19:21:02 <Hello71> if you want microvm to be extra fast you need -bios qboot.bin
2021-07-02 19:40:26 <minimal> Hello71: yeah, true, however firecracker claims not to have any BIOS at all so I'd expect it to be faster (whether measurable or not) than microvm at startup as a result
2021-07-02 20:23:33 <AppAraat[m]> Are rng-tools required for a Pi ZW? https://wiki.alpinelinux.org/wiki/Raspberry_Pi_Zero_W_-_Installation#Create_cache_folder_and_add_rng-tools_packages
2021-07-02 20:24:53 <AppAraat[m]> "I decided to not install the rngd related packages at this point to see how a reboot looked, answer is not good. The dhcp request just times out. Running setup-alpine again at this point also doesn't work." - Is the rng on the Pi _so_ bad? :S
2021-07-02 20:27:37 <c705> it's no bother to install and run haveged, which I tend to do anyways
2021-07-02 20:29:43 <Hello71> haveged adds dubious security
2021-07-02 20:30:03 <AppAraat[m]> instead of rng-tools? I'd favor that for the fact it only has 1 dependency instead of 3 like with rng-tools
2021-07-02 20:30:34 <AppAraat[m]> but I'm very much anything except an expert on crypto
2021-07-02 21:13:07 <c705> haveged is obsolete since 5.6 anyways apparantely
2021-07-02 21:22:26 <AppAraat[m]> TIL
2021-07-02 21:23:36 <minimal> AppAraat: rng-tools supports both the HWRNG and also jitterentropy-rng
2021-07-02 21:23:42 <minimal> so no need for haveged
2021-07-02 21:24:46 <AppAraat[m]> but yes need for rng-tools
2021-07-02 21:25:18 <AppAraat[m]> (I'm still getting used to the fact of adding packages I want to the sdcard before booting it and connecting to the wifi :p)
2021-07-02 21:26:48 <AppAraat[m]> In any case, this guide might be out of date: https://wiki.alpinelinux.org/wiki/Raspberry_Pi_Zero_W_-_Installation - since new instructions are here: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12353#note_164623
2021-07-02 21:27:16 <AppAraat[m]> This guide already seems to reflect that: https://wiki.alpinelinux.org/wiki/Raspberry_Pi_-_Headless_Installation
2021-07-02 21:28:01 <minimal> well you need something to keep feeding entropy, rngd will feed from both the RPI's hwrng and from jitterentropy-rng (which is similar to haveged), plus it supports other entropy sources if you have them
2021-07-02 21:29:16 <AppAraat[m]> I'm still wondering why that isn't possible OOTB on a Pi ZW, but as I said I am a huge crypto noob.
2021-07-02 21:29:32 <minimal> AppAraat: "cat /proc/sys/kernel/random/entropy_avail" will give you an idea of how much entropy is available at any point in time
2021-07-02 21:29:34 <AppAraat[m]> actually, scratch crypto :p
2021-07-02 21:29:58 <ikke> These devices do not have a lot of sources of entropy
2021-07-02 21:30:20 <minimal> AppAraat: out of the box? well you need something to keep feeding entropy into /dev/random, that's what rngd (or haveged) do
2021-07-02 21:31:28 <AppAraat[m]> but where do they then get it from? :D
2021-07-02 21:31:50 <AppAraat[m]> my laptop is hovering around 4k entropy btw, is that good?
2021-07-02 21:31:52 <minimal> ikke: kernel gets some entropy from network activity and keyboard activity, plus RPI has its HWRNG. Obviously if you're using a RPI as a headless server then there will no no keyboard activity
2021-07-02 21:32:14 <minimal> Apparaat: as a rule of thumb anything above 1000 is "good"
2021-07-02 21:33:13 <ikke> minimal: event the zero?
2021-07-02 21:33:27 <AppAraat[m]> ok then, my alpine VM is around 125 lol
2021-07-02 21:33:39 <minimal> now if you're using a lot of crypto stuff that can affect available entropy. Ideally you want rngd running early on first boot, *before* you host's SSH keys are created
2021-07-02 21:34:10 <minimal> ikke: dunno about the Zero, never used one. I'm think of the RPI 2/3/4
2021-07-02 21:34:46 <minimal> AppAraat: alpine VM running on QEMU/KVM? If so then you should enable the virtio rng so it gets entropy from the host machine
2021-07-02 21:35:29 <AppAraat[m]> yes, QEMU/KVM. But wouldn't my entropy also increase if I install rng-tools?
2021-07-02 21:35:40 <AppAraat[m]> (instead of virtio rng)
2021-07-02 21:36:49 <minimal> well its a *virtual* machine so for jitterentropy-rng/haveged which rely on the slighty-unpredictable behaviour of physical CPUs a virtual CPU will be less unpredicatible
2021-07-02 21:37:42 <minimal> so run qemu with "-object rng-random,id=rng0 -device virtio-rng-pci,rng=rng0,max-bytes=1024,period=1000"
2021-07-02 21:37:53 <minimal> and load the virtio_rng module in the VM
2021-07-02 21:39:20 <minimal> On Alpine rng-tools "out-of-the-box" takes its entropy from 2 sources: jitterentropy and HWRNG - if you use those qemu options then you'll have a /dev/hwrng, if you don't use them then you won't and so will only have 1 source
2021-07-02 21:39:24 <AppAraat[m]> ah ok so it's also kernel mod that needs to be loaded by the guest?
2021-07-02 21:39:43 <minimal> I'm the Alpine rng-tools package maintainer BTW :-)
2021-07-02 21:40:03 <AppAraat[m]> oh nice, I was checking it out just now :p
2021-07-02 21:41:50 <minimal> AppAraat: actually its slightly confusing - the jitterentropy-rng kernel module is used for the kernel's own "special" (early init?) pool, whereas rngd uses libjitterentropy which effectively has the same code as the kernel module but rngd uses it to then top-up the normal entropy pool
2021-07-02 21:42:38 <AppAraat[m]> I'm using virt-manager, and I've noticed in "Add hardware" I can add the hosts /dev/urandom or /dev/random btw
2021-07-02 21:43:09 <AppAraat[m]> (or any other custom rng daemon exposed as a /dev/ file)
2021-07-02 21:44:20 <minimal> have a read of https://fedoraproject.org/wiki/Features/Virtio_RNG
2021-07-02 21:45:55 <AppAraat[m]> "entropy starvation" is _such_ a weird concept to wrap my head around.
2021-07-02 21:46:15 <AppAraat[m]> like... how on earth can you run out of randomness ???
2021-07-02 21:46:23 <minimal> AppAraat: not sure about virt-manager, I tend to use qemu cli directly
2021-07-02 21:46:46 <AppAraat[m]> yeah that's one of my goals too
2021-07-02 21:47:21 <AppAraat[m]> actuall virtsh but it's kind of light on guides
2021-07-02 21:47:27 <minimal> because there are different types of RNGs - a PRNG (pseudo-RNG) is not really random, its "sort of" random one its been seeded, the entropy pool is used to seed it
2021-07-02 21:48:24 <minimal> also if you run a lot of crypto stuff (VPN or SSL or SSH) and check the available entropy you'll see it drop over time if you don't have enough entropy
2021-07-02 21:49:40 <minimal> AppAraat: did I mention before my script to build Alpine disk images? :-)
2021-07-02 21:49:58 <AppAraat[m]> entropy is both mysterious and cool
2021-07-02 21:50:09 <AppAraat[m]> minimal: no but I am very interested :D
2021-07-02 21:50:30 <minimal> Apparaat: https://github.com/dermotbradley/create-alpine-disk-image
2021-07-02 21:50:50 <minimal> for now only properly tested with QEMU and Virtualbox VMs
2021-07-02 21:51:18 <minimal> it will set up all the RNG-related stuff
2021-07-02 21:51:59 <AppAraat[m]> wow sweet :D
2021-07-02 21:53:19 <AppAraat[m]> so using this you can basically create an image which is ready to yeet over to an sdcard which can be used in a Pi directly?
2021-07-02 21:54:02 <AppAraat[m]> unfortunately no armhf yet, but the concept is super cool
2021-07-02 21:56:49 <minimal> AppAraat: yeah I previously used an Ansible playbook to create disk image to "dd" onto SD cards for my RPIs, this script is intended to do the same, I just haven't had time to focus on completing the RPI work. What would you use armhf with? Its not much work, in theory, for me to add other arches
2021-07-02 21:57:22 <AppAraat[m]> for the Pi ZW
2021-07-02 21:57:48 <AppAraat[m]> but then again for learning purposes I want to create my own script which does that first
2021-07-02 21:57:57 <ikke> hmm, can the wireless adapter act as a source of entropy?
2021-07-02 21:58:03 <minimal> plus it uses cloud-init so with the physical machine images there's a small partition containing the cloud-init YAML files so you can either those before 1st boot to define hostname/network details/timezone/etc and it gets autoconfigured then during 1st boot
2021-07-02 21:58:03 <ikke> sounds like a good source
2021-07-02 21:58:23 <minimal> s/either/edit/g
2021-07-02 21:58:23 <alpine-meetbot> minimal meant to say: plus it uses cloud-init so with the physical machine images there's a small partition containing the cloud-init YAML files so you can edit those before 1st boot to define hostname/network details/timezone/etc and it gets autoconfigured then during 1st boot
2021-07-02 21:58:57 <AppAraat[m]> that's pretty cool
2021-07-02 21:59:17 <AppAraat[m]> is it also possible to define which packages get added? As in, not downloaded but transferred from local disk?
2021-07-02 21:59:31 <AppAraat[m]> ikke: or perhaps even a mic if it's available?
2021-07-02 21:59:37 <minimal> ikke: vague think I've seen someone doing that a while ago. rngd does support cheap USB SDR (soft-defined radio) devices where you tune them to a unused frequency as the "static" noise is somewhat random
2021-07-02 22:00:01 <ikke> right
2021-07-02 22:01:19 <minimal> AppAraat: with cloud-init you mean? Sure, in the user-data.yaml file you can use the "packages:" section to list what additional stuff you want added
2021-07-02 22:01:56 <AppAraat[m]> oh I just meant in general, but that's pretty cool yeah
2021-07-02 22:02:22 <minimal> ikke: Cloudflare (still I think) have a wall of Lava Lamps on shelves in their HQ's lobby with webcam(s) pointing to them as Lava Lamps are a good source of entropy (the hot wax inside moves in unpredictable ways)
2021-07-02 22:02:34 <ikke> yeah, that's pretty cool (or hot0
2021-07-02 22:02:44 <AppAraat[m]> lol
2021-07-02 22:04:25 <minimal> radio static includes some "solar system noise", I'd assume in theory WiFi unused channels might too but there's probably more chance of real traffic appearing on any WiFi channel
2021-07-02 22:06:23 <minimal> "rngd -l" shows the possible entropy sources (note there's 2 alternative Alpine packages "rng-tools" (with a small set) and "rng-tools-extra" (with things like the SDR support))
2021-07-02 22:12:03 <minimal> ikke: there was a guy in NL who used to write several small daemons for getting entropy from different devices (e.g. sound card input, webcams). With a cheap-n-nasty webcam if you block the lens (so it can't see anything) and turn up the gain the chip "tries" to make out an image from the darkness and they result is a degree of entropy)
2021-07-02 22:14:06 <AppAraat[m]> oh that's pretty cool
2021-07-02 22:14:30 <AppAraat[m]> I bet you could do the same with mic background hum
2021-07-02 22:15:15 <AppAraat[m]> https://wiki.alpinelinux.org/wiki/Raspberry_Pi_-_Headless_Installation#1_-_Prepare_the_SD_card - "Use the a command to toggle the partition as bootable." - This is meaning the first partition in this case, correct?
2021-07-02 22:18:21 <minimal> AppAraat: its the small FAT partition, typically the 1st one (not sure if it must always be the 1st)
2021-07-02 22:18:57 <AppAraat[m]> ah ok, I'll add that to that wiki article later.
2021-07-02 22:19:08 <AppAraat[m]> yeah it needs to be FAT in order to boot apparently
2021-07-02 22:19:26 <AppAraat[m]> #justweirdpithings
2021-07-02 22:32:00 <minimal> AppAraat: hopefully I'll find time to test my script for RPI images in a week or so
2021-07-02 22:33:31 <AppAraat[m]> oh that's ok, right now I'm just messing around with it anyway
2021-07-02 22:34:24 <AppAraat[m]> I'll try to follow the "manual" methods first and then transition to using more automation tools (like your image builder)
2021-07-02 22:39:22 <AppAraat[m]> btw, is there a way to download packages from https://pkgs.alpinelinux.org/ ?
2021-07-02 22:40:29 <minimal> what do you mean exactly? you want to store the files locally somewhere?
2021-07-02 22:41:06 <AppAraat[m]> yeah, in order to include them in the /boot/cache in order to install them after first boot on the Pi ZW
2021-07-02 22:41:20 <AppAraat[m]> https://wiki.alpinelinux.org/wiki/Raspberry_Pi_Zero_W_-_Installation#Create_cache_folder_and_add_rng-tools_packages
2021-07-02 22:42:10 <AppAraat[m]> basically like a web based apk fetch :p
2021-07-02 22:45:57 <minimal> well you could manually fetch the files from one of the mirrors (https://mirrors.alpinelinux.org/) but then you'd also have to figure out the dependancies and also fetch them
2021-07-02 22:46:38 <AppAraat[m]> just the dependencies? Or also dependencies of dependencies etc. etc.
2021-07-02 22:47:33 <minimal> yupe :-) you'd have to do the work manually that "apk add" automatically figures out lol
2021-07-02 22:47:51 <AppAraat[m]> :O
2021-07-02 22:48:34 <AppAraat[m]> oh in case of rng-tools it's pretty straight-forward. Both jitterentropy-library and libcrypto1.1 depend on musl, which should be installed already by default.
2021-07-02 22:49:20 <minimal> BTW in that article, just above the section you referenced, it talks about entries to put into usercfg.txt. I seem to remember that "gpu_mem=" *only* works if its in config.txt file, I could be wrong but I think thats the case
2021-07-02 22:49:31 <Hello71> c705: not exactly but pretty close
2021-07-02 22:51:16 <Hello71> minimal: actually it's more confusing than that, there's mini-jitterentropy (my name) which is used always, crypto/jitterentropy which is used for i think internal users (ipsec?) and userspace
2021-07-02 22:53:25 <AppAraat[m]> minimal: Seems so yeah: https://www.raspberrypi.org/documentation/configuration/config-txt/memory.md
2021-07-02 22:57:08 <AppAraat[m]> ah, in config.txt it says in last line `include usercfg.txt`
2021-07-02 22:57:52 <AppAraat[m]> In fact: https://paste.debian.net/plain/1203193
2021-07-02 22:59:03 <minimal> Hello72: yeah and the guy who wrote jitterentropy-rng also wrote libjitterentropy and jitterentropy-rngd (userspace daemon). I used to maintain jitterentropy-rngd until I enabled the jitter support in rng-tools. I noticed that (at least then) his jitterentropy-rngd did *not* depend on libjitterentropy but rather he'd pasted basically the same library code into the jitterentropy-rngd source :-(
2021-07-02 23:00:21 <minimal> AppAraat: yes config.txt includes usercfg.txt but my point was that I *think* the cpu setting MUST be in config.txt or else it is not acted upon (somnething to do with the bootloader can only set it early in boot before it loads any included files)
2021-07-02 23:00:30 <minimal> s/cpu/gpu/g
2021-07-02 23:00:30 <alpine-meetbot> minimal meant to say: AppAraat: yes config.txt includes usercfg.txt but my point was that I *think* the gpu setting MUST be in config.txt or else it is not acted upon (somnething to do with the bootloader can only set it early in boot before it loads any included files)
2021-07-02 23:03:26 <AppAraat[m]> hmm ok, well in any case I'll hopefully find out soon enough once I finally boot this thing.
2021-07-02 23:03:52 <AppAraat[m]> it shouldn't really matter all too much considering the RAM usage of Alpine :p
2021-07-02 23:04:49 <AppAraat[m]> I wonder why that guide also mentions downloading `sysfsutils`
2021-07-02 23:05:11 <AppAraat[m]> it's definitely not a dependency of `rng-tools`
2021-07-02 23:11:05 <minimal> AppAraat: from memory I think it was a dependancy but the rng-tools author rewrote the bit of code (4 or 5 months ago?) to remove the need for sysfsutils
2021-07-02 23:11:38 <AppAraat[m]> oh cool, in that case I'll remove that from that article.
2021-07-02 23:12:10 <AppAraat[m]> omg... I'm downloading all kinds of stuff like wpa_supplicant and its deps but apparently there's already lot of stuff in the Pi armhf image: https://paste.debian.net/plain/1203194
2021-07-02 23:12:29 <minimal> AppAraat: yeah 6.12 removed that: https://github.com/nhorman/rng-tools/releases/tag/v6.12
2021-07-02 23:18:14 <AppAraat[m]> minimal: IMO rng-tools should also get added to the Pi armhf image.
2021-07-02 23:18:32 <AppAraat[m]> it just makes much more sense
2021-07-02 23:20:16 <minimal> true. Don't have any armhf devices here so haven't looked at it. Smallest RPI I have is an RPI2 :-)
2021-07-02 23:21:35 <AppAraat[m]> oh you should check out the Pi ZW then, it's so deliciously tiny! And super power efficient too. I can power one with my 20 Ah powerbank for like a week or so.
2021-07-02 23:23:32 <minimal> I'm more into RPIs with POE as network servers
2021-07-02 23:24:31 <AppAraat[m]> yeah IIRC for that you'd need to add a PoE adapter to it, increasing the cost.
2021-07-02 23:24:59 <AppAraat[m]> (like... considerably even, almost double the amount of a Pi ZW)
2021-07-02 23:25:08 <minimal> plus an RTC
2021-07-02 23:25:21 <AppAraat[m]> RTC = real time clock?
2021-07-02 23:25:24 <minimal> yupe
2021-07-02 23:25:55 <AppAraat[m]> oh I never really thought about the purpose of it, why are you using it?
2021-07-02 23:28:48 <minimal> I use RPIs as homebuild "blades" - each mounted vertically on a board and connected to a POE switch to run network services such as DHCP, DNS, Radius, Docker, etc
2021-07-02 23:34:18 <AppAraat[m]> oh cool, like a Pi cluster?
2021-07-02 23:34:35 <AppAraat[m]> and you sync clocks among them using an RTC/
2021-07-02 23:34:38 <AppAraat[m]> ?
2021-07-02 23:36:32 <minimal> the RTCs are to keep accurate time (even if you sync swclock with NTP when you reboot the PIs clock goes back to 1970 during boot until it regains NTP sync)
2021-07-02 23:37:27 <minimal> I actually use some RPIs as my NTP servers for the local network (have a RPI with GPS hat for getting time)
2021-07-02 23:37:58 <minimal> some of the RPIs are in a cluster but some of them are just standalone servers
2021-07-02 23:40:14 <AppAraat[m]> ah yeah I've noticed that clock reset on the pi
2021-07-02 23:40:29 <AppAraat[m]> so that's what an RTC can be used for!
2021-07-02 23:41:09 <minimal> you can get RTCs for RPI for about $5/£5
2021-07-02 23:42:48 <AppAraat[m]> yea but for the purposes of my usage I don't think I'll be needing one any time soon :p
2021-07-02 23:43:27 <AppAraat[m]> also... first boot, yay! https://paste.debian.net/1203198/
2021-07-02 23:43:37 <AppAraat[m]> I must say, alpine is booting incredibly fast, like wow.
2021-07-02 23:43:49 <AppAraat[m]> (I mean, compared to Raspbian)
2021-07-02 23:53:07 <AppAraat[m]> interesting, I think I can hear some noise from the Pi ZW during disk activity. Maybe it's capacitor whine or something.
2021-07-02 23:55:16 <minimal> disk activity? or SDcard activity?
2021-07-02 23:56:18 <AppAraat[m]> yeah I meant SD
2021-07-02 23:57:13 <minimal> notice the "clock skew detected" message in your boot output? That's because of the absence of an RTC
2021-07-02 23:58:21 <AppAraat[m]> nice, it hasn't even booted up and it's already skewing lol
2021-07-03 00:00:45 <minimal> yeah its due to an OpenRC "fudge" - immediately at power on the clock defaults to 1970. When OpenRC runs it checks if the time if older than the timestamp of one of its files (that it "touch"s on shutdown) and if so gives that skew message.
2021-07-03 00:01:32 <AppAraat[m]> aah, that's pretty smart.
2021-07-03 00:01:48 <minimal> i.e. it spots if time has gone backwards
2021-07-03 00:02:38 <AppAraat[m]> yeah but I guess in this case timestamps were preserved in the FS somehow because it couldn't touch anything on shutdown since it was the first boot.
2021-07-03 00:03:06 <AppAraat[m]> which would mean that timestamps get preserved in FAT32
2021-07-03 00:03:24 <minimal> then (from memory) it sets the system time to the date/time of that file......which is sort of fine if you just rebooted as the clock will be only 1 or 2 mins slow - but if you shutdown the RPI 1 week ago and turn it on now then the time will be 1 week wrong..............until you run NTP client to update the clock
2021-07-03 00:04:28 <minimal> AppAraat: its the date/time of the relevant file, i.e. when you do an "ls -l" anywhere you see the last changed date/time of files, it uses one of its own files at shutdown to "tag" that date/time
2021-07-03 00:05:51 <AppAraat[m]> yeah but in this case how could touch a file upon shutdown if it hasn't been shutdown before?
2021-07-03 00:06:03 <AppAraat[m]> *could it touch a file
2021-07-03 00:06:44 <minimal> right, but that file already has a date/time from when the Alpine image was built
2021-07-03 00:07:49 <AppAraat[m]> yeah that's what I meant, and as such I today learned that FAT32 apparently preserves this date/time for files.
2021-07-03 00:47:05 <c705> what's the easiest way to add x-bash-completion for each x in /etc/apk/world if x-bash-completion exists? 
2021-07-03 01:41:09 <Hello71> apk add bash-completion
2021-07-03 04:48:14 <kevint> I booted up my laptop today and after entering grub I get: "mount: mounting /dev/mapper/vg0-root on /sysroot failed: I/O error." And then I get launched into an initramfs emergency shell. My laptop was booting fine yesterday. Anyone have any advice for how to proceed?
2021-07-03 06:32:36 <AppAraat[m]> minimal: hmm, despite starting rngd, entropy is still 10x smaller it seems: https://paste.debian.net/1203213/
2021-07-03 06:36:17 <AppAraat[m]> And at some point the setup-alpine actually stops rngd: https://paste.debian.net/1203214/
2021-07-03 07:13:55 <AppAraat[m]> If there's not sufficient entropy, does the process of generating cryptography simply take longer, or is it also inherently less secure when finally generated?
2021-07-03 08:28:18 <kevint> I'm getting some kind of weird circular conflict with util-linux-misc: https://paste.debian.net/1203217/ anyone know how to fix this?
2021-07-03 11:52:00 <Aftershock> hi I've been trying to install alpine on a raspberry pi and I've run into some issues. I was wondering if the current headless guide is working. My issues occur when the setup-alpine runs and tries to install to the disk
2021-07-03 12:02:49 <clandmeter> We do not support install to disk
2021-07-03 12:02:53 <clandmeter> On rpi
2021-07-03 12:07:07 <Aftershock> https://wiki.alpinelinux.org/wiki/Raspberry_Pi_-_Headless_Installation
2021-07-03 12:07:12 <Aftershock> does that work?
2021-07-03 12:07:19 <Aftershock> using sys mode?
2021-07-03 12:07:28 <ikke> People have managed to get disk setup to work
2021-07-03 12:07:46 <Aftershock> how would you recommend running a rpi? on a different os?
2021-07-03 12:12:18 <ikke> Aftershock: what is the exact issue you are running into?
2021-07-03 12:12:39 <Aftershock> I'm just retrying now
2021-07-03 12:13:05 <Aftershock> it says ERROR tempory issue please try again later
2021-07-03 12:13:11 <Aftershock> or something similar
2021-07-03 12:13:16 <ikke> That's a network / dns issue
2021-07-03 12:13:23 <Aftershock> when it install the file system to disk
2021-07-03 12:13:36 <Aftershock> I'm manually selecting mirrors this time
2021-07-03 12:13:50 <Aftershock> I can install things with apk add like normal
2021-07-03 12:39:28 <Aftershock> can I mkfs.ext4? or do I need to install an apk for that file system type
2021-07-03 12:39:41 <yyp> apk add e2fsprogs
2021-07-03 12:39:54 <Aftershock> yo thanks
2021-07-03 12:41:14 <ikke> note that the boot partition needs to be fat
2021-07-03 12:48:43 <Aftershock> do you have an example of cmdline.txt 
2021-07-03 12:49:00 <Aftershock> under /media/mmcblk0p1/
2021-07-03 12:50:59 <yyp> modules=loop,squashfs,sd-mod,usb-storage quiet console=tty1
2021-07-03 12:53:03 <Aftershock> root=mmcblk0p2
2021-07-03 12:53:04 <Aftershock> ?
2021-07-03 12:53:36 <Aftershock> should I put a space after tty1 before that
2021-07-03 12:55:57 <yyp> Yes
2021-07-03 12:56:08 <yyp> also it should be root=/dev/mmcblk0p2
2021-07-03 12:56:18 <Aftershock> holy
2021-07-03 12:56:25 <Aftershock> that was it
2021-07-03 12:56:25 <Aftershock> ok
2021-07-03 12:56:35 <yyp> Are you doing a sys mode install?
2021-07-03 12:56:43 <Aftershock> yeah
2021-07-03 12:56:52 <Aftershock> following this
2021-07-03 12:56:53 <Aftershock> https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi
2021-07-03 12:57:02 <Aftershock> first time messing with alpine
2021-07-03 12:57:06 <yyp> It's kinda broken
2021-07-03 12:57:08 <Aftershock> mostly an archie boi
2021-07-03 12:57:18 <yyp> You need to format in fat32 iirc otherwise it won't boot
2021-07-03 12:57:27 <Aftershock> mine seems to boot
2021-07-03 12:57:37 <yyp> huh
2021-07-03 12:57:37 <Aftershock> after fixing that cmdline.txt
2021-07-03 12:57:45 <Aftershock> I have a vfat boot
2021-07-03 12:58:02 <Aftershock> hmm actually though p2 is boot
2021-07-03 12:58:20 <Aftershock> so think its going straight off ext4
2021-07-03 12:59:14 <Aftershock> how would I tell if my system is currently booting off the ext4 sys partition
2021-07-03 12:59:24 <Aftershock> or if it is in diskless mode?
2021-07-03 12:59:37 <yyp> lsblk?
2021-07-03 13:00:00 <yyp> It should show if the sys partition is mounted to /
2021-07-03 13:01:30 <Aftershock> mmcblk0p2 179:2    0 29.3G  0 part /
2021-07-03 13:01:38 <Aftershock> that's it right?
2021-07-03 13:02:00 <yyp> Yep
2021-07-03 13:02:09 <Aftershock> alright 
2021-07-03 13:02:14 <Aftershock> feeling good
2021-07-03 13:02:15 <Aftershock> ahahah
2021-07-03 13:02:26 <Aftershock> thanks for the help appreciate it
2021-07-03 13:03:13 <yyp> Does Wi-Fi work after sys install?
2021-07-03 13:03:22 <Aftershock> I'm using eth
2021-07-03 13:03:25 <yyp> :(
2021-07-03 13:03:29 <Aftershock> but I believe it should
2021-07-03 13:04:11 <yyp> At least, does `ifconfig wlan0` work?
2021-07-03 13:04:51 <Aftershock> sick it has doas
2021-07-03 13:05:04 <Aftershock> error fetching interface information: Device not found
2021-07-03 13:05:06 <Aftershock> :/
2021-07-03 13:05:16 <yyp> I had the same :(
2021-07-03 13:05:28 <Aftershock> but I think I need to enable wifi in usercfg.txt
2021-07-03 13:05:31 <Aftershock> let me check
2021-07-03 13:08:51 <Aftershock> so I can clone this ssd so I don't need to set this up again on like 12 other rpis?
2021-07-03 13:11:08 <ikke> Aftershock: should be possible, yes
2021-07-03 13:14:53 <yyp> There doesn't seem to be an option to enable wifi
2021-07-03 13:15:43 <ikke> I think you need to install some broadcom drivers
2021-07-03 13:15:58 <yyp> Maybe the alpinelinux.org/downloads image is somehow different from the one you get with setup-disk?
2021-07-03 13:17:45 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.arm.sh
2021-07-03 13:20:31 <Aftershock> yeah I got lost a bit
2021-07-03 13:20:54 <Aftershock> I was hoping to be able to set up a rpi as a hotspot/wifi extender
2021-07-03 13:21:02 <Aftershock> maybe can't with alpine
2021-07-03 13:21:31 <Aftershock> do you think alpine is a good choice for servers?
2021-07-03 13:21:34 <Naib> should be able to
2021-07-03 13:21:55 <Aftershock> I run many servers and previously ran them all on fedora
2021-07-03 13:22:11 <Aftershock> but looking to switch
2021-07-03 13:22:33 <ikke> Don't you need 2 wifi adapters for an extender?
2021-07-03 13:22:38 <Aftershock> tossing between alpine/rocky/void/arch
2021-07-03 13:22:54 <Aftershock> ikke: I don't know
2021-07-03 13:23:25 <Aftershock> I haven't tried it yet, my brother was tring to set one up 
2021-07-03 13:23:40 <Aftershock> he normally uses his laptop as an extender 
2021-07-03 13:23:46 <Naib> why arch?  arch on a server is ... risky
2021-07-03 13:23:53 <Naib> they are a bit too bleeding edge
2021-07-03 13:23:54 <Aftershock> like using ethernet plug
2021-07-03 13:24:01 <ikke> I've been running arch on a server for years
2021-07-03 13:24:27 <Naib> i did as well... until their "upgrade" policy screwed everything
2021-07-03 13:24:34 <Aftershock> yeah I mean as much as there is errors that you can get usually they only need a bit of manual fixing
2021-07-03 13:24:49 <Aftershock> but yeah I kind like rocky as it is very solid
2021-07-03 13:24:56 <Aftershock> centos
2021-07-03 13:25:42 <Aftershock> only problem is rpm is probably not as good as deb/pacman
2021-07-03 13:26:04 <Naib> im waiting to see what Ansys and co do w.r.t. centos vs rocky. but whatever the direction for the host system, the 10+ containers for webapps will be alpine :)
2021-07-03 13:26:28 <Aftershock> I was thinking of using a alpine kvm host
2021-07-03 13:26:35 <Aftershock> is that a dumb idea?
2021-07-03 13:26:43 <ikke> I don't think so
2021-07-03 13:37:04 <yyp> Looks like setup-disk is missing some files in / like bcm2837-rpi-3-b.dtb
2021-07-03 13:37:21 <yyp> Probably that's why wifi didn't work
2021-07-03 13:39:55 <Aftershock> hmm yeah ok
2021-07-03 13:57:14 <minimal> kevint: yes the util-linux-misc issue was a side-effect of my MR to create the util-linux-misc subpackage. I did raise another MR, !22651, a week ago to fix this but it ha not been merged yet.
2021-07-03 13:58:07 <Aftershock> welllllllll searx doesn't like alpine 
2021-07-03 13:58:12 <Aftershock> damn
2021-07-03 14:00:40 <minimal> Aftershock: yes Alpine makes a great server OS. I have a script that I'm working on for building Alpine server images so that you just "dd" the created image onto disk or SDcard. RPI support in not tested yet. Its at https://github.com/dermotbradley/create-alpine-disk-image
2021-07-03 14:01:08 <Aftershock> nice
2021-07-03 14:01:37 <Aftershock> have you run it as a kvm host for multiple virtual machines?
2021-07-03 14:04:14 <minimal> Aftershock: I've been focusing more on containers but have a partly-finished OS config for running Nomad which can manage both containers and VMs
2021-07-03 14:07:52 <minimal> AppAraat: with your low entropy did you check which sources rngd was using? "rngd -l" should show you which are enabled. Also BTW cat-ting /proc/sys/kernel/random/entropy_avail too often can itself reduce entropy a little as whenever any binary is run (such as "cat") the ASLR (address randomisation) security feature uses a small amount of entropy
2021-07-03 14:11:57 <bananicorn> Hello :)
2021-07-03 14:13:52 <bananicorn> I am currently figuring out how to build a custom image that includes eg. the ZFS kernel module so that I can install alpine's root on ZFS. I know the Extended ISO exists, but I am using this as an opportunity to learn how to build a custom image while I am at it and have my own selection of pre-installed APKs.
2021-07-03 14:14:39 <ikke> bananicorn: look at the scripts dir in aports
2021-07-03 14:14:46 <ikke> it contains the scripts that builds the images
2021-07-03 14:14:50 <bananicorn> I have written my own mkimg.zfs.sh file already
2021-07-03 14:14:56 <ikke> ok
2021-07-03 14:15:12 <bananicorn> I am struggling with configuring it correctly though, I seem to be unable to load the ZFS kernel module
2021-07-03 14:15:44 <bananicorn> Perhaps I am forgetting something, since the mkimage.sh guide and the extended alpine iso contains a bunch of other stuff too
2021-07-03 14:16:03 <bananicorn> When running "modprobe zfs" I get "module zfs not found in modules.dep"
2021-07-03 14:16:07 <ikke> can you show what you have so far?
2021-07-03 14:16:16 <bananicorn> I'll show you a paste of my mkimg.zfs.sh
2021-07-03 14:17:32 <bananicorn> https://bin.disroot.org/?5895919613def196#2vXnqqFTGRQKBV4cxnmQNaWVuJpVLDuAKYJzESB1HN3m
2021-07-03 14:17:38 <bananicorn> Thanks for taking the time ikke :)
2021-07-03 14:19:07 <bananicorn> I did "apk add zfs e2fsprogs" on the booted system, do I need to add e2fsprogs to the mkimg profile too? Not sure if that is the issue
2021-07-03 14:21:47 <ikke> bananicorn: you need to install zfs-lts as well
2021-07-03 14:21:52 <ikke> zfs is just userspace
2021-07-03 14:21:57 <ikke> zfs-lts contains the kernel modules
2021-07-03 14:22:10 <bananicorn> You mean I need to do that on the booted system?
2021-07-03 14:22:40 <ikke> bananicorn: You mean booted from the image?
2021-07-03 14:22:52 <bananicorn> Yeah
2021-07-03 14:23:14 <bananicorn> Or where do I need to add zfs-lts?
2021-07-03 14:23:39 <ikke> Oh, I see mkimage.base.sh already takes care of that
2021-07-03 14:24:00 <bananicorn> If I do "apk add zfs-lts" on the booted image, I just get a whole lot of "file not found" errors and some others too
2021-07-03 14:24:13 <bananicorn> and a lot of "failed to rename x to x"
2021-07-03 14:24:32 <bananicorn> and some errors relating to /lib/modules being read-only
2021-07-03 14:24:45 <ikke> yeah, that makes sense, modloop is read-only
2021-07-03 14:27:37 <bananicorn> I am not sure what is missing either way, it would be great if you have any other ideas. I have tried to use the "extended" profile and the mkimage.sh wiki page as references (since they also contain zfs) but as I mentioned, I am quite confused about what is needed and what is unrelated
2021-07-03 14:28:44 <ikke> it appears that the zfs module is missing from the modloop, but I'm not entirely sure why
2021-07-03 14:31:41 <bananicorn> ikke: While building the image, I can see this:
2021-07-03 14:31:47 <bananicorn> >>> mkimage-x86_64: Building zfs            
2021-07-03 14:32:00 <bananicorn> >>> mkimage-x86_64: --> kernel x86_64 lts ffc2d40827ebb1aebf924c1eb54a58e8d5434f0f linux-lts l
2021-07-03 14:32:00 <bananicorn> inux-firmware wireless-regdb xtables-addons-lts zfs-lts                                       
2021-07-03 14:32:00 <bananicorn> Warning: extra firmware " " not found!
2021-07-03 14:32:26 <bananicorn> Not sure what the extra firmware warning is about, but it clearly seems to be including zfs-lts :/
2021-07-03 14:46:57 <bananicorn> My issue may be caused by an outdated build machine, don't look into it too much before I tried again after an update :)
2021-07-03 14:53:49 <bananicorn> ikke: It was indeed an issue of my build machine, it was still running 3.13...
2021-07-03 14:55:05 <bananicorn> I noticed that it was building the kernel with zfs-lts before downloading it from the mirror that I specified, so I thought that it might be using the wrong version, which would explain why it does not show up for the current kernel of v3.14
2021-07-03 14:55:51 <bananicorn> After installing zfs-lts locally and upgrading to 3.14, I built it again, gave it another shot and voila, "modprobe zfs" works just fine on the resulting image.
2021-07-03 14:56:02 <bananicorn> Thanks for your support anyway ikke! :)
2021-07-03 14:57:25 <ikke> NP, glad you found it
2021-07-03 16:53:17 <EnfauKerus> hello I got some problem installing alpine on my pi zero w with sys mode
2021-07-03 16:53:52 <yyp> What's the problem?
2021-07-03 16:54:09 <EnfauKerus> linux-firmware-brcm package have symlink from cypress folder but linux-rpi didnt depend on linux-firmware-cypress
2021-07-03 16:54:35 <EnfauKerus> i am using v3.14
2021-07-03 16:55:06 <EnfauKerus> oh this is fixed on edge
2021-07-03 16:55:27 <EnfauKerus> sorry for bothering
2021-07-03 16:58:37 <EnfauKerus> will this get fixed on v3.14?
2021-07-03 17:01:33 <yyp> I don't think Alpine will patch already released version
2021-07-03 17:03:13 <EnfauKerus> mmh thank you
2021-07-03 17:09:49 <PureTryOut> yyp: yes it will, if security issues or bugs are found
2021-07-03 17:09:57 <PureTryOut> Just no major package upgrades or new packages
2021-07-03 17:10:09 <yyp> Makes sense
2021-07-03 18:40:21 <kevint> minimal thanks for linking that issue!
2021-07-03 18:41:51 <minimal> kevint: seems like ikke merged it after he saw my posting on here so it should be rolled out the the mirrors by now
2021-07-03 18:42:05 <ikke> :)
2021-07-03 18:42:10 <kevint> Great, thank you
2021-07-03 18:43:09 <minimal> sorry for the issue in the first place, unfortunately no matter how much you test stuff locally some problem can always sneak in
2021-07-03 18:47:43 <c705> no kidding. I test things twice and I always still manage to have problems get by
2021-07-03 18:51:42 <minimal> c705: maybe we need a new repo called "extremely-edgy" that stuff goes into first ;-)
2021-07-03 23:25:37 <GreyDev> I'm doing some custom things using the mkimage script: I gather this is the same way the official releases are built?
2021-07-03 23:26:02 <GreyDev> are the build logs public at all? I'd like to see if my output matches up as I'm having some issues
2021-07-03 23:26:29 <GreyDev> I know the rest of aports has public build logs, but I can't find the mkimage logs
2021-07-04 09:35:35 <markand> did something change with 4.14? I no longer get wlan0 on my pi 4 after updating
2021-07-04 09:35:56 <markand> dmesg complains about brcmfmac not able to find the brcmfmac43455-sdio.bin
2021-07-04 09:36:06 <markand> to my surprise I see that it's a symbolic link to cypress?
2021-07-04 09:36:10 <markand> lrwxrwxrwx    1 root     root            31 Jul  4  2021 /lib/firmware/brcm/brcmfmac43455-sdio.bin -> ../cypress/cyfmac43455-sdio.bin
2021-07-04 09:38:16 <markand> and indeed it works by installing linux-firmware-cypress now 
2021-07-04 18:15:48 <gera> Hi. I want to install alpine on my laptop. Which version has ext4 filesystem support?
2021-07-04 18:16:05 <Misthios> every version
2021-07-04 18:16:24 <Misthios> i believe ext4 is the default
2021-07-04 18:16:47 <gera> I tried standard and extended and couldn't mount my partitions. /proc/filesystems doesn't list ext4
2021-07-04 18:17:40 <ikke> modprobe ext4
2021-07-04 18:17:47 <ikke> or mount -t ext4 ..
2021-07-04 18:18:07 <gera> ok, i'll try it
2021-07-04 19:04:24 <gera> i've mounted my efi partition to /mnt/boot but setup-disk is installing grub like its mbr
2021-07-04 19:04:43 <gera> how do i make it onstall to efi directory?
2021-07-04 19:04:54 <raven523> did you boot with UEFI?
2021-07-04 19:05:26 <gera> yes, my laptop only supports uefi
2021-07-04 19:05:56 <nickersonm> gera: I use `apk add efibootmgr grub-efi; BOOTLOADER=grub; USE_EFI=1; setup-disk -m sys <target>`
2021-07-04 19:09:08 <gera> nickersonm: still installing as mbr\=
2021-07-04 19:12:08 <Hello71> how did you determine you are using efi
2021-07-04 19:12:16 <Hello71> and what do you mean "like its mbr"
2021-07-04 19:13:08 <nickersonm> gera: 3.14?  and you can check EFI functionality via `efibootmgr` which will list the EFI boot entries
2021-07-04 19:13:34 <ikke> mount | grep efi
2021-07-04 19:15:17 <gera> nickersonm: https://files.catbox.moe/0m3fwc.jpg
2021-07-04 19:15:34 <gera> maybe i should chroot and do grub-install?
2021-07-04 19:17:08 <nickersonm> gera: oh yes, I forgot that portion
2021-07-04 19:17:09 <nickersonm> chroot $MNT grub-install --target=x86_64-efi --efi-directory=/boot
2021-07-04 19:17:09 <nickersonm> chroot $MNT grub-mkconfig -o /boot/grub/grub.cfg
2021-07-04 19:17:09 <nickersonm> for n in proc sys dev etc/resolv.conf; do mount --rbind /$n $MNT/$n; done
2021-07-04 19:17:09 <nickersonm> chroot $MNT apk del syslinux
2021-07-04 19:18:03 <gera> maybe --bootloader-id=alpine also
2021-07-04 19:21:41 <gera> should the correct mount point for rfi be /boot/efi instead of /boot ? i dont think grub and kernels should be on efi part
2021-07-04 19:21:48 <gera> efi*
2021-07-04 19:22:58 <nickersonm> gera: this is what I've recently used.  I just modified it for ext4, so there could be mistakes with the root FS or mounting root, but it's otherwise been successful: https://rentry.co/gcnww
2021-07-04 19:26:08 <minimal> gera: Grub doesn't need the kernel/initramfs in the same directory as its EFI config, so mounting the ESP partition on /boot/efi makes sense.
2021-07-04 19:26:42 <minimal> whereas with Syslinux UEFI support the kernel/initramfs need to be in the same directory as Syslinux files so there I had to mount the ESP partition as /boot
2021-07-04 19:29:45 <gera> i installed grub and it boots
2021-07-04 19:29:59 <gera> ty nicketsonm for your script
2021-07-04 19:30:12 <gera> i used /boot/efi as mountpoint
2021-07-05 02:23:18 <gavodavo> Hi guys so I have a problem with my rpi4. Basically a long time I go I set it up as a traditional sys-based installation(I don't think that's relevant but I'll leave it just in case). Today I upgraded the distro to 3.14.0, however now the kernel can't find the wlan broadcom firmware blob on my device
2021-07-05 02:24:32 <gavodavo> Ah wait
2021-07-05 02:24:43 <gavodavo> ```
2021-07-05 02:24:45 <gavodavo> ∅ /run/media/me/d1839b0f-103a-4651-afc9-b6e13f5d0bf8/lib/firmware/brcm
2021-07-05 02:24:45 <gavodavo> ❯ file brcmfmac43455-sdio.bin
2021-07-05 02:24:45 <gavodavo> brcmfmac43455-sdio.bin: broken symbolic link to ../cypress/cyfmac43455-sdio.bin
2021-07-05 02:28:32 <zcrayfish> linux-firmware-cypress  not installed?
2021-07-05 02:28:41 <zcrayfish> Someone mentioned that earlier.
2021-07-05 02:28:44 <gavodavo> Yes after installing `linux-firmware-cypress` it works now :)
2021-07-05 02:28:53 <zcrayfish> sweet
2021-07-05 12:11:21 <Hello71> maybe we should backport that
2021-07-05 13:28:35 <danopia> I'm updating a CI pipeline from alpine 3.12 to alpine 3.14 and, for some reason I cannot figure out, ruby-related things are acting broken when the 3.14 image is run on my CI host. has anyone seen ruby acting like $PATH is messed up?
2021-07-05 13:29:45 <danopia> I call ghi which crashes saying `No such file or directory - git (Errno::ENOENT)` (ghi execs git internally); I installed ruby-irb and it crashes on start with `No such file to load -- irb/error.rb (LoadError)` so I can't even debug what ruby is doing that easily :)
2021-07-05 13:30:06 <ikke> Check the 3.14 release notes
2021-07-05 13:30:20 <ikke> Especially the wiki part
2021-07-05 13:31:16 <danopia> I'm looking at https://wiki.alpinelinux.org/wiki/Release_Notes_for_Alpine_3.14.0 - are you referring to the faccessat2 part?
2021-07-05 13:32:02 <danopia> Or am I just on the totally wrong page haha
2021-07-05 13:33:04 <ikke> No, that indeed 
2021-07-05 13:33:25 <ikke> Not sure if that is the issue, but a likely culprit 
2021-07-05 13:33:46 <danopia> Well it would explain why I can't repo on my laptop with the same exact image
2021-07-05 13:34:07 <danopia> Ok, so alpine 3.14 just can't be used properly on CI providers that don't offer docker 20, such as circleci
2021-07-05 13:36:41 <danopia> I'm going to verify my side with 3.13 for now but it's a bit annoying because 95% of my pipeline is working fine, there's just a legacy ruby step at the very end that isn't happy :)
2021-07-05 13:37:32 <herag> So this may be a newb thing, but here goes. I have used Alpine Linux with XFCE for my main desktop computer for years now. I recently decided I wanted to switch to DWM. So I downloaded the source for DWM and Dmenu and St. Compiled and installed them. Then I went through the wiki about DWM and installed all the deps that it said were necessary. But now, when I startx, I get an
2021-07-05 13:37:34 <herag> error and it won't start. I get `/dev/dri/card0: permission denied` and this is on an ASUS EeePC 1005HAB 32-Bit... I only join the channel because my search-fu skills have come to an end and I'm not sure what is causing this... it is probably something simple that I'm overlooking. Any ideas?
2021-07-05 13:43:42 <trylletre> herag: is your user in the video group
2021-07-05 13:44:15 <herag> ... omg... I'm an idiot... hold on... let me check. If this is it... feel free to ban me forever.
2021-07-05 13:50:31 <herag> trylletre: Okay, so now the log output says `xf86EnableIOPorts: failed to set IOPL for I/O {Operation Not Permitted` as the first EE
2021-07-05 13:51:17 <trylletre> perhaps also the input group if you didnt have that either, but i dont know about that error specifically
2021-07-05 13:51:40 <trylletre> you also need udev/mdev, at least per the 3.14 changelog as X doesnt have root anymore
2021-07-05 13:53:36 <herag> I did add input to my user groups. Oh, okay. I will tinker with both of those ideas. I need to setup my permissions again anyway. I was just trying to get DWM setup before I did that, but I guess I shouldn't cut corners. I'll check back when I have my groups all sorted and udev/mdev, if I don't already have those. I'm a former Slackware user, so I'm very accustomed to all these
2021-07-05 13:53:38 <herag> things happening automagically for me. :D
2021-07-05 14:12:27 <herag> trylletre: okay... it worked. Adding input worked... I think. I added a bunch groups actually, but input was one of them. I don't know why I always forget about Linux permissions. Thank you so very much!
2021-07-05 14:12:48 <trylletre> happy to help :)
2021-07-05 19:34:19 <kmmndr> Hi all :-)
2021-07-05 19:34:36 <kmmndr> Is there something wrong on gitlab ?
2021-07-05 19:34:53 <kmmndr> I'm trying to create a MR having only 1 commit
2021-07-05 19:35:11 <kmmndr> But it seems to include many more from master
2021-07-05 19:35:27 <ikke> kmmndr: link?
2021-07-05 19:36:20 <kmmndr> ikke: https://gitlab.alpinelinux.org/kmmndr/aports/-/commits/ruby-build
2021-07-05 19:38:08 <ikke> https://gitlab.alpinelinux.org/kmmndr/aports/-/merge_requests/new?utf8=%E2%9C%93&merge_request%5Bsource_project_id%5D=544&merge_request%5Bsource_branch%5D=ruby-build&merge_request%5Btarget_project_id%5D=1&merge_request%5Btarget_branch%5D=master
2021-07-05 19:39:18 <ikke> kmmndr: You need to create the merge request against the right project
2021-07-05 19:39:19 <kmmndr> Arff ... I'm a silly boy, thanks for you help 
2021-07-05 19:39:33 <ikke> No problem, when I followed your link, and create an MR, I got the same
2021-07-05 19:40:06 <ikke> If you go to branches, and there create an MR, it seems to create it against your local master, which is behind
2021-07-05 19:40:32 <ikke> instead, you need to make it against alpine/aports
2021-07-05 19:40:37 <kmmndr> That's it, and I haven't seen it at first glance
2021-07-05 19:43:01 <kmmndr> Is there a quick tip to build many package at once ?
2021-07-05 19:43:55 <kmmndr> For example, when a package require multiple dependencies from testing, built and added in the same MR 
2021-07-05 19:45:11 <kmmndr> something like git diff | grep APKBUILD | sort by commit
2021-07-05 19:46:50 <ikke> https://gitlab.alpinelinux.org/alpine/infra/docker/alpine-gitlab-ci/-/blob/master/overlay/usr/local/bin/build.sh#L115
2021-07-05 19:49:31 <kmmndr> Great, thanks again ikke ! :-)
2021-07-06 10:52:44 <dalz> hello, I'm trying to run rspamd but it segfaults every time I start it, what could be the problem?
2021-07-06 10:53:11 <lopid> anything useful when you run it manually?
2021-07-06 10:53:20 <dalz> nope, just "Segmentation fault"
2021-07-06 10:53:43 <dalz> running it with su rspamd -s /bin/sh -c 'rspamd --no-fork'
2021-07-06 10:54:10 <lopid> from whence was it installed?
2021-07-06 10:54:44 <dalz> just apk add rspamd
2021-07-06 10:54:48 <dalz> i'm on alpine 3.14
2021-07-06 10:55:58 <dalz> raspberry pi version if it matters
2021-07-06 10:56:43 <clandmeter> dalz: did you upgrade recently?
2021-07-06 10:58:00 <dalz> it's a very recent install, and i ran apk update and upgrade when installing
2021-07-06 10:58:23 <clandmeter> did you run upgrade -a?
2021-07-06 10:58:25 <lopid> update-conf
2021-07-06 10:58:52 <dalz> clandmeter: trying it now
2021-07-06 10:59:04 <dalz> OK: 786 MiB in 315 packages
2021-07-06 10:59:13 <dalz> lopid: no output
2021-07-06 10:59:30 <clandmeter> apk -U upgrade -a
2021-07-06 10:59:35 <clandmeter> thats what i normally use 
2021-07-06 10:59:42 <lopid> fwiw mine won't start because the cpu doesn't suppose ssse3 instructions
2021-07-06 10:59:46 <lopid> support
2021-07-06 11:00:06 <dalz> clandmeter: still nothing
2021-07-06 11:00:32 <clandmeter> which arch are you running?
2021-07-06 11:01:04 <dalz> aarch64
2021-07-06 11:02:15 <dalz> oh well
2021-07-06 11:02:17 <dalz> https://github.com/rspamd/rspamd/issues/3563
2021-07-06 11:04:05 <lopid> can you examine the coredump on musl?
2021-07-06 11:06:23 <dalz> if by "examine" you mean "follow some instructions" sure, otherwise i think it's out of my reach :)
2021-07-06 11:06:32 <dalz> how do i get the core dump?
2021-07-06 11:06:33 <ikke> uname -c unlimited
2021-07-06 11:06:36 <ikke> then run it again
2021-07-06 11:07:05 <dalz> uname: unrecognized option: c
2021-07-06 11:07:15 <ikke> sorry, ulimit, not uname
2021-07-06 11:07:33 <clandmeter> :p
2021-07-06 11:07:42 <dalz> still segfaults
2021-07-06 11:07:50 <clandmeter> yes
2021-07-06 11:07:51 <ikke> yes, but now there should be a coredump file
2021-07-06 11:07:55 <clandmeter> but you will have a core
2021-07-06 11:08:09 <clandmeter> in your working directory by default
2021-07-06 11:08:32 <dalz> ok i see it
2021-07-06 11:18:37 <clandmeter> looks like it segfaults on v3.14 but not on edge
2021-07-06 11:21:57 <dalz> let me try
2021-07-06 11:22:56 <clandmeter> strace says it wants to open /var/lib/rspamd//b24672aa66af71b15f4d08d7edeaa5a31068f47e6937c1f791c9c77a17c0042c.hsmp but its does not exist and then sigsegv
2021-07-06 11:34:40 <dalz> upgraded to edge and now it doesn't boot anymore :D
2021-07-06 11:36:55 <dalz> nevermind i turned it off and on again and it works now
2021-07-06 11:37:00 <dalz> rspamd runs fine
2021-07-06 11:37:06 <dalz> thanks everyone
2021-07-06 11:44:22 <clandmeter> dalz: glad you fixed it, but we do not advise running on edge.
2021-07-06 11:44:49 <clandmeter> if you want you can create an issue on gitlab.a.o and somebody could have a look at it later when they have time.
2021-07-06 11:50:02 <dalz> clandmeter: i was using arch linux before i think i'll be fine ;)
2021-07-06 11:50:11 <dalz> but sure i'll file an issue
2021-07-07 00:24:45 <eletrotupi> hello! I was wondering how should ruby dependencies be handled on alpine? should we follow similar to how python handle dependencies?
2021-07-07 02:03:09 <Aftershock> Has anyone had trouble building python wheels in alpine? I feel like I'm doing something wrong. I've tried to install py3-lxml but pip still complains that lxml is missing and if I try to build lxml it fails.
2021-07-07 02:03:37 <Aftershock> same for apks like cython
2021-07-07 02:03:44 <Aftershock> pip can't seem to see them
2021-07-07 02:54:25 <Hello71> Aftershock: ten bucks says you're mixing release and edge
2021-07-07 02:54:34 <Hello71> eletrotupi: #alpine-devel
2021-07-07 14:01:32 <yann-kaelig> Hello
2021-07-07 14:02:01 <yann-kaelig> I'm tryiing to find what is an apk database , where can I find it or how can I create it
2021-07-07 14:04:08 <ikke> /lib/apk/db
2021-07-07 14:14:05 <onelegend> I like this a lot
2021-07-07 14:17:11 <lopid> 
ACTION blushes

2021-07-07 14:17:50 <yann-kaelig> well, I missed something, this folder doesn't exist and from the content search on the alpine website can not find any of the files in this folder. 
2021-07-07 14:18:57 <yann-kaelig> apk-tools doesn't contain anything like a /lib/apk/db, I don't know where to look
2021-07-07 14:20:50 <yann-kaelig> how is this /lib/apk/db created
2021-07-07 14:30:20 <yann-kaelig> Am I the only one with a 502 bad gateway for git.alpinelinux.org ?
2021-07-07 14:32:18 <ikke> No our monitoring system indicates it as well 
2021-07-07 14:44:35 <programmerq> yann-kaelig▸ 'lib-apk-db' appears in /lib/libapk.so.3.12.0 so I would venture a guess that apk itself creates those files as it does its operations.
2021-07-07 14:45:22 <ikke> Unless you create a chroot from scratch, you should already have it 
2021-07-07 14:47:23 <programmerq> hey ncopa -- a few years ago you gave a presentation on the history of alpine linux. Do you have slides for that anywhere?
2021-07-07 14:47:56 <programmerq> or anything similar-- I'd like to do a similar presentation at my current company for some of the junior tech folks
2021-07-07 14:48:25 <yann-kaelig> it's not a chroot from scratch but I have built abuild and apk-tools-static on another distro. I tried to build apk-tools but it failed with ERROR: Failed to open apk database: No such file or directory. Obviously there is something I do not understand with this apk database
2021-07-07 14:49:03 <ikke> Huh, just when building apk-tools? 
2021-07-07 14:50:22 <ncopa> programmerq: which presentation? the one at dockercon?
2021-07-07 14:51:01 <programmerq> ncopa▸ I seem to recall something being presented internally. You just went over the basic history and how apk is tiny and fast and awesome and why libmusl was chosen.
2021-07-07 14:51:28 <programmerq> if there's another set of slides with that same vibe, I'll be happy.
2021-07-07 14:52:32 <yann-kaelig> ikke: That was my fault, I have enabled the check part and that why apk-tools failed. Thx
2021-07-07 14:53:12 <programmerq> (internally when we were both at docker-- sometime within the first year that you were there, if I recall correctly)
2021-07-07 14:54:30 <ncopa> right. i think i remember. I doubt i still have the slides but I will have a look
2021-07-07 14:55:06 <programmerq> it's not a big deal-- don't go out of your way. It's a pretty informal presentation I'm wanting to do.
2021-07-07 14:55:34 <programmerq> if you do share anything I would super appreciate it. :)
2021-07-07 14:56:26 <ikke> yann-kaelig: FYI, apk has an option to initialize a new DB 
2021-07-07 14:56:39 <ncopa> programmerq: im a bit impressed that you still remember it. My memory is that everyone was sleeping except for one or two who was giving attention to be polite :D
2021-07-07 14:56:55 <programmerq> I was joined over zoom
2021-07-07 14:57:24 <programmerq> but that was honestly when I started really paying attention to alpine and dug into apk
2021-07-07 14:57:29 <programmerq> and now apk is my favorite
2021-07-07 14:57:35 <ncopa> :)
2021-07-07 15:07:36 <Slence> All right. Time to migrate to Alpine Linux for my main Linux laptop.
2021-07-07 15:12:00 <ncopa> Slence: best wishes!
2021-07-07 15:13:34 <Slence> ncopa: Thanks. Can't be worse than installing OpenBSD ;)
2021-07-07 15:14:10 <Slence> But, the minimalism is appealing to me, and since I tinker a lot with postmarketOS/sxmo, I might as well get the base linux daily driven as well.
2021-07-07 15:25:12 <BHRobinson> Hi all. I am trying to install qemu on 3.14 but it cannot find the packages. Is there something special that needs to be done?
2021-07-07 15:26:30 <programmerq> "apk update && apk add qemu" works on a fresh 3.14 for me.
2021-07-07 15:27:32 <BHRobinson> I get a no such package. 
2021-07-07 15:28:07 <BHRobinson> right now I am running the standard image, but I did try the virt as well. 
2021-07-07 15:28:37 <BHRobinson> and apk search qemu shows only vde2 and xen
2021-07-07 15:29:45 <lopid> is it in the community repo?
2021-07-07 15:30:01 <ikke> Yes, you need to make sure you have the community repository enabled
2021-07-07 15:31:16 <BHRobinson> there you go... I am not thinking :) Community is still remarked out
2021-07-07 15:31:18 <BHRobinson> thank you!
2021-07-07 15:46:11 <darkelvenangel> Hello, I need some assistance, I'm trying to write a script that will run when the system is powering off but not rebooting. I've setup /etc/local.d and now I need how to determine if this is a poweroff or reboot.  Runlevels are the same for both so I need something else.  Thanks.
2021-07-07 15:52:19 <ikke> darkelvenangel: isn't it better to use a service in the shutdown runlevel?
2021-07-07 15:54:12 <darkelvenangel> ikke: possibly, however when I was testing reboot seem to also use the shutdown runlevel
2021-07-07 15:54:28 <ikke> Yeah, not sure if you can differentiate between that
2021-07-07 15:56:29 <lopid> maybe there's something in /sys or /proc that you can examine
2021-07-07 15:58:07 <darkelvenangel> Ah, thank you.  Do you know if Alpine uses runlevel numbers like Gentoo does?  
2021-07-07 15:58:30 <Slence> Ah, everything seems to work fine. Just some issues getting dwm to run with startx in .profile.
2021-07-07 15:58:59 <Slence> I followed the guide at https://wiki.alpinelinux.org/wiki/Dwm. Assuming I did something wrong, but not sure what. Perhaps just some X error with resolution or whatnot.
2021-07-07 15:59:11 <lopid> they both use openrc, so… maybe
2021-07-07 15:59:50 <ikke> afaik we only use named runlevels
2021-07-07 16:01:07 <Slence> "Fatal server error. no screens found".
2021-07-07 16:01:14 <Slence> Yeah, probably some X-related stuff.
2021-07-07 16:01:18 <darkelvenangel> I've try logging the named runlevels so if that's the case it's not going work.  
2021-07-07 16:02:25 <Slence> The guide says to use sudo, but I assume it should not matter that I use doas instead.
2021-07-07 16:02:44 <ikke> no, that should not matter
2021-07-07 16:02:54 <Slence> Right, so do I need to do some X configuration to get it working?
2021-07-07 16:03:00 <Slence> X does not seem to recognize my screen.
2021-07-07 16:03:11 <Slence> Maybe some missing graphics drivers?
2021-07-07 16:03:51 <ikke> do you have the correct graphics drivers installed?"
2021-07-07 16:03:57 <ikke> check the xorg log
2021-07-07 16:04:39 <Slence> It is trying to load modules fbdev (does not exist) and vesa (module does not exist) and it gets permission denied on /dev/dri/card0.
2021-07-07 16:04:51 <ikke> Ah, are you member of video?
2021-07-07 16:05:18 <Slence> No, I have followed https://wiki.alpinelinux.org/wiki/Dwm 1:1.
2021-07-07 16:05:31 <Slence> Nothing beyond that, so I need to add my user to 'video' group. Gotcha!
2021-07-07 16:06:18 <lopid> and relog in
2021-07-07 16:07:10 <Slence> There we go. That did it. Thanks for the help!
2021-07-07 16:07:18 <Slence> May I suggest adding that part to the dwm guide?
2021-07-07 16:07:27 <Slence> Or maybe I just missed it in the adduser section.
2021-07-07 16:07:37 <Naib> so how is the openRC replacement coming? Alpine might want the alternative sooner than later after gentoo just canned opentmpfiles
2021-07-07 16:07:51 <ikke> Naib: Waiting for funding
2021-07-07 16:08:01 <Naib> the update said tehre is funding
2021-07-07 16:08:54 <Naib> but since opentmpfiles is an openrc project and they just sat on a CVE for years... they pulled the plug and the alterantive is systemd-tmpfiles but that uses alot of systemd code and it might not work with musl
2021-07-07 16:10:14 <Slence> Ah. Touchpad does not work. On to fixing that then. :)
2021-07-07 16:10:21 <ikke> Naib: source?
2021-07-07 16:10:32 <ikke> Slence: input group :)
2021-07-07 16:10:37 <Naib> https://bugs.gentoo.org/751415
2021-07-07 16:10:44 <ikke> and make sure you have input drivers
2021-07-07 16:11:04 <Slence> ikke: Oh, I see. Thanks!
2021-07-07 16:11:06 <Naib> https://gitweb.gentoo.org/repo/gentoo.git/log/sys-apps/opentmpfiles?showmsg=1
2021-07-07 16:11:19 <Naib> all versions got hard-masked on gentoo
2021-07-07 16:12:04 <ikke> Naib: in any case, the first focus of the new service manager is to handle services
2021-07-07 16:12:46 <Naib> ikke: idd, it was always neferious why packages needed to start supporting systemd tmpfile but it snuck in ...
2021-07-07 16:13:19 <ikke> skarnet differentiates between service handling and policy handling
2021-07-07 16:14:05 <Naib> good :) gentoo contorted them with the direction of openrc 
2021-07-07 16:14:26 <Naib> rather then ensuring things were done completly, simply and well
2021-07-07 16:15:11 <ikke> https://skarnet.com/projects/service-manager.html#plan
2021-07-07 16:21:13 <Slence> Ok. Up and running now with dwm. Thanks!
2021-07-07 16:22:07 <Slence> The font seems to assume I am 19 years old with 20/20 vision, so increasing that by like 500x is the next task on the todo. :)
2021-07-07 16:22:44 <Hello71> more that it assumes you're using a 640x480 CRT
2021-07-07 16:23:06 <Slence> :D
2021-07-07 16:23:35 <Slence> I am very very happy with how minimal this install is and how it assumes nothing regarding packages I may or may not want.
2021-07-07 16:24:02 <Slence> One thing that has kept me away from Linux Distributions for a long time is systemd and bloat in most dists.
2021-07-07 16:24:18 <Slence> This solves both of my issues.
2021-07-07 16:24:19 <minimal> Slence: you called? ;-)
2021-07-07 16:24:22 <Slence> Thanks for creating this!
2021-07-07 16:25:17 <Slence> minimal: Hah! Trust me, I know what it feels like to have a nick that often gets hilighted. For a long time, my away.log was full of "SILENCE, I KEEL YOU". Now it's mostly people talking about the ircd command /SILENCE.
2021-07-07 16:25:32 <yann-kaelig> I may need glasses but I can't see anything from the apk man page or even with apk --help about how to initialise a  new database
2021-07-07 16:38:12 <Slence> I am very happy with this install.
2021-07-07 16:48:49 <Slence> What's the recommended way for installing different locales for st?
2021-07-07 16:49:00 <Slence> Need SE keyboard input, as defined during setup-alpine.
2021-07-07 16:49:21 <Slence> Or, not just for st but as the default keyboard input
2021-07-07 16:50:48 <Slence> musl-locales?
2021-07-07 16:53:24 <Slence> Ah, setup-keymap.
2021-07-07 16:53:42 <Slence> At this point I would happily welcome a "RTFM!".
2021-07-07 16:53:46 <Slence> I deserve it.
2021-07-07 16:57:58 <Slence> Hmm. Is it not supposed to work after setup-keymap and defining the layout I want?
2021-07-07 16:58:18 <Slence> loadkmap file looks fine.
2021-07-07 17:02:32 <ikke> yann-kaelig: somehow that option is missing from the help
2021-07-07 17:02:47 <ikke> apk add --initdb --allow-untrusted
2021-07-07 17:09:17 <yann-kaelig> thx a lot ikke, you saved my day
2021-07-07 17:09:43 <Slence> st does not seem to respect loadkmap file at all, despite restarting.
2021-07-07 17:09:50 <Slence> Nor does other programs.
2021-07-07 17:09:56 <yann-kaelig> so, it's working, I can build and install apk package on my non alpine distro, that nice and very interesting
2021-07-07 17:10:11 <Slence> Am I missing something here?
2021-07-07 17:11:34 <Slence> https://techoverflow.net/2021/05/17/how-to-change-keyboard-layout-in-alpine-linux/
2021-07-07 17:11:54 <Slence> That one tells me it should work automagically and be persistent across reboots, but it doesnt work with or without reboot.
2021-07-07 17:11:58 <Slence> It worked fine during setup-alpine.
2021-07-07 17:12:02 <Slence> But that did not persist.
2021-07-07 17:17:30 <yann-kaelig> which one of consolekit2 and elogind is used by default on Alpine ?
2021-07-07 17:17:36 <Hello71> setup-keymap only controls console keymap
2021-07-07 17:17:42 <Hello71> x keymap is separate
2021-07-07 17:17:55 <Hello71> for confusing and mostly dumb historical reasons
2021-07-07 17:18:00 <Hello71> yann-kaelig: none
2021-07-07 17:18:05 <yann-kaelig> :)
2021-07-07 17:18:20 <Hello71> but consolekit and consolekit2 are obsolete
2021-07-07 17:19:29 <Slence> Hello71: Ah, so, how do I fix X keymap then? Just export LANG etc_
2021-07-07 17:19:31 <Slence> etc?
2021-07-07 17:19:46 <Hello71> no, that's for output
2021-07-07 17:19:57 <PureTryOut> consolekit2 is obsolete yes, but it actually started being maintained again funnily enough
2021-07-07 17:20:03 <Hello71> huh
2021-07-07 17:20:06 <PureTryOut> yann-kaelig: Alpine uses neither "by default"
2021-07-07 17:20:26 <Hello71> does rootless x actually work? i think that's the only thing people actually care about
2021-07-07 17:20:35 <Hello71> PureTryOut: didn't I say that already
2021-07-07 17:22:38 <PureTryOut> I guess you did 🤷‍♂️
2021-07-07 17:24:42 <yann-kaelig> Sorry, I was not clear enough. Taking a look at plasma, there is no dependency with elogind or consolekit2. Maybe it's somewhere else, or the user has the choice ?
2021-07-07 17:25:43 <trylletre> you might not need either? you can run X with just mdev/udev and the right permissions in rootless without elogind/ck
2021-07-07 17:25:46 <Slence> Hello71: Most of the guides I find online suggest setup-keymap should solve this for me, but I take it thats not the case?
2021-07-07 17:26:17 <Slence> To be clear, what Im after is SE keyboard layout.
2021-07-07 17:26:22 <trylletre> Slence: for xorg use setxkbmap inside the x session
2021-07-07 17:26:35 <Slence> Thanks! Will try that.
2021-07-07 17:31:12 <Slence> Ah, great. That worked. 
2021-07-07 17:36:55 <trylletre> you might also want to specify -model pc104 or 105 depending on your keyboard
2021-07-07 17:37:06 <trylletre> most se keyboards are 105, im unsure on what default gets picked
2021-07-07 17:37:43 <PureTryOut> yann-kaelig: plasma depends on polkit-kde-agent, which depends on polkit-qt which depends on polkit-elogind
2021-07-07 17:37:52 <PureTryOut> And well that last one pulls in elogind
2021-07-07 17:39:10 <Slence> trylletre: The default worked fine for me. Thanks anyway. :)
2021-07-07 17:39:19 <Slence> Just running setxkbmap on login now.
2021-07-07 17:39:33 <Slence> And dwmstats, because that's also nice to have.
2021-07-07 17:39:45 <Slence> So, all good now. Working perfectly.
2021-07-07 18:22:45 <nickersonm> Does ifupdown-ng support bridge definitions, or do I need to use the 'bridge' package?  I see 'bridge' just installs openrc scripts to call `brctl`.
2021-07-07 18:41:57 <yann-kaelig> I changed the path of the REPODEST in abuild.conf and I removed the old path but now when I initialise a db I get a WARNING: Ignoring /home/yann-kaelig/repo/v3.14/main/apkbuild: No such file or directory . How can I remove this warning, I cheakced in /var/cache/apk but there is nothing
2021-07-07 18:42:35 <yann-kaelig> checked*
2021-07-07 18:43:38 <yann-kaelig> Where is this path still recorded ?
2021-07-07 19:13:04 <yann-kaelig> well, I lost to much time with apkbuild, it's a shame but to much thing s are done in the shadow. not the package builder I'm looking for. :(
2021-07-07 19:13:18 <yann-kaelig> thx
2021-07-07 20:45:44 <Slence> Alpine is doing wonders to my battery life.
2021-07-07 20:46:20 <Slence> Or perhaps it's dwm.
2021-07-07 20:53:04 <graywolf> Hello. Quick question, what is upgrade story for firefox? Does it stay on major version at the time of alpine version release or are the upgrade done in all supported versions (of alpine)?
2021-07-07 20:58:11 <ikke> firefox is in community, which is supported at the moment at most until the next release
2021-07-07 20:58:14 <Xe> is there a start-stop-daemon flag that reads envvars from a file to put them into the environment of the running service?
2021-07-07 20:58:35 <Xe> or at least one to passthrough stuff in the current environment to it
2021-07-07 20:59:06 <Xe> i'm in a bit of a "god is dead" kind of situation where I need that behavior and I don't see a flag for it in the manpage
2021-07-07 21:01:46 <lopid> i don't know about start-stop-daemon, but can you source an env file?
2021-07-07 21:02:21 <Xe> that's what i do currently
2021-07-07 21:02:27 <lopid> oh
2021-07-07 21:02:29 <Xe> it seems to shed that
2021-07-07 21:03:00 <Xe> here's the openrc script in question: https://github.com/tailscale/tailscale/blob/main/cmd/tailscaled/tailscaled.openrc
2021-07-07 21:03:31 <lopid> you'll have to export them
2021-07-07 21:03:48 <Xe> as in `export source`?
2021-07-07 21:04:00 <lopid> no, export each var that is sourced
2021-07-07 21:04:30 <Xe> but there is potentially an uncountable number of things in that file being sourced
2021-07-07 21:04:38 <lopid> 
ACTION nods

2021-07-07 21:04:54 <Xe> basically i want the behavior that systemd's [Service].EnvironmentFile has
2021-07-07 21:06:32 <lopid> you could mark each var for export in the sourced file itself
2021-07-07 21:06:36 <Xe> does openrc-run support $() expansions?
2021-07-07 21:07:02 <graywolf> ikke: Ok but are newer versions backport to the current stable (at the moment 3.14)? Because for example for postgresql afaik they are not, so for firefox it should be the same? But I'm not sure it is, therefore I am asking.
2021-07-07 21:08:15 <lopid> if it's posix it should
2021-07-07 21:08:16 <Xe> set -a before and set +a after seemed to do what i want
2021-07-07 21:08:29 <ikke> graywolf: they can be backported, but it's the responsibility of the maintainer to do that
2021-07-07 21:09:24 <lopid> ah, right
2021-07-07 21:09:41 <ikke> Xe: right, that's how I would solve it
2021-07-07 21:10:10 <graywolf> Does that mean that to have "safe" (for some definition of safe) firefox version I should go for -esr, since that should be supported at the same version mostly by firefox upstream? (I realize that maintainer still needs to release new version of the package and all of that)
2021-07-07 22:15:08 <nickersonm> Does anyone have an opinion on using libvirt vs. using lxd + qemu directly?  Primarily Alpine containers, perhaps some alpine-on-rpi or Debian-on-x64 VMs.
2021-07-07 22:18:46 <onelegend> I run LXD containers directly on a bare metal instance of Alpine Linux
2021-07-07 22:23:15 <niallr> Hello!
2021-07-07 22:23:39 <niallr> I'm new to alpine linux, am setting up a machine following this webpage.
2021-07-07 22:23:41 <niallr> https://wiki.alpinelinux.org/wiki/LVM_on_LUKS
2021-07-07 22:23:46 <nickersonm> onelegend: using `lxc` to manage, or something else?
2021-07-07 22:24:02 <niallr> Using the gpt method.
2021-07-07 22:24:38 <niallr> The problem I'm running into is that when I follow the steps exactly, the resulting grub config seems to missing modules that i'm used to seeing in FDE setups
2021-07-07 22:24:49 <niallr> and sure enough, when I try and boot, I get a
2021-07-07 22:25:09 <niallr> error:disk 'lvmid/<blahblahblah>' not found.
2021-07-07 22:26:05 <niallr> Following the steps again just leads to the same non-booting setup for me, and I'm not sure how to manually fix the grub config (if that even is the problem here.)
2021-07-07 22:27:01 <dowNz> Hey guys.. I have some strange problem.. when I am running Nextcloud on my Odroid-C2 with docker images (using arm64v8/alpine) I am getting Segmentation fault when I try to run php index.php in docker... but when I gzip it, untar it on x86 machine and run it with x86 images everything is working well..
2021-07-07 22:27:12 <dowNz> can it be that my Odroid is corrupted?
2021-07-07 22:27:22 <dowNz> or it's some bug in alpine or php?
2021-07-07 22:46:12 <Slence> Who on earth wrote this stuff: http://harmful.cat-v.org/society/sweden
2021-07-07 22:46:16 <Slence> And who hurt that person.
2021-07-07 22:46:22 <Slence> Why is it linked from suckless?
2021-07-07 22:46:25 <Slence> So many questions.
2021-07-07 22:47:16 <c705> wow, someone posted some dumbass shit on the internet?! unbelievable!!
2021-07-07 22:47:25 <c705> better get AP in here
2021-07-07 22:48:41 <Slence> Well, it's linked from suckless so it seems someone supports the view there. But, I will ask in #suckless.
2021-07-07 22:53:00 <c705> wow people have opinions?! and they link to thinks on their webpage?!?!
2021-07-07 22:55:58 <invoked> you can ask in #cat-v
2021-07-07 22:56:07 <invoked> at your own risk
2021-07-07 22:57:53 <Slence> I think I will pass. I am now fairly sure that whoever is behind cat-v is doing some serious meth.
2021-07-07 22:58:19 <Slence> I have never seen a collection of more clueless teenage rage before.
2021-07-07 22:58:36 <invoked> did you look at the harmful.cat-v.org landing page?
2021-07-07 22:59:27 <Slence> c705: This is not opinions. This is poorly informed rage disguised as fact. 
2021-07-07 22:59:32 <Slence> invoked: yes.
2021-07-07 22:59:47 <Slence> It claims to be "not politically correct". It would have been enough with "not correct".
2021-07-07 23:00:14 <invoked> right. so, opinions being what they are, you also have yours.
2021-07-07 23:00:34 <c705> Slence: same difference
2021-07-07 23:00:46 <Slence> invoked: These are not opinions.
2021-07-07 23:00:58 <Slence> invoked: An opinion is "i think oranges are tasty".
2021-07-07 23:01:03 <invoked> i guess you didn't read the landing page then.
2021-07-07 23:01:07 <Slence> I did.
2021-07-07 23:01:21 <invoked> you obviously didn't.
2021-07-07 23:01:25 <c705> yes, yes they are. staged as fact, nonetheless, but unless you take everyting someone says at face value (i don't recommend) then it is an opinion
2021-07-07 23:01:46 <Slence> invoked: What part do you feel I missed?
2021-07-07 23:01:52 <invoked> "this is my humble contribution to attempt to collect all the things that bother me", for starters
2021-07-07 23:01:57 <Slence> yes
2021-07-07 23:01:59 <c705> anyways, what's your point. what do you want us to do about it?
2021-07-07 23:01:59 <Slence> I read that.
2021-07-07 23:02:00 <Slence> "This is my humble contribution to attempt to collect all the things that bother me, the goal is not to offend anyone but entertain and try to dispel widespread silly ideas that infest our minds and perhaps in while doing that make the world a bit more reasonable and pleasant place to live in.
2021-07-07 23:02:05 <Slence> "
2021-07-07 23:02:27 <Slence> Well, #suckless is a better place to ask, but they also forward to #cat-v. I do not understand why this crap is linked from suckless.
2021-07-07 23:02:34 <invoked> you really should share your opinions in #cat-v
2021-07-07 23:02:35 <c705> and how does alpine-linux relate to this?
2021-07-07 23:02:44 <dowNz> shouldn't this discussion be in #alpine-offtopic ?
2021-07-07 23:02:53 <Slence> c705: It does not. #suckless is the place for this. Now im just replying to you.
2021-07-07 23:02:58 <Slence> So I guess we should all change channel. :)
2021-07-07 23:03:03 <c705> good idea
2021-07-07 23:05:14 <dowNz> and also my problem is doing this - https://pastebin.com/TqypRdmA
2021-07-07 23:05:47 <dowNz> for me it looks like some issue with memory module, but on another hand everything else is working well there.. 
2021-07-07 23:36:37 <Slence> dowNz: Also, I -just- discovered the offtopic channel. My apologies.
2021-07-07 23:36:54 <dowNz> Slence: no worries :-)
2021-07-07 23:55:41 <dowNz> mkey.. downgraded to alpine:3.11 and working well..
2021-07-07 23:58:13 <dowNz> 3.12 working as well
2021-07-08 00:00:26 <dowNz> 3.13 as well.. strange..
2021-07-08 00:02:29 <dowNz> and 3.14 = Segmentation fault
2021-07-08 02:09:46 <niallr> I think I found what was tripping up my insta
2021-07-08 02:10:03 <niallr> grub-install --target=x86_64-efi --efi-directory=/boot/efi
2021-07-08 02:10:06 <niallr> produces
2021-07-08 02:10:14 <niallr> sh: /etc/default/grub: line 5: cryptkey: not found
2021-07-08 02:10:14 <niallr> Installing for x86_64-efi platform.
2021-07-08 02:10:14 <niallr> Installation finished. No error reported.
2021-07-08 02:10:56 <niallr> but my GRUB_CMDLINE_LINUX_DEFAULT looks like cryptroot=UUID=<UUID> cryptdm=lvmcrypt cryptkey
2021-07-08 02:22:03 <niallr> Ok, now grub works successfully
2021-07-08 02:22:25 <niallr> but when I boot, screen freezes on Loading initial ramdisk
2021-07-08 02:33:05 <niallr> sshd is enabled to start at boot
2021-07-08 02:33:22 <niallr> but no route to host, so I think it is a true freeze and not a display freeze
2021-07-08 02:33:59 <niallr> my mkinitfs.conf is
2021-07-08 02:34:19 <niallr> ata base ide scsi usb virtio ext4 lvm cryptsetup cryptkey
2021-07-08 02:34:41 <niallr> and I am booting a laptop, from a sata connected ssd
2021-07-08 02:35:02 <niallr> is there anything I am missing that is crashing at the ramdisk?
2021-07-08 02:54:11 <niallr> adding kms and keymap to initrd features worked
2021-07-08 02:54:16 <niallr> I wish I knew why
2021-07-08 02:56:52 <ohheyitsbrian> hey everyone - i'm getting the dreaded "Ignoring https://dl-cdn.alpinelinux.org/alpine/v3.13/main: temporary error (try again later)" error while trying to build one of our repositories. Odd thing is that it's only affecting one of our repositories, others are working fine on the same build server. Before anyone asks, it's a 64-bit build not 32-bit so it's not subject to the libseccomp bug in alpine 3.13. :) I'm
2021-07-08 02:56:52 <ohheyitsbrian>  at a loss for debugging this one. curling the server works fine, curling our own package mirror works fine too. Using the package mirror causes the same error as the regular alpine package host.
2021-07-08 02:57:47 <ohheyitsbrian> Forgot to mention, the error occurs when trying to `apk update` or really do anything with `apk`, like `apk add`. Thanks!
2021-07-08 02:58:27 <ohheyitsbrian> Fun fact though... while reading through the source for apk fetch, I found the easter egg for `apk fetch coffee` and `apk fetch coffee --force` :D
2021-07-08 03:30:31 <ohheyitsbrian> Ugh... solved. `docker build --network host -t ...` :facepalm:
2021-07-08 03:45:22 <niallr> hpooray
2021-07-08 13:34:10 <Aftershock> ok today I learned: apk del * works as expected and is not a good idea
2021-07-08 13:35:17 <ikke> Right... 
2021-07-08 14:54:30 <ncopa> :D
2021-07-08 14:55:52 <ncopa> reminds me of that time where i did apk del * on stage (with /etc/apk/world backed up and a copy of a static busybox) to show how to do a restore using /etc/apk/world
2021-07-08 14:56:13 <lopid> how would i discover which package provides EXTERN.h? i know the answer is perl-dev, only because i have a vm that happens to have that already installed. https://pkgs.alpinelinux.org/packages?name=*EXTERN.h&branch=v3.14&arch=x86_64 shows no results
2021-07-08 14:56:17 <ncopa> except that xfce4-terminal didnt like that at all
2021-07-08 14:56:49 <ncopa> apk info --who-owns /path/to/EXTERN.h
2021-07-08 14:57:00 <lopid> that works if the package is installed
2021-07-08 14:57:19 <ncopa> yup. if its not installed you'll have to search pkgs.alpinelinux.org
2021-07-08 14:57:38 <lopid> what query would i use? the ones i tried showed no results
2021-07-08 14:58:14 <ncopa> click on "contents" at the top
2021-07-08 14:58:17 <ncopa> https://pkgs.alpinelinux.org/contents?file=EXTERN.h&path=&name=&branch=edge
2021-07-08 14:58:54 <lopid> ah! thank you
2021-07-09 02:18:53 <bleb> i manually installed a program with its man page under /usr/local/share/man
2021-07-09 11:59:03 <Hello71> ok
2021-07-09 14:18:34 <bleb> and, i was wondering, how would i get that to show up in when i type man <programname>?
2021-07-09 14:19:11 <bleb> now i get "man: No entry for castget in the manual."
2021-07-09 14:25:42 <bleb> the man(1p) man page doesn't describe where it looks for pages or how to make it look in a particular place
2021-07-09 14:25:49 <bleb> so i'm not sure where else to look for documentation
2021-07-09 14:27:13 <danieli> bleb: I can't see any file called 'castget' included in any alpine package, where is it from?
2021-07-09 14:27:27 <danieli> generally, you want <package>-doc
2021-07-09 14:29:26 <bleb> sorry, there i posted some background info much earlier
2021-07-09 14:29:32 <bleb> <bleb> i manually installed a program with its man page under /usr/local/share/man
2021-07-09 14:29:54 <bleb> now i'm trying to get it to show up in the manual
2021-07-09 14:30:31 <bleb> looks like there is a man man page in section 1, though it defaults to the posix page from section 1p
2021-07-09 15:47:08 <trylletre> you probably want MANDATORY_MANPATH in /etc/man_db.conf or somesuch to contain /usr/local/share/man ?
2021-07-09 15:47:27 <trylletre> bleb: ^
2021-07-09 18:05:16 <nickersonm> Anyone here use lxd on Alpine?  I assume you use the package on edge/testing?
2021-07-09 18:09:27 <c705> that's the only pre-built package avaliable
2021-07-09 18:13:17 <nickersonm> Yeah, wondering if people actually using it prefer that or compile their own.  The package seems good, except some possibly unnecessary dependencies that I'll check (iproute2 for one, I have ifupdown-ng already).
2021-07-09 18:15:24 <ikke> they are complementary
2021-07-09 18:16:19 <nickersonm> ikke: oops, thanks for pointing that out.  Got iproute2 confused with ifupdown2.
2021-07-09 19:45:19 <veecue> Hey, quick question, is there any sensible alternative to "systemctl is-system-running --wait" with openrc?
2021-07-09 19:49:59 <lopid> what does that do?
2021-07-09 19:52:43 <Hello71> /bin/true
2021-07-09 19:52:44 <veecue> Checks if all services have been launched sucessfully and blocks if still starting up. Closest thing I know is "! rc-status -c" but that doesn't wait for starting services
2021-07-09 19:53:13 <Hello71> imo is-system-running is not that well defined
2021-07-09 19:53:29 <Hello71> or more precisely it's an antipattern
2021-07-09 19:53:41 <veecue> but why though?
2021-07-09 19:53:51 <invoked> the only init with an existential crisis
2021-07-09 19:55:25 <trylletre> you can wait for starting services by making your own service and depending on something you need cant you
2021-07-09 19:55:38 <trylletre> i cant think of why you would specifically need something like 'all the services are started'
2021-07-09 19:57:11 <veecue> to verify that a host has booted up again sucessfully after a reboot
2021-07-09 20:02:52 <trylletre> i think you can use a `after *` in the depend() of a service
2021-07-09 20:02:55 <trylletre> unsure how well that works
2021-07-09 20:03:23 <veecue> and then let that service touch a file in /dev/shm and wait for that? could work, but not too pretty :/
2021-07-09 20:04:20 <trylletre> since it runs last everything before it is started, so unless the priors crash you can make it do what you want to do normally
2021-07-09 20:04:36 <trylletre> im unsure on how is-system-running works in regards to things failing some time later after starting
2021-07-09 20:05:48 <veecue> Depends. If the failure occurs before the system gets "ready" it fails right away, if the failure occurs after the system has been ready for some time, it will be unnoticed
2021-07-09 20:06:13 <veecue> and the thing I want to do is resume a script executed from ssh
2021-07-09 20:08:43 <Hello71> if you want to check a service
2021-07-09 20:08:45 <Hello71> then check a service
2021-07-09 20:09:06 <Hello71> it's the same broken thinking as network-online
2021-07-09 20:17:33 <veecue> the "after *" thing seems to work, thanks a lot trylletre. Still not ideal, maybe s6-init will provide something better when it comes
2021-07-09 20:19:37 <trylletre> glad you found something working at least :)
2021-07-09 20:22:33 <invoked> in runit you'd just append something to /etc/runit/1, if i understand what veecue is trying to do
2021-07-10 02:37:22 <bleb> <trylletre> you probably want MANDATORY_MANPATH in /etc/man_db.conf or somesuch to contain /usr/local/share/man ?
2021-07-10 02:37:33 <bleb> trylletre: are you getting that from some documentation somewhere?
2021-07-10 02:37:53 <bleb> dont see anything about MANDATORY_MANPATH in man.conf(5) or mandoc.db(5)
2021-07-10 02:37:58 <trylletre> man 5 manpath
2021-07-10 03:11:54 <bleb> hmm dont have that, haven't installed man-db
2021-07-10 03:12:03 <bleb> wonder what that's for...
2021-07-10 03:15:38 <bleb> oh i do have man-db but its an older version, maybe the new version has manpath(5)
2021-07-10 03:16:08 <trylletre> maybe i just have weird things
2021-07-10 03:17:39 <trylletre> i dont have man.conf for instance, but that does mention the manpath
2021-07-10 03:18:06 <trylletre> to be honest- no idea how the man stack works myself :)
2021-07-10 03:20:29 <bleb> i was trying to figure out what is the default man search path
2021-07-10 03:20:35 <bleb> i dont have man.conf either
2021-07-10 03:21:32 <bleb> man(1) makes multiple references to "standard directories" and a "standard search path" but never says what it is
2021-07-10 03:22:50 <bleb> i tried putting MANDATORY_MANPATH /usr/local/share/man in /etc/man_db.conf and running makewhatis; still no man page for castget
2021-07-10 03:30:20 <trylletre> running mandb works for me
2021-07-10 03:31:01 <trylletre> if you echo $MANPATH do you see the path
2021-07-10 03:33:22 <trylletre> when you run a man command its using that path from the env, so if its not there it wont check the correct paths, but idk what sets it for your shell env system-wise
2021-07-10 03:48:18 <bleb> jk i thought "installed size" meant the amount currently taken up by an installed package
2021-07-10 03:48:32 <bleb> it's actually the size that _would_ be taken up if the package were installed
2021-07-10 03:48:37 <bleb> so i never had man-db
2021-07-10 03:48:58 <bleb> but after updating i for some reason have the man page manpath(5)
2021-07-10 03:51:27 <bleb> ah, in my confusion i had installed man-db-doc manually
2021-07-10 03:51:50 <bleb> still don't know the purpose of man-db, since man pages are working fine without it
2021-07-10 03:52:36 <bleb> only its hard to identify where the system gets its default man search path, or what it is
2021-07-10 03:57:31 <trylletre> if you manually set it and it works i guess everything is fine
2021-07-10 04:04:01 <atka> hello everyone! nice to see you after the move to oftc
2021-07-10 04:05:48 <bleb> trylletre: looks like man-db is one man system and mandoc is another; alpine has a package for each
2021-07-10 04:06:08 <trylletre> yep
2021-07-10 04:06:10 <bleb> hi atka
2021-07-10 04:06:42 <bleb> ohhh but the docs package requires mandoc?
2021-07-10 04:12:42 <atka> anyone install alpine on luks in the last couple months? I'm using my old notes that worked in the past but the last 2 attempts have failed to unlock after grub. anything changed?
2021-07-10 04:19:04 <atka> bleb: hello to you!
2021-07-10 08:30:03 <kevint> atka do you get any specific error?
2021-07-10 08:30:30 <kevint> I'm using alpine on luks fine right now, installed a few months ago
2021-07-10 08:31:36 <atka> I get to grub but after trying to boot linux-lts I get dumped into recovery console
2021-07-10 08:31:42 <atka> here's my notes I'm using
2021-07-10 08:31:43 <atka> https://p.teknik.io/PiLgZ
2021-07-10 08:41:07 <kevint> Hmm maybe you need `GRUB_PRELOAD_MODULES="luks cryptodisk part_gpt` and `GRUB_ENABLE_CRYPTODISK=y` ? 
2021-07-10 08:41:35 <kevint> This is the guide I followed, it uses lvm though but that still might apply: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Grub_with_UEFI
2021-07-10 08:42:04 <atka> thanks for the info
2021-07-10 08:42:24 <atka> I've never needed that before and been using alpine with luks for a few years
2021-07-10 08:42:38 <atka> but haven't done a fresh install for abit
2021-07-10 08:42:49 <kevint> Ah okay I see
2021-07-10 08:43:45 <atka> well I guess I haven't done a fresh instal since 3.13.1 whenever that was
2021-07-10 08:43:58 <atka> I'll git that guide a try and see if it works, thanks again!
2021-07-10 08:44:06 <atka> s/git/give
2021-07-10 08:44:06 <alpine-meetbot> atka meant to say: I'll give that guide a try and see if it works, thanks again!
2021-07-10 08:44:23 <atka> heh, thats pretty cool
2021-07-10 08:48:49 <kevint> No problem, hope it helps!
2021-07-10 12:34:04 <jaafarrc> I'm trying to use setup-interfaces to connect to my wireless network, but when the script scans for networks, it doesn't find mine.
2021-07-10 12:34:23 <ikke> What kind of network is it?
2021-07-10 12:34:35 <jaafarrc> It's an open network.
2021-07-10 16:16:34 <jaafarrc> Oh, by the way, I forgot to mention: when setup-interfaces lists the networks, it renders "\x00" multiple times on two lines. One has more "\x00"-s than the other.
2021-07-10 16:17:07 <jaafarrc> I don't know why that is. Maybe it is detected some network with a non-Latin name? Seems unlikely.
2021-07-10 16:18:00 <jaafarrc> And it seems to be listing a printer... not a network... what?
2021-07-10 16:18:24 <jaafarrc> Nevermind, there is no printer.
2021-07-10 16:21:05 <minimal> atka: in your set of commands, shouldn't "cryptdm= " be "cryptdm=cryptroot"?
2021-07-10 16:23:42 <minimal> atka: also kevint's suggestions are for full encryption (inc. /boot) whereas your notes are for encrypting only rootfs
2021-07-10 17:14:21 <Hello71> jaafarrc: i believe this is caused by poorly programmed hidden networks
2021-07-10 17:15:01 <jaafarrc> My network isn't hidden, though.
2021-07-10 17:15:09 <Hello71> what is your wireless card? and network config? n? a? ac?
2021-07-10 17:15:30 <jaafarrc> What do you mean "n" "a" "ac"?
2021-07-10 17:17:40 <jaafarrc> Hello71: ?
2021-07-10 18:49:57 <jaafarrc> Hey, I think I fixed my problem. I found alternative way to connect to the Wi-Fi, so, that's cool.
2021-07-10 19:45:03 <atka> minimal: "cryptdm= " can be any name you want, I made my notes generic in case someone else wanted to do something different.
2021-07-10 19:45:34 <minimal> atka: yes but I was pointing out that no value was specified for cryptdm
2021-07-10 19:46:03 <atka> minimal: I did notice that kevint's suggestion included encrypted boot, and will probabably retry and set it up that way next try.
2021-07-10 19:46:32 <minimal> atka: as I mentioned, his suggestion is for when you also want /boot encrypted
2021-07-10 19:47:00 <minimal> I've been doing a lot of LUKS testing recently so if you have a particular issue I can likely help
2021-07-10 19:47:54 <minimal> BTW there seems to be some issue with LUKS v2 (including /boot) encryption - I haven't gotten that working yet, seems like an GRUB 2.06 issue
2021-07-10 19:48:01 <atka> ok, thanks for the offer. I was just curious if something has changed since 3.13.1, as that had happened on an old void install, over the years the syntax had changed slightly
2021-07-10 19:48:21 <atka> oh, that's nice to know
2021-07-10 19:48:28 <minimal> I think its likely a GRUB 2.06 change in behaviour
2021-07-10 19:49:07 <atka> I noticed a few luks 2 issues, for instance a system that only has 1 gb of ram cannot unlock my luks 2 disks but can unlock my v1 disks no problem
2021-07-10 19:49:34 <atka> notice anything like that?
2021-07-10 19:50:25 <minimal> actually you might be hitting an issue I'm currently investigating - GRUB with encrypted root and unencrypted /boot (i.e. separate partition) where I think the grub.cfg created is not right - it need to look on the unencrypted /boot for kernel and initramfs but it seems to be looking at encrypted root partition
2021-07-10 19:51:49 <minimal> atka: I've got scripting to create encrypted disk images: https://github.com/dermotbradley/create-alpine-disk-image
2021-07-10 19:53:41 <atka> have you tried specifying --boot-directory=/boot and --efi-directory=/boot/efi in grub?
2021-07-10 19:53:59 <minimal> atka: might be related to Argon - with LUKS1 there's only PBKDF2 whereas for LUKS2 there is either PBKDF2, Argon2i and Argon2id. I think one of the Argons is the default for LUKS2 and those are *designed* to be memory intensive as part of their security
2021-07-10 19:54:26 <minimal> you could try selecting PBKDF2 for LUKS2 and it should use less memory
2021-07-10 19:54:45 <atka> minimal: thanks, that makes sense, it did give me a memory related error
2021-07-10 19:55:38 <minimal> from the cryptsetup manpage: "For Argon2i/id,  there  is  also memory cost (memory required during the process of key derivation)"
2021-07-10 19:57:16 <atka> thanks, I'm sure that is what the issue was
2021-07-10 19:57:21 <nickersonm> FWIW for my chromebook, I need to use `cryptsetup luksFormat --pbkdf-memory 131072 --type luks2 "${DEV1}${PART}2"` (not running Alpine, but using GRUB + LVM-on-LUKS)
2021-07-10 19:57:55 <nickersonm> Without the pbkdf-memory reduction from default, it can't unlock anything.
2021-07-10 20:00:17 <atka> how much ram do you have? I haven't noticed it on systems with 4gb or more but my system with 1gb had issues
2021-07-10 20:00:20 <minimal> nickersonm: haven't gotten that far with LUKS2 on Alpine - there's something fishy with GRUB 2.06 as in the grub.cfg that it creates the "insmod luks2" and "cryptomount -u" stuff is missing whereas its correctly put in grub.cfg when using LUKS1
2021-07-10 20:01:00 <nickersonm> atka: iirc my chromebook only has 2GB
2021-07-10 20:01:47 <atka> nickersonm: nice to know, that could cause some issues for sure like use in a minimal vm etc
2021-07-10 20:03:05 <minimal> atka: again from the manpage: "The parameters can be time, memory and parallel cost." Choose your weapon! :-)
2021-07-10 20:03:09 <nickersonm> minimal: ah; I have not played around much with Alpine and LUKS, as I mostly just use Alpine on VMs (plus a new physically secured NAS I'm working on now)
2021-07-10 20:03:38 <atka> minimal: time it is!
2021-07-10 20:04:10 <minimal> nickersonm: as mentioned above I've scripted the creation of Alpine disk images for VMs, Cloud, and physical machines and LUKS is one of the options for the disk images
2021-07-10 20:04:33 <minimal> so I've been testing lots of scenarios recently...
2021-07-10 20:04:43 <atka> thanks for the info minimal , nickersonm , I have to run but will be around
2021-07-10 20:05:14 <minimal> atka: have a look at the script I mentioned - you'll see the cryptsetup options I'm using
2021-07-10 20:05:26 <atka> minimal: k, will do
2021-07-10 20:08:00 <nickersonm> minimal: yeah, I have your github repo starred to investigate once I start working on deploying Alpine VMs to replace my web infrastructure (currently 3 badly aging manually managed debian instances)
2021-07-10 20:10:31 <minimal> nickersonm: what your intended VM host "mechanism"? I've only extensively tested QEMU (cli) and Virtualbox so far
2021-07-10 20:12:27 <nickersonm> minimal: current development hypervisor is Hyper-V, soon to be replaced with LXD-on-Alpine (maybe QEMU as well if I need it); deployment will be to DigitalOcean or Vultr
2021-07-10 20:14:13 <minimal> haven't looked at Hyper-V (I've no Windows boxes anymore), so its possible the kernel modules referenced in the script for Hyper-V might ned tweaked
2021-07-10 20:15:33 <nickersonm> alpine-virt boots fine with no tweaks fwiw.  I imagine I'll end up needing QEMU for fully packaging the VMs, though.
2021-07-10 20:17:03 <minimal> nickersonm: I was referring to my script where I tweak the initramfs config for various situations to only contain relevant kernel modules (so a disk image for QEMU doesn't have any Hyper-V/Virtualbox/etc drivers in its initramfs)
2021-07-10 20:17:50 <nickersonm> Oh, for initramfs, yeah.
2021-07-10 20:19:02 <minimal> plus I do have config to disable the loading of some kernel modules as well again depending on the disk image type
2021-07-10 20:20:08 <minimal> So I create a /etc/modprobe.d/disable-drivers-modules.conf file
2021-07-10 20:21:06 <nickersonm> very true to your handle :)
2021-07-10 20:23:37 <minimal> nickersonm: yes and no :-) Using cloud-init brings in python deps but for me its worth the increase in size for the flexibility that cloud-init provides. Plus it helps security, by some unmeasurable value, to prevent unrequired/unexpected kernel modules from loading
2021-07-10 20:27:37 <minimal> nickersonm: you might be interested in the remote (i.e. SSH-based) LUKS unlocking I've been working on
2021-07-10 20:34:56 <nickersonm> minimal: have you seen alpine-lift?  Provides basic cloud-init functionality.  I've not looked in to it in detail yet, but it's something I have bookmarked to consider for my future web infrastructure.
2021-07-10 20:36:49 <minimal> nickersonm: looked at it (and all the other cloud-init alternatives) in the past. It (and the others) may be more lightweight that cloud-init but none of them have anywhere near the degree of functionality and breadth of DataSources as cloud-init
2021-07-10 20:38:01 <nickersonm> Yeah, definitely much simpler
2021-07-10 20:38:08 <minimal> if there was a Golang equivalent (with most/all functionality and DataSources) of cloud-init I'd be interested
2021-07-10 21:05:55 <Razapui23> Hi. Just out of the box after installing alpine, the keyboard barely works (no shift key, actuall shift key is altgr).
2021-07-10 21:06:14 <Razapui23> problem seems to be I'm missing the kbd-bkeymaps package
2021-07-10 21:07:00 <lopid> did you answer setup-keymap correctly?
2021-07-10 21:07:17 <Razapui23> Yes, but setup-keymap has erros in it.
2021-07-10 21:07:29 <Razapui23> As in, there is a error message it could not cd
2021-07-10 21:07:59 <lopid> indeed sounds like the installation went awry
2021-07-10 21:08:00 <Razapui23> to //usr/share/bkeymaps (yes, //")
2021-07-10 21:20:23 <Razapui23> I think there is a bug in the de-T3 keymap
2021-07-10 21:20:43 <Razapui23> de-de works fine. de-T3 has broken shift keys.
2021-07-11 02:59:49 <atka> hello, anyone running btrfs on luks as their main filesystem?
2021-07-11 03:36:20 <atka> nm, got it to boot finally
2021-07-11 03:51:23 <atka> on another topic, anyone running any wayland tiling wm besides sway? I am kinda wanting to try wayland again but I'm more of a dwm fan than i3 fan.
2021-07-11 06:24:11 <Misthios> I use river
2021-07-11 06:29:44 <atka> do you find it to be pretty stable?
2021-07-11 06:34:37 <atka> Misthios: and no pango from the looks of it which is nice.
2021-07-11 07:06:24 <Misthios> Yeah havent had a problem yet, only problem i had was alpine not being able to build a newer version due to a newer zog version which requires a newer llvm version
2021-07-11 07:09:12 <atka> Misthios: ah, ok, thanks for the feedback. I just really want dwm for wayland but the current alternatives seem to have issues. I might give river a try though
2021-07-11 07:22:55 <Misthios> Hf
2021-07-11 10:27:54 <donoban> atka: I use it bot with plain boot partition
2021-07-11 10:28:31 <donoban> ah you fixed it great
2021-07-11 12:32:59 <lopid> how can i tell what changed in alpine-baselayout 3.2.0-r15 -> 3.2.0-r16?
2021-07-11 12:33:22 <lopid> having difficulty identifying it from https://git.alpinelinux.org/aports/log/main/alpine-baselayout
2021-07-11 12:49:12 <graywolf> lopid: git diff $(git log --oneline -2 -Gpkgrel=15 -- main/alpine-baselayout | tail -n 1 | cut -d\  -f1)..$(git log --oneline -2 -Gpkgrel=16 -- main/alpine-baselayout | tail -n 1 | cut -d\  -f1) -- main/alpine-baselayout/
2021-07-11 12:49:36 <graywolf> I think that is correct? Also, there is probably more easy way, so someone will possibly enlighten us both
2021-07-11 12:50:10 <lopid> without checking out the source code
2021-07-11 12:51:29 <graywolf> ah no idea then
2021-07-11 12:52:05 <lopid> thanks anyway
2021-07-11 13:27:17 <jaafarrc> How do I check if a certain package is installed with apk?
2021-07-11 13:27:55 <lopid> apk info -I package
2021-07-11 13:32:18 <jaafarrc> lopid: I just get a help page.
2021-07-11 13:32:49 <lopid> list, ssorry
2021-07-11 13:35:00 <jaafarrc> Okay, thanks.
2021-07-11 16:30:07 <jaafarrc> Hey, I'm trying to use Sway, and it doesn't work at all. If start Sway, it prints two lines and completely freezes the tty; I can't go to any other tty nor can I Ctrl+C or Ctrl+D.
2021-07-11 16:44:09 <jaafarrc> Nevermind, I fixed the freezing but Sway still doesn't work.
2021-07-11 20:51:46 <atka> donoban: would you mind sharing your btrfs subvolume scheme? I'm tryning to come up with a decent flat layout. I'm aiming for having snapshots of / and /home, but was thinking of having a subvolume for docker containers and snapshots as well. Do you mount your top-level anywhere?
2021-07-12 10:26:27 <donoban> atka: my scheme is very simple, I have @ROOT and @HOME, initially I wanted to use @ROOT based on 3.13 and @EDGE on edge but docker created his own subvolume in '/var/lib/docker' inside @ROOT, so if you want to have an indepently docker subvolume, create it and adjust your fstab before instaling and run docker stuff
2021-07-12 10:29:54 <donoban> atka: I remember that I had some problem for enable compression, like fstab option being ignored
2021-07-12 13:59:41 <lopid> in /etc/profile, HOSTNAME is set, but not marked for export
2021-07-12 16:27:11 <jaafarrc> https://termbin.com/0qz3
2021-07-12 16:27:28 <jaafarrc> This is the error I get when I try to run any Qt5 Flatpak.
2021-07-12 16:35:49 <lopid> why wayland?
2021-07-12 16:36:18 <jaafarrc> Because I use it.
2021-07-12 16:42:33 <jaafarrc> lopid: I use Sway.
2021-07-12 16:43:32 <jaafarrc> By the way, I could just run the Qt5 apps with Xwayland, but that's a hacky workaround I don't like, and the apps I'? using do support Wayland natively.
2021-07-12 16:44:24 <jaafarrc> I'mz
2021-07-12 17:02:31 <invoked> there's some QT env var you need to set, iirc, one minute
2021-07-12 17:02:58 <invoked> QT_QPA_PLATFORM=wayland
2021-07-12 17:03:00 <invoked> do you set that?
2021-07-12 17:03:49 <invoked> with flatpak there is some way you have to pass it on the commandline when you run it
2021-07-12 17:04:09 <invoked> i'm not in a position to dig that up, but it's something like --env=var=foo
2021-07-12 17:06:00 <invoked> but tbh, if the flatpak is built properly you shouldn't have to do that
2021-07-12 17:20:24 <jaafarrc> invoked: Yes, I've set it already.
2021-07-12 17:21:16 <jaafarrc> Actually, if I *un*-set it, it would just render with Xwayland after failing to use Wayland directly.
2021-07-12 17:25:51 <invoked> i thought you were trying to do pure wayland
2021-07-12 17:26:13 <invoked> (or i had assumed you had xwayland disabled, bad assumption either way)
2021-07-12 17:26:31 <jaafarrc> Yes, I am doing pure Wayland.
2021-07-12 17:26:48 <jaafarrc> invoked: I'm just telling you what happens if I un-srt.
2021-07-12 17:26:54 <jaafarrc> unset it*
2021-07-12 17:26:54 <invoked> gotcha
2021-07-12 17:27:55 <jaafarrc> --env=var=foo doesn't seem to do anything
2021-07-12 17:47:16 <invoked> flatpak run --env=QT_QPA_PLATFORM=wayland <package>
2021-07-12 17:47:30 <invoked> i think that's how it works, but i'm not checking :)
2021-07-12 17:50:02 <invoked> although, if the package isn't built with wayland support then it's not going to magically work
2021-07-12 17:57:37 <jaafarrc> That also didn't work. :(
2021-07-12 18:11:58 <invoked> like i said, it's one thing if upstream says they support wayland. how it's packaged on flathub/et al is something else.
2021-07-12 18:12:35 <invoked> i didn't look at your paste. flatpak (at the time you install something) will probably explicitly list if it supports wayland (along with X11 and what else)
2021-07-12 18:12:54 <invoked> if it doesn't, it probably won't
2021-07-12 18:22:04 <atka> donoban: thanks a lot, that is very useful info!
2021-07-12 18:22:45 <jaafarrc> invoked: It lists that? How do I check?
2021-07-12 18:29:32 <invoked> jaafarrc: it's at install time. flatpak will list out what permissions it needs for the package before you say Y. wayland will be listed if it's explicitly supporting it.
2021-07-12 18:29:53 <jaafarrc> Alright.
2021-07-12 18:30:00 <invoked> it'll list other things like, "ipc" "network" so on
2021-07-12 19:14:53 <atka> donoban: fyi when it comes to mount options and subvolumes, only mount options on the first mounted subvolume on the btrfs filesystem will be honored. so if you had a subvolume without compression mount options get mounted first the others won't be mounted with the compression options. also the compression won't apply to current files only new ones after the fact. I dunno if thats what you were running into but just letting you kn
2021-07-12 19:48:33 <jaafarrc> invoked: Thank you so much.
2021-07-12 23:49:28 <jaafarrc> how do I get tput?
2021-07-12 23:58:47 <invoked> jaafarrc: tput is part of ncurses
2021-07-12 23:59:00 <jaafarrc> Ok, thanks.
2021-07-13 13:42:16 <ncopa> apk add cmd:tput
2021-07-13 13:58:56 <invoked> yeah, i should have put it that way.
2021-07-13 14:00:13 <ikke> explicitly adding cmd:<cmdname> can cause issues when upgrading though
2021-07-13 14:00:21 <ikke> ran into that a few times
2021-07-13 14:14:32 <lopid> might want to add -i to that
2021-07-13 14:14:39 <lopid> and/or -s
2021-07-13 14:57:55 <telmich> hello
2021-07-13 14:58:12 <telmich> is it possible that something recently broke with regards to checkpath / /run?
2021-07-13 14:58:21 <telmich> Just upgraded knotd and I am getting this:
2021-07-13 14:58:33 <telmich>  * checkpath: /run/knot/: could not open knot: No such file or directory
2021-07-13 15:05:27 <telmich> nvm, I was on an old version from 3.12
2021-07-13 18:55:47 <telmich> Sorry for asking again, but does anyone have a clean solution to specificy `mount --make-shared /sys` and `mount --make-shared /run` to take place every boot?
2021-07-13 18:56:49 <ikke> I guess there is no mount option for that that you can set in fstab?
2021-07-13 19:07:07 <telmich> ikke: you guessed correctly
2021-07-13 19:08:19 <telmich> Also /run and /sys are probably be mounted outside of the fstab approach
2021-07-13 19:09:23 <ikke> there is a sysfs service
2021-07-13 19:10:11 <telmich> and that has sysfs_opts
2021-07-13 19:10:23 <telmich> hmm, seems hardcoded
2021-07-13 19:11:00 <ikke> I think openrc itself handles /run
2021-07-13 19:42:07 <telmich> Kinda fixed: https://code.ungleich.ch/ungleich-public/ungleich-k8s/-/tree/master/alpine-linux
2021-07-13 19:48:46 <proycon> I'm having an odd issue resolving hostnames in a local (wireless) network, ping and other tools can't resolve the hostname, but tools like nslookup, dig, and host have no issue when I query the exact nameserver I specified in /etc/resolv.conf: https://download.anaproy.nl/resolveissue.txt 
2021-07-13 19:49:00 <proycon> Does anyone see what I'm missing?
2021-07-13 19:50:34 <lopid> epson6fb191.home is resolved by the nameserver?
2021-07-13 19:50:56 <lopid> looks like it
2021-07-13 19:51:51 <xordspar0> ping and other tools use the libc DNS implementation, while nslookup, dig, host, and the rest of the BIND tools use an internal implementation.
2021-07-13 19:51:52 <lopid> i have not had this problem, using dnsmasq
2021-07-13 19:52:29 <proycon> hmmm, but the internal implementation should also simply rely on /etc/resolv.conf right?
2021-07-13 19:52:30 <xordspar0> Another tool that uses the libc impl is `getent ahosts`. It may give you move details.
2021-07-13 19:52:46 <proycon> ah, thanks for the tip, I didn't know that one
2021-07-13 19:54:03 <proycon> the fact that NetworkManager set up the connection  shouldn't be an issue right?
2021-07-13 19:54:22 <proycon> I didn't see anything out of the ordinary in its configuration either
2021-07-13 19:57:57 <xordspar0> I'm not sure, but I would be surprised if that caused a problem.
2021-07-13 19:58:29 <xordspar0> musl libc DNS does have a few limitations, but I can't tell if those apply in your case. https://wiki.musl-libc.org/functional-differences-from-glibc.html#Name-Resolver/DNS
2021-07-13 19:58:34 <proycon> the odd thing is also that everything else (external hostnames) resolves just fine, it's just the internal ones
2021-07-13 19:59:03 <proycon> ah thanks, I'll read that
2021-07-13 20:01:30 <proycon> the paragraph about 'domain' and 'search' could be relevant
2021-07-13 20:08:40 <xordspar0> Yes...Alpine has had search support since 3.4, circa 2017. I'm not sure about the ndots part because I don't know where that is configured.
2021-07-13 20:09:54 <proycon> in resolv.conf it seems
2021-07-13 20:10:37 <proycon> well, I'll have to try solve it further tomorrow cause the remote system is down now...
2021-07-13 20:13:22 <xordspar0> Ah, looking at the source code it seems the default ndots is 1 if it's not set. So that means that only domains with 0 dots get searched normally unless you add ndots configuration to your resolv.conf.
2021-07-13 20:14:12 <xordspar0> Hopefully that gives you some things to try. Good luck tomorrow.
2021-07-13 20:14:22 <lopid> that explains why i don't have a problem - using no dots
2021-07-13 20:15:33 <proycon> I guess the domain "home" counts as 0 dots
2021-07-13 20:16:13 <proycon> at least I have some things to try when the system is back up again
2021-07-13 20:33:26 <ijr22> does anyone here use alpine for desktop
2021-07-13 20:34:24 <telmich> that was a short visit, and I wanted to say yes to irj22
2021-07-13 20:34:39 <yyp> ^ same
2021-07-13 20:34:56 <lopid> you can say it to me if you'd like
2021-07-13 20:35:01 <yyp> I just installed it today, everything works good so far
2021-07-13 20:35:48 <telmich> lopid: I am using Alpine on almost all desktops!
2021-07-13 20:35:56 <lopid> cor!
2021-07-13 20:35:58 <telmich> (even though technically they are notebooks...)
2021-07-13 20:36:11 <lopid> why did you both choose alpine?
2021-07-13 20:36:53 <yyp> lopid: because apk is great :)
2021-07-13 20:37:37 <lopid> you chose it for the package manager?
2021-07-13 20:38:39 <yyp> Not only because of the package manager of course, but that was the main feature for me
2021-07-13 20:38:49 <yyp> The community is really good as well
2021-07-13 20:39:00 <lopid> :)
2021-07-13 20:43:35 <yyp> err, maybe not because of a package manager, but because of good packaging in general. I just want to make a package for everything I touch because it works so nicely unlike my previous Arch + AUR + duct tape setup :D
2021-07-13 20:45:06 <lopid> one day i'll get around to that, too
2021-07-14 01:10:59 <eletrotupi> hm, gitlab.alpinelinux.org is getting gateway timeout to me, anyone else experiencing issues?
2021-07-14 01:32:34 <fluix> same
2021-07-14 02:35:40 <bleb> busybox ash has its own implementation of test
2021-07-14 02:35:50 <bleb> is there any way to get documentation for this?
2021-07-14 02:38:53 <bleb> ah so with coreutils installed, /usr/share/man/man1 contains both test.1.gz and test.1p.gz
2021-07-14 02:39:24 <bleb> test.1.gz has the coreutils man page, i assume test.1p.gz has the posix man page, which busybox probably approximates
2021-07-14 02:39:58 <bleb> but is there any way to access test.1p.gz through the man command?
2021-07-14 02:40:23 <raven523> aman 1p test
2021-07-14 02:40:29 <raven523> *man, not aman
2021-07-14 02:40:53 <bleb> man: No entry for test in section 1p of the manual.
2021-07-14 02:40:58 <bleb> man 1p test works for you?
2021-07-14 02:41:08 <raven523> oh right the system I tried that on is gentoo
2021-07-14 02:41:39 <bleb> do you know if it has mandoc or man-db
2021-07-14 02:41:54 <raven523> probably man-db
2021-07-14 02:43:01 <raven523> it looks like busybox test is documented in man busybox
2021-07-14 02:43:54 <raven523> ... or not. The other commands are documented, but not test
2021-07-14 02:44:12 <bleb> lol
2021-07-14 02:49:13 <bleb> after switching to man-db, man 1p test works
2021-07-14 02:49:36 <bleb> but ofc man-db is not supported by the docs package so its not worth it to switch
2021-07-14 02:55:50 <Hello71> did you try man /usr/share/man/man1/test.1p.gz
2021-07-14 02:56:08 <bleb> another point of confusion: how does busybox decide when to use a builtin and when to check if $PATH leads to something else
2021-07-14 02:58:24 <bleb> maybe some of the "applets" are actually shell builtins, and ash uses those without checking in $PATH
2021-07-14 07:14:45 <omni> test.1p.gz comes from the package man-pages from https://www.kernel.org/doc/man-pages/ and test.1.gz from the package coreutils-doc
2021-07-14 07:23:44 <omni> bleb: I don't think ash has test as a builtin, rather it will use what's in your path, which could be busybox test
2021-07-14 07:24:53 <bleb> omni: i'm aware of the origin of the different man pages.  problem is that mandoc can't access the 1p page if both are installed.
2021-07-14 07:24:54 <omni> from within ash, try 'apk info -W $(which test)' or 'apk info -W $(which \[)' and you'll know
2021-07-14 07:26:17 <bleb> omni: this points me to coreutils test, which supports -l to get the length of a string, but the version that ash is actually using doesn't support it
2021-07-14 07:27:15 <bleb> so it seems some busybox applets are shell builtins which prevent the use of an external program of the same name
2021-07-14 07:27:40 <omni> that's odd
2021-07-14 07:28:05 <yyp> ${#var} is the proper way to get length of the string
2021-07-14 07:29:08 <omni> https://busybox.net/downloads/BusyBox.html doesn't say much more for test than "Check file types, compare values etc. Return a 0/1 exit code depending on logical value of EXPRESSION"
2021-07-14 07:30:52 <omni> yyp: right
2021-07-14 07:31:14 <bleb> omni: i wonder if that page eventually filters down to the busybox man page
2021-07-14 07:31:57 <bleb> current version doesn't document test
2021-07-14 07:32:46 <bleb> busybox is underdocumented anyway
2021-07-14 07:33:12 <bleb> and i think the builtin superseding a program in $PATH is pretty normal, at least it says so in test(1)
2021-07-14 07:33:35 <bleb> mandoc being unable to find test(1p) seems like a real problem though
2021-07-14 07:34:32 <yyp> I have a patched version of man-pages package, where mandoc is able to find {0,1,3}p pages
2021-07-14 07:34:51 <bleb> oh cool
2021-07-14 07:34:56 <bleb> what had to change
2021-07-14 07:35:31 <bleb> (i assumed it was a mandoc problem because man-db was able to find them)
2021-07-14 07:37:07 <yyp> I removed a part of prepare that moved man{0,1,3}p into man{0,1,3} (removing p)
2021-07-14 07:38:27 <omni> I installed the man-pages package and get the 1p pages, but I don't have coreutils-doc with competing section 1 pages installed
2021-07-14 07:39:36 <omni> so I get TEST(1P) when 'man test' or 'man 1 test'
2021-07-14 07:39:46 <yyp> (just in case, my patched version is here: https://git.sr.ht/~yerinalexey/apkbuilds/tree/main/item/local/man-pages/APKBUILD)
2021-07-14 07:40:27 <bleb> yyp: what happens if you install coreutils-doc?
2021-07-14 07:40:37 <omni> but 'man -s 1p test' and it says "man: No entry for test in section 1p of the manual."
2021-07-14 07:41:23 <bleb> omni: yeah that's what i get, except with coreutils-doc i simply can't access test(1p)
2021-07-14 07:42:22 <omni> bleb: because you'll get the coreutils-doc test(1) instead, right?
2021-07-14 07:43:49 <bleb> omni: right
2021-07-14 07:44:15 <yyp> bleb: even if you have coreutils-doc, you can access test(1p) with `man 1p test`
2021-07-14 07:44:15 <bleb> and with man-db instead of mandoc, man 1p test works
2021-07-14 07:45:45 <bleb> yyp: nice
2021-07-14 07:48:18 <bleb> yyp: is there any downside to your change to the man-pages package?
2021-07-14 07:48:26 <bleb> have you tried getting it merged
2021-07-14 07:51:06 <yyp> It may be broken with man-db; no, didn't try merging it upstream
2021-07-14 08:02:25 <bleb> i might try it with man-db tomorrow to see
2021-07-14 08:03:16 <bleb> yyp: your repo's readme speaks of "prolog man pages," I think you may have meant POSIX
2021-07-14 10:21:19 <frojnd> Anyone uses fish shell ona remote server?
2021-07-14 10:21:28 <frojnd> I can't seem to configure the shell
2021-07-14 10:26:02 <frojnd> when writing in fish_config I see this: https://imgur.com/a/QOqKmUX
2021-07-14 10:54:44 <drgibbon> hello all, what is the "best" ARM download for raspi 2 model b? armv7?
2021-07-14 10:55:02 <drgibbon> or just armhf?
2021-07-14 10:56:53 <ikke> A 900MHz quad-core ARM Cortex-A7 CPU
2021-07-14 10:56:58 <ikke> so armv7 should work
2021-07-14 10:58:09 <drgibbon> ok thanks
2021-07-14 10:58:22 <drgibbon> is there some advantage to armv7 over armhf?
2021-07-14 10:58:30 <ikke> More packages
2021-07-14 10:58:44 <ikke> armhf is loosing support upstream, so more packages are disabled
2021-07-14 10:59:49 <ikke> And I guess architectural advantages as well
2021-07-14 11:01:40 <drgibbon> armv7 it is then!
2021-07-14 11:13:55 <drgibbon> thanks, time to test out Alpine
2021-07-14 17:01:34 <HiHo> Hello, what kernel version does 3.14.0 come with?
2021-07-14 17:04:17 <minimal> Hiho: https://pkgs.alpinelinux.org/packages?name=linux-lts&branch=v3.14
2021-07-14 17:04:40 <HiHo> Thanks!
2021-07-14 17:28:22 <ijohnr> how can I bring up networking without waiting for udhcpc at boot
2021-07-14 17:30:07 <donoban> do you mean setup manual configuration?
2021-07-14 17:30:30 <donoban> static ip/gateway/etc...
2021-07-14 17:30:57 <ijohnr> No, when I boot I have to wait for udhcpc to discover, is there a way that that can be daemonized or something so it runs in the background instead and continues booting
2021-07-14 17:32:14 <Misthios> Change runlevel
2021-07-14 17:32:24 <ijohnr> To what?
2021-07-14 17:32:38 <Misthios> Ehm default?
2021-07-14 17:32:50 <ijohnr> Thanks!
2021-07-14 17:34:52 <minimal> ijohnr: from memory udhcpc does run in background (eventually) after a few tries in the foreground. If you have static network config defined in /etc/network/interfaces then it udhcpc shouldn't be running
2021-07-14 17:36:33 <donoban> he left :\
2021-07-14 17:40:21 <minimal> ijohnr: from memory udhcpc does run in background (eventually) after a few tries in the foreground. If you have static network config defined in /etc/network/interfaces then it udhcpc shouldn't be running
2021-07-14 17:40:44 <ijohnr> Is there a way to have it immediately run in the background
2021-07-14 17:40:52 <ijohnr> I am using dhcp
2021-07-14 17:41:24 <minimal> ijohnr: and then what about all the other services that depend on "net"?
2021-07-14 17:41:57 <ijohnr> I don't think I have any other services that depend on it
2021-07-14 17:43:01 <ijohnr> and if I do I can disable them
2021-07-14 17:44:14 <minimal> ijohnr: is it really a problem to wait for 1 or 2 secs for DHCP to complete (assuming you actually have a DHCP server)
2021-07-14 17:44:45 <ijohnr> This is not a server, it is a desktop, so it becomes a problem when I have to wait that long everytime I boot
2021-07-14 17:46:56 <minimal> is 1 or 2 seconds "long"?
2021-07-14 17:47:13 <ijohnr> Long enough to be nuisance, especially on a laptop
2021-07-14 17:47:31 <minimal> I don't see an option for udhcpc to do what you want, you could install one of the the DHCP clients that may provide a way
2021-07-14 17:47:47 <minimal> s/the the/the other/
2021-07-14 17:47:47 <alpine-meetbot> minimal meant to say: I don't see an option for udhcpc to do what you want, you could install one of the other DHCP clients that may provide a way
2021-07-14 17:49:17 <ijohnr> ok, is there a way to use dhcpcd with /etc/network/interfaces
2021-07-14 17:49:33 <minimal> yes, install it
2021-07-14 17:50:02 <minimal> assuming you're using a recent Alpine release (with ifupdown-ng as it autodetects multiple DHCP clients)
2021-07-14 17:51:04 <ijohnr> Is there a way to make dhcpcd run in the background with the networking service?
2021-07-14 17:51:19 <minimal> look at /usr/libexec/ifupdown-ng/dhcp script - it detects dhcpdc, dhclient, and udhcpc. I'm using dhcpdc here
2021-07-14 17:52:12 <minimal> oops, dhcpcd
2021-07-14 17:52:38 <ijohnr> because dhcpcd doesn't fork to the background immediately be default
2021-07-14 17:53:14 <minimal> I don't see any option for dhcpcd to background initially but booting a VM here a min ago it literally got its IP via DHCP in approx 1 second, hardly a long time to wait...
2021-07-14 17:53:31 <ijohnr> is there any client that can?
2021-07-14 17:54:49 <minimal> you're spending more time trying to resolve this "problem" that you'll probably be waiting for DHCP at boot for the next 10 years......
2021-07-14 17:54:57 <minimal> s/that/than/
2021-07-14 17:54:57 <alpine-meetbot> minimal meant to say: you're spending more time trying to resolve this "problem" than you'll probably be waiting for DHCP at boot for the next 10 years......
2021-07-14 17:55:47 <ijohnr> I'm a perfectionist, it's why I am using linux
2021-07-14 17:58:26 <ijohnr> All im looking for is a way do make dhpc immediately fork into the background
2021-07-14 17:59:03 <minimal> well either the existing software supports that or it doesn't - in which case you'll have to rewrite it
2021-07-14 17:59:26 <minimal> neither udhcpc nor dhcpcd appear to support *immediate* backgrounding
2021-07-14 17:59:51 <ijohnr> I guess I'll try to modify udhpc or dhcpcd and upload a fork request upstream
2021-07-14 18:00:34 <Hello71> udhcpc &
2021-07-14 18:00:59 <minimal> or speed up your DHCP server?
2021-07-14 18:01:24 <ijohnr> Hello: YES, how can I make /etc/networking/interfacing run that automatically
2021-07-14 18:01:42 <ijohnr> minimal: its on a netgear router, that's not really an option
2021-07-14 18:01:59 <minimal> ijohnr: I already pointed you to the ifupdown-ng script that runs udhcpc - any changes would go there
2021-07-14 18:04:24 <ijohnr> ok I'll try it now
2021-07-14 18:11:07 <ijohnr> ok, I got it to run in the background, how can I make the output of dhcpcd silent
2021-07-14 18:11:21 <ijohnr> I've already tried adding > /dev/null at the end
2021-07-14 18:11:27 <fluix> >/dev/null 2>&1
2021-07-14 18:11:42 <ijohnr> thank you
2021-07-14 18:15:41 <ijr_> YES! that worked, alpine is suprisingly good on the desktop
2021-07-14 18:30:59 <ijr_> Can you setup dwm with a custom config on apkbuild
2021-07-14 20:02:31 <ijohnr> Even though I added acf-openssh the tab isn't showing up in acf
2021-07-14 20:02:54 <jaafarrc> Hey, how do I get QtGtkStyle? It's like some sorta plugin for Qt5 that makes the Qt theme mimic the GTK theme.
2021-07-14 20:04:51 <ijohnr> Do you mean /usr/include/QtGui/QGtkStyle
2021-07-14 20:05:12 <jaafarrc> I guess.
2021-07-14 20:05:54 <ijohnr> its in qt-dev
2021-07-14 20:06:40 <jaafarrc> I can't seem to find that package.
2021-07-14 20:07:22 <ikke> https://pkgs.alpinelinux.org/contents?file=QGtkStyle&path=&name=&branch=edge
2021-07-14 20:07:30 <ikke> Cannot find a package that has it either
2021-07-14 20:07:36 <ijohnr> https://pkgs.alpinelinux.org/package/v3.3/main/x86/qt-dev
2021-07-14 20:07:44 <ikke> v3.3...
2021-07-14 20:07:50 <ijohnr> https://pkgs.alpinelinux.org/contents?repo=main&page=27&arch=x86&branch=v3.3&name=qt-dev
2021-07-14 20:07:56 <ijohnr> oh, wrong version
2021-07-14 20:07:58 <ikke> that's ancient :D
2021-07-14 20:12:16 <jaafarrc> ijohnr: What's the right version, then?
2021-07-14 20:13:03 <ijohnr> well it's not in 3.14 or edge so I dont know
2021-07-14 20:20:25 <jaafarrc> Aw, man.
2021-07-14 20:27:17 <invoked> "aw, man" -- classic linux user's lament
2021-07-14 20:28:24 <jaafarrc> True.
2021-07-14 20:50:41 <ijohnr> Is there an official forum for alpine linux?
2021-07-14 20:52:11 <ikke> No\
2021-07-14 20:52:30 <ijohnr> Why not?
2021-07-14 20:52:59 <ikke> We had one, but it was a wasteland
2021-07-14 20:53:50 <ikke> There is a mailing list
2021-07-14 20:54:02 <ijohnr> Ah, too bad, but I guess irc and the wiki work pretty well, plus it seems like a lot of poeple post a question on a forum and just never follow up, so maybe it's for the best
2021-07-14 20:55:32 <invoked> a web forum would work if all posts older than a year were purged continuously
2021-07-14 20:55:40 <invoked> well, possibly.
2021-07-14 20:56:31 <lopid> even those with useful answers?
2021-07-14 20:56:53 <invoked> that's stackexchange's job. dispense answers from 10 years ago and let users blindly apply them
2021-07-14 21:01:30 <ijohnr> I have still had some good answers from the arch and gentoo forums from the late 2000's
2021-07-14 21:25:24 <ksr> Hi, Any idea what gid: 8, group: mem is used for? https://git.alpinelinux.org/alpine-baselayout/tree/group
2021-07-15 00:09:35 <ksr> What is the usecase of system group: mem (gid:8) in Alpine Linux?
2021-07-15 06:36:48 <iggy> nothing in my alpine systems have gid=8
2021-07-15 08:17:39 <danieli> ksr: what does ls -l /dev/mem show?
2021-07-15 08:17:55 <danieli> I'm guessing it's the same as kmem in debian
2021-07-15 08:18:13 <lopid> root:root in my alpine
2021-07-15 08:18:44 <ikke> Can be a remnant of grsec?
2021-07-15 08:19:26 <lopid> "sudo find / -group mem 2> /dev/null" = no results
2021-07-15 09:00:16 <danieli> it's a remnant of <something> for sure, does git blame say anything interesting?
2021-07-15 09:05:45 <ikke> Was imported into aports already with that group
2021-07-15 09:05:52 <ikke> e1a9254211b885e6e44538e53dc3f9f341f261f0
2021-07-15 09:10:45 <donoban> I have root:kmem
2021-07-15 09:36:45 <thresh> hi.  I have a question on virtual packages handling with apk. I have a package that depends on a virtual provides, that is fulfilled by one or more "real" packages.  However when I try to install that said package, I get "unsatisfiable constraints".  Please check https://gist.github.com/thresheek/f0a8283559254f1b4bef5139646e5694 for actual output.  I wonder if there can be a workaround to force apk 
2021-07-15 09:36:51 <thresh> to choose the most recent version?
2021-07-15 09:37:37 <thresh> (those packages are coming from my own repositories, so I can fix them on my own)
2021-07-15 09:38:01 <ikke> Does adding a provider_priority help?
2021-07-15 09:42:55 <thresh> ikke, thanks! that looks like just what I'd need.  I guess I can bump it the same as pkgrel and then downgrade when a new pkgver arrives.
2021-07-15 10:06:19 <donoban> does anybody know what is the properly way for handling power (basically put the computer to sleep) with sway or other minimal desktops? I would like to only put it to sleep if it's running from battery
2021-07-15 10:09:21 <lopid> perhaps xfce4-settings-manager can do it for you
2021-07-15 10:09:37 <donoban> uhM
2021-07-15 10:09:49 <lopid> though that might require an xfce power daemon…
2021-07-15 10:10:36 <donoban> I wonder if using swayidle I can run any command
2021-07-15 10:11:05 <donoban> so maybe a simple solution is a script, if running_batteri, pm-suspend
2021-07-15 10:11:51 <donoban> I prefer to avoid another desktop dependency
2021-07-15 10:13:10 <lopid> maybe something in /proc and/or /sys can tell you that
2021-07-15 10:14:15 <donoban> /sys/class/power_supply/BAT0/status
2021-07-15 13:13:15 <caskd> does anyone have experience with booting the ipxe kernel from grub?
2021-07-15 13:13:41 <caskd> trying to get alpine to boot on a device without putting the minimal base onto it
2021-07-15 13:31:36 <getreal> Can't decide between gentoo and alpine for my desktop. I want to try replacing windows entirely on it, and I assume doing so requires enough applications I might prefer being able to run glibc applications
2021-07-15 13:32:42 <getreal> Alpine still uses the LTS kernel, right?
2021-07-15 13:36:05 <lopid> 5.10.43 right now
2021-07-15 13:36:46 <ikke> so, yes
2021-07-15 13:37:22 <getreal> I guess not everything's so bad about LTS; does that include wireguard, yet?
2021-07-15 13:37:34 <ikke> yes
2021-07-15 13:37:43 <getreal> i can't think of any new features i'd care about
2021-07-15 13:39:17 <lopid> i had musl trouble only with tinyfugue, but i don't use it for my desktop
2021-07-15 13:39:31 <getreal> and has anyone ever done a daily driver installed in data-disk mode instead of system-disk mode?
2021-07-15 13:39:56 <getreal> does the performance increase even matter with modern kernels caching so much in ram?
2021-07-15 13:40:56 <ikke> run-in-ram is not primarly meant for performance
2021-07-15 13:42:32 <getreal> so if i run it in ram, does that mean ALL filesystem changes sit in ram until lbu? even stuff like bookmarks from firefox?
2021-07-15 13:42:40 <getreal> or do those touch disk in data-disk mode
2021-07-15 13:46:57 <ikke> if your home is mounted on a disk, then they are persisted
2021-07-15 13:47:24 <ikke> root is mounted as tmpfs, so anything that is not mounted on a disk is ephemeral unless persisted with LBU
2021-07-15 15:38:28 <donoban> getreal: if you need some glibc dependant apps you can use flatpak/lxc/docker/vms...
2021-07-15 15:38:54 <getreal> Really? Can I run appimage garbage using that kind of thing?
2021-07-15 15:40:02 <donoban> uhM AppImage should work too
2021-07-15 15:42:56 <invoked> https://github.com/AppImage/AppImageKit/issues/1015
2021-07-15 15:44:05 <donoban> hehe maybe don't :D
2021-07-15 15:48:55 <invoked> i don't get the decision criteria between alpine and gentoo though. musl isn't first class with gentoo, and besides everything else being done very differently (openrc aside)
2021-07-15 15:49:31 <invoked> "as a desktop" is overly broad when compared these two systems.
2021-07-15 15:49:38 <getreal> invoked: both really provide a lot of decent choice, and the groups behind them are usually wildly more helpful than any other group i've seen online
2021-07-15 15:50:44 <getreal> i can't stand the arch crowd (who can?), and anything with systemd as its default is obviously out of the question. i've been burned by windows for the last time, so why would i take the same dive with sinisterd?
2021-07-15 15:52:43 <invoked> sure, if it's choice you're after, gentoo is less opinionated and there are more options there. blueness is a good steward also. that said i think alpine out of the box just works better.
2021-07-15 15:52:51 <invoked> as a desktop though, it's a whole other thing.
2021-07-15 15:54:28 <getreal> aye, though like i've had in the past, alpine was perfectly fine on my laptop, doing all i asked it to do
2021-07-15 15:54:49 <getreal> the desktop might be a bit more demanding, but as long as amd drivers work at-all i don't think i'll have many problems
2021-07-15 15:55:16 <getreal> the gentoo for desktop thing was just if i got really dedicated to choosing everything, but alpine's packages are almost always new enough that it's not important to compile
2021-07-15 15:56:49 <invoked> yeah, like i said musl isn't first class on gentoo. either way, a desktop on a musl system, you should be ready to solve your own problems
2021-07-15 15:58:38 <getreal> in some way, none of software development is ever solving your own problem.
2021-07-15 15:59:27 <getreal> as long as VMs work, i presume it's not impossible on a desktop to do most of what i care about regardless of what's going on.
2021-07-15 16:02:17 <graywolf> Hello, I've installed bash on alpine 3.14 with bash-completion and getting: -bash: syntax error near unexpected token `(' error when pressing tab. Anyone encountered this?
2021-07-15 16:07:24 <donoban> uhM, I can't reproduce it
2021-07-15 16:09:31 <invoked> maybe it's a bug in one of the -bash-completion packages. you could try removing the ones you have installed one by one until the problem clears up
2021-07-15 16:09:36 <donoban> graywolf: does it happens whatever you have written in the command line?
2021-07-15 16:12:36 <graywolf> donoban: seems to happen only with vim
2021-07-15 16:12:45 <graywolf> (like: vim ../<TAB>)
2021-07-15 16:13:00 <graywolf> Will try to remove it
2021-07-15 16:13:50 <donoban> it doesn't fail here
2021-07-15 16:14:06 <donoban> maybe do you have some strange filename?
2021-07-15 16:16:43 <graywolf> No that I know of. Ok, but closing the terminal (ssh session) and reconnecting fixed it. 1. I feel stupid for not trying that first 2. I do not like transient errors like this :/
2021-07-15 16:16:52 <graywolf> Sorry
2021-07-15 16:20:26 <donoban> hehe no worries
2021-07-15 17:20:12 <getreal> thinking about deniable encryption, how does that work with backups?
2021-07-15 17:20:21 <getreal> seems like that's a bit of a hassle.
2021-07-15 17:20:49 <caskd> what does deniable encryption have to do with backups?
2021-07-15 17:22:23 <getreal> think about it
2021-07-15 17:35:10 <caskd> i seriously have no clue
2021-07-15 17:39:30 <lopid> 
ACTION thinks about it

2021-07-15 20:06:08 <danieli> i also don't quite see what it has to do with backups specifically
2021-07-16 00:05:49 <omni> plausible deniable backups?
2021-07-16 00:06:26 <omni> when you fail to restore from backup you can just say you didn't have them in the first place and noone can prove you wrong
2021-07-16 01:27:08 <ksr> danieli: Doesn't have a /dev/mem in my docker instance. Yeah, I assume it's similar to kmem in other distros.
2021-07-16 01:30:09 <ksr> Thanks everyone for looking at mem group.
2021-07-16 01:32:44 <ksr> My use case is a very odd one where, I have to use GID:8 in a Alpine Linux docker instance so that it plays nicely with a mapped FS. I was just evaluating whether there are any security risks of using GID:8 (mem) in Alpine Linux. Looks like there's none. Because I couldn't find anything owned by that group or any other mentions of "mem" group apart form /etc/group files.
2021-07-16 01:38:56 <getreal> lmfao, omni
2021-07-16 01:39:19 <getreal> danieli: if i have encrypted backups laying around, they're encrypted. they're encrypted data that can be rubberhosed
2021-07-16 01:39:46 <getreal> deniable encryption is about avoiding coercion, so all deniably encrypted data must be also deniably backed up.
2021-07-16 01:45:27 <Hello71> of course you don't have /dev/mem in docker
2021-07-16 01:45:42 <Hello71> kmem is totally different and also doesn't exist anymore
2021-07-16 01:45:56 <Hello71> if you're using recent kernel then use idmap
2021-07-16 07:58:06 <getreal> Is there a way to use efibootmgr/EFISTUB straight out of the gate, without setting up grub at all, but still using the defaults from setup-alpine for other things?
2021-07-16 08:39:15 <Wyk72> hello everyone
2021-07-16 08:39:20 <Wyk72> I have a question 
2021-07-16 08:39:39 <Wyk72> I'm in trouble with ZFS on Alpine Linux
2021-07-16 08:40:03 <Wyk72> when I install the system it keeps installing the wrong kernel and I don't know why
2021-07-16 08:40:25 <lopid> what's "the wrong kernel"?
2021-07-16 08:40:26 <Wyk72> the release for 3.14 is  5.10.43-0-lts
2021-07-16 08:40:54 <Wyk72> but after the setup script I get kernel 5.10.50-0-lts
2021-07-16 08:41:02 <Wyk72> and it screws up my ZFS setup badly
2021-07-16 08:41:15 <Wyk72> because modules are missing
2021-07-16 08:41:35 <Wyk72> and also in the mkinitfs features there's something wrong
2021-07-16 08:42:26 <Wyk72> the file /etc/mkinitfs/features.d/zfs.modules ha a reference to a non-existant module -> extra/zpios
2021-07-16 08:42:38 <lopid> sounds like the zfs package has jumped ahead?
2021-07-16 08:42:49 <Wyk72> that is not present anymore in the extra category of modules
2021-07-16 08:43:03 <Wyk72> it's kinda messy
2021-07-16 08:44:20 <lopid> zfs 2.0.3-r1?
2021-07-16 09:34:00 <danieli> ksr: ah, you didn't mention that it's a container and not a real machine, it won't have that device in that case
2021-07-16 09:35:34 <danieli> Hello71: hahaha, i didn't see your reply (or that the person left) when i wrote that :)
2021-07-16 13:41:11 <getreal> Is setup-alpine just a big ol' bash script?
2021-07-16 13:42:36 <ikke> yes
2021-07-16 13:42:37 <ikke> well
2021-07-16 13:42:43 <ikke> posix sh, but yes
2021-07-16 13:42:59 <ikke> (with a bith of allowance from ash :P)
2021-07-16 14:00:08 <getreal> ikke: And I can't configure efistub with this, I presume?
2021-07-16 14:00:26 <getreal> Also, can I change to data/sys/lvm modes AFTER setting up without much effort?
2021-07-16 14:01:36 <ikke> afaik, setup-disk does not support efistub
2021-07-16 14:01:46 <getreal> aww
2021-07-16 14:01:58 <getreal> maybe i could just read how setup-alpine works and hope it's self-documented
2021-07-16 14:02:04 <getreal> and do my own line by line setup
2021-07-16 14:02:18 <ikke> Look at sstup-disk
2021-07-16 14:02:39 <ikke> setup-alpine calls several other scripts
2021-07-16 14:03:18 <getreal> oh, that's fun
2021-07-16 14:05:55 <getreal> when did pipewire get popular? i've heard a lot of people switching to it on other distros
2021-07-16 14:06:08 <nickersonm> getreal: you can do `setup-alpine`, choose 'none' for disk selection, and then do your own disk setup.  Or set up to disk as usual and do your own EFISTUB installation.  For a BTRFS filesystem, I personally run `setup-alpine` with 'none' disk, set up the target with BTRFS subvolumes etc., then do `setup-disk -m sys $MNT`.
2021-07-16 14:07:09 <getreal> is there a way to cheat and still get a data install with arcane setups like that?
2021-07-16 14:07:43 <nickersonm> getreal: `setup-disk -m data $MNT`
2021-07-16 14:07:50 <getreal> hell ye.
2021-07-16 14:07:58 <nickersonm> https://wiki.alpinelinux.org/wiki/Alpine_setup_scripts#setup-disk
2021-07-16 14:08:22 <ikke> But with a data installation, you still boot from the bootmedia, right?
2021-07-16 14:08:48 <getreal> ah, really? i had no clue
2021-07-16 14:09:30 <getreal> oh, you can do setup-bootable and copy it to a partition.
2021-07-16 14:09:52 <getreal> "In data disk mode, the boot device may also remain the initial (and possibly read-only) installation media, or be copied to a partition (e.g. /dev/sdXY) with setup-bootable. " per wiki/Installation
2021-07-16 15:11:54 <getreal> ah fuck i just realized i have to set stuff up manually anyway to do uefi because i've got the board on uefi only
2021-07-16 15:12:02 <getreal> alright, i'm gonna do this over the weekend
2021-07-16 15:50:45 <nickersonm> getreal: this is how I do an EFI install https://rentry.co/ieuxd
2021-07-16 15:52:25 <greatwielder[m]> is there a solution to the firefox problem of: Unable to load hand2 from the cursor theme    spam in terminal when hovering over anything which would trigger a cursor icon change
2021-07-16 15:52:49 <greatwielder[m]> tried to find cursor themes and even installed some to no effect
2021-07-16 17:55:16 <lemontree> test
2021-07-16 18:04:52 <yyp> test passed
2021-07-16 18:11:47 <lemontree> :D thanks
2021-07-16 19:36:32 <bl4ckb0ne> is there a package that creates the plugdev group or do I have to make it myself
2021-07-16 19:38:13 <ikke> ther are packages that create it, but they are not generic packages
2021-07-16 19:42:05 <bl4ckb0ne> seems that plugdev was not the group I was looking for
2021-07-16 20:11:18 <unrznbl[m]> wondering if there is an intentional reason for not having a gdb-arm-none-eabi package? there are a few others already: gcc, newlib... https://pkgs.alpinelinux.org/packages?name=*arm-none-eabi&branch=edge&arch=x86_64 
2021-07-16 20:36:29 <bl4ckb0ne> out of curiosity, anybody got drm content to play on a browser? 
2021-07-16 20:37:49 <jaafarrc> I'm trying to make my Wacom driver work. I have xf86-wacom-input and the wacom_w8001.ko, and, the computer still doesn't seem to recognize my Wacom. (It's a Wacom Intuos Art.)
2021-07-16 20:38:23 <jaafarrc> s/my Wacom driver/my Wacom tablet/
2021-07-16 20:38:23 <alpine-meetbot> jaafarrc meant to say: I'm trying to make my Wacom tablet work. I have xf86-wacom-input and the wacom_w8001.ko, and, the computer still doesn't seem to recognize my Wacom. (It's a Wacom Intuos Art.)
2021-07-16 20:39:17 <jaafarrc> (I forgot to mention: wacom_w8001.ko is a kernel module I have.)
2021-07-16 20:42:06 <ikke> Do you see anything in dmesg?\
2021-07-16 20:43:42 <jaafarrc> ikke: dmesg says "new usb device, Product: Intuos PTS, Manufacturer: Wacom Co.,Ltd".
2021-07-16 20:45:14 <ikke> so it is recognized
2021-07-16 20:46:19 <jaafarrc> Well, if I run `libwacom-list-local-devices` from the `libwacom` package, I get: `Failed to find any devices known to libwacom.`
2021-07-16 21:22:09 <jaafarrc> ikke: So, I can't tell.
2021-07-16 21:29:09 <lemontree> jaafarrc: do you run libwacom-list-local-devices as root?
2021-07-16 21:29:21 <jaafarrc> Oh, I didn't.
2021-07-16 21:29:35 <jaafarrc> Still...
2021-07-16 21:31:45 <lemontree> https://github.com/linuxwacom/libwacom#adding-tablet-descriptions-to-libwacom
2021-07-16 21:31:49 <lemontree> this might be helpful
2021-07-16 21:32:25 <lemontree> jaafarrc: ^
2021-07-16 21:34:40 <lemontree> -> use libwacom-list-devices
2021-07-16 21:36:35 <jaafarrc> Hmmm, this is interesting.
2021-07-16 21:37:55 <jaafarrc> lemontree: My computer recognized my Wacom as "Intuos PTS" with PID (Product ID) 033c. If I use `libwacom-list-devices", I can see "Wacom Intuos Pen & Touch Snall" with PID 033c.
2021-07-16 21:38:11 <jaafarrc> s/Snall/Small/
2021-07-16 21:38:11 <alpine-meetbot> jaafarrc meant to say: lemontree: My computer recognized my Wacom as "Intuos PTS" with PID (Product ID) 033c. If I use `libwacom-list-devices", I can see "Wacom Intuos Pen & Touch Small" with PID 033c.
2021-07-16 21:39:31 <lemontree> so basically you need to follow the steps in the readme it seems
2021-07-16 21:40:06 <lemontree> (i think this readme describes kind of your situation)
2021-07-16 21:45:43 <jaafarrc> lemontree: It looks like there's a .tablet file for my tablet already, but, if I run `libwacom-update-db`, I get: Unable to find udev base directory: /etc/udev/hwdb.d
2021-07-16 21:45:49 <jaafarrc> (I use eudev)
2021-07-16 21:47:19 <lemontree> did you run that as root? eudev things can be skipped when using root, so if that works you know you need to add a udev rule
2021-07-16 21:47:51 <jaafarrc> lemontree: I did run it with root already.
2021-07-16 21:52:29 <lemontree> sorry, i don't know much but i'd try to look what the script does and fix things it expects
2021-07-16 21:52:37 <lemontree> for instance, here is your error thrown: https://github.com/linuxwacom/libwacom/blob/master/tools/libwacom-update-db.py#L229
2021-07-16 21:52:50 <lemontree> (i think)
2021-07-16 21:54:02 <jaafarrc> lemontree: Alright, I guess the Wacom can wait.
2021-07-16 21:55:01 <lemontree> hm, but i think at least this error is simple to fix: just create /etc/udev/hwdb.d
2021-07-16 21:56:50 <jaafarrc> I assume that's a directory.
2021-07-16 21:57:58 <lemontree> yes probably
2021-07-16 21:58:08 <jaafarrc> Btw, I don't seem to have /etc/libwacom.
2021-07-16 22:00:57 <jaafarrc> lemontree: I also ran `doas libwacom-update-db /etc/libwacom` after creating the directory, and I get "no such file or directory: 'systemd-hwdb'."
2021-07-16 22:02:20 <lemontree> yes this is also explained in the README it seems: https://github.com/linuxwacom/libwacom#libwacom-110-and-newer
2021-07-16 22:02:27 <lemontree> Copy the .tablet file into /etc/libwacom and run the libwacom-update-db tool:
2021-07-16 22:04:44 <jaafarrc> Is /etc/libwacom a file or a directory?
2021-07-16 22:05:01 <lemontree> ah, regarding systemd-hwdb: that is this: https://man.archlinux.org/man/systemd-hwdb.8.en - maybe you can try to understand what this systemd thing does and look what the replacement on alpine linux is
2021-07-16 22:05:42 <lemontree> then you can create a custom libwacom-update-db.py file which works correclty for alpine and send a merge request for that to alpine packages
2021-07-16 22:06:24 <lemontree> jaafarrc: ah even better you can try --skip-systemd-hwdb-update 
2021-07-16 22:06:59 <lemontree> look what nice things you can find when looking at the source code https://github.com/linuxwacom/libwacom/blob/master/tools/libwacom-update-db.py#L196
2021-07-16 22:07:38 <jaafarrc> Oh, that's cool.
2021-07-16 22:07:58 <jaafarrc> It told me to unplug and replug my device; I did that, but... nothing.
2021-07-16 22:08:20 <jaafarrc> Also, I still don't know if /etc/libwacom is a file or a directory.
2021-07-16 22:08:20 <lemontree> nad the answer to if /etc/libwacom is a file or dir is here: https://github.com/linuxwacom/libwacom/blob/master/tools/libwacom-update-db.py#L185
2021-07-16 22:09:35 <lemontree> and now i need to go to bed - but i have shown you the method to fix your problem. basically try to understand how things work
2021-07-16 22:09:39 <lemontree> good luck
2021-07-16 22:10:00 <jaafarrc> lemontree: Still nothing...
2021-07-16 22:10:16 <jaafarrc> And I don't know how to read Python.
2021-07-16 22:12:54 <lemontree> i too don't know how to read python but you can learn it
2021-07-16 22:13:07 <lemontree> you can read english that is enough
2021-07-16 22:18:30 <jaafarrc> Oh, my God... why.
2021-07-16 22:23:03 <jaafarrc> Alright... next question: Can I downgrade from edge to stable?
2021-07-16 22:59:21 <getreal> Thanks a ton, nickersonm; I'll look into that while I'm back and forth for work
2021-07-16 22:59:47 <getreal> Ah wait, you still using dat grub
2021-07-16 22:59:51 <getreal> I'll modify from there.
2021-07-16 23:24:23 <Hello71> jaafarrc: unless you specifically need something only in xf86-input-wacom you're probably better off with xf86-input-libinput
2021-07-16 23:24:48 <Hello71> you can downgrade at your own risk
2021-07-16 23:24:52 <Hello71> apk upgrade -a
2021-07-16 23:25:31 <jaafarrc> Do I need the libinput package for xf86-input-libinput?
2021-07-16 23:36:29 <Hello71> https://pkgs.alpinelinux.org/package/edge/community/x86_64/xf86-input-libinput Depends: libinput-libs
2021-07-16 23:36:47 <Hello71> https://pkgs.alpinelinux.org/contents?branch=edge&name=libinput&arch=x86_64&repo=community Contents: /usr/bin/...
2021-07-16 23:45:14 <jaafarrc> Oh, my God, I didn't pay attention to that. That's embarrasing, I'm sorry.
2021-07-17 00:04:06 <kevint> bl4ckb0ne: only on flatpak Firefox :(
2021-07-17 00:07:58 <getreal> Ah wait, I'll have to modify quite significantly, as setup-disk doesn't actually allow linux to be its own bootloader
2021-07-17 00:34:14 <bl4ckb0ne> kevint: yeah i saw that the widevine lib segfaults on musl
2021-07-17 00:34:26 <bl4ckb0ne> well at leasti have a reason to move out of spotify
2021-07-17 00:48:15 <getreal> Actually does the kernel built for alpine come with CONFIG_EFI_STUB=y?
2021-07-17 00:51:10 <getreal> I swear it did previously, or my last time doing this wouldn't have worked
2021-07-17 01:52:04 <onelegend> I love Alpine Linux
2021-07-17 04:02:37 <omni> me too!
2021-07-17 08:25:20 <donoban> who doesn't? :D
2021-07-17 16:01:56 <onelegend> I know a lot of people run alpine linux in their docker containers, but I use alpine linux to run docker!
2021-07-17 16:05:20 <lopid> reasonable
2021-07-17 17:22:36 <Lorenzoas3ii[m]> 
ACTION  < https://matrix.org/_matrix/media/r0/download/matrix.org/UypUgcOHYiHSrLNJdWeHAcnG/message.txt >

2021-07-17 17:23:40 <ikke> please, for the irc users here, do not use the matrix multiline feature for the entire message
2021-07-17 17:23:46 <ikke> we just get an url without context
2021-07-17 17:25:03 <ikke> Lorenzoas3ii[m]: What is the command you are executing?
2021-07-17 18:15:54 <danieli> yes, what is the command and have you preloaded any of the official signing keys?
2021-07-17 18:54:59 <Lorenzoas3ii[m]> <ikke "Lorenzo (as3ii): What is the com"> 
2021-07-17 18:55:15 <Lorenzoas3ii[m]> <ikke "please, for the irc users here, "> oh sorry
2021-07-17 18:56:19 <Lorenzoas3ii[m]> 
ACTION  < https://matrix.org/_matrix/media/r0/download/matrix.org/vFBQpJZFSNinGJNecqMXagcc/message.txt >

2021-07-17 18:58:16 <ikke> Where do you have the keys for this repositoy?
2021-07-17 18:58:29 <ikke> You can try adding --hostkeys, which should copy the keys from your host
2021-07-17 18:59:55 <Lorenzoas3ii[m]> sorry, I'm noob with alpine, I'm following https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage
2021-07-17 19:05:26 <ikke> Did you try adding --hostkeys?
2021-07-17 19:05:34 <Lorenzoas3ii[m]> <ikke "You can try adding --hostkeys, w"> nothing changed
2021-07-17 19:06:32 <jaafarrc> Hey, I'm here for my Wacom tablet again. So, it's does get recognized, it does have a udev entry, it just that nothing happens when I try to use the tablet.
2021-07-17 19:12:13 <Lorenzoas3ii[m]> 
ACTION  < https://matrix.org/_matrix/media/r0/download/matrix.org/zszPMuGosYnrhSoSRwPjvGtB/message.txt >

2021-07-17 19:12:56 <ikke> Lorenzoas3ii[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/zszPMuGosYnrhSoSRwPjvGtB/message.txt >
2021-07-17 19:27:48 <Lorenzoas3ii[m]> 
ACTION  < https://matrix.org/_matrix/media/r0/download/matrix.org/lcqacgUiXdcJpnojakKRhVbt/message.txt >

2021-07-17 19:29:29 <ikke> https://tpaste.us/DXDX 
2021-07-17 19:31:01 <Lorenzoas3ii[m]> damn matrix bridge... sorry
2021-07-17 19:41:21 <Lorenzoas3ii[m]> sh mkimage.sh --tag "v3.14" --outdir ~/iso-aarch64 --arch "aarch64" --repository "repo-main" --repository "repo-community" --profile "rpi"
2021-07-17 19:41:24 <Lorenzoas3ii[m]> maybe if I don't insert links the message will be sent as normal text
2021-07-17 19:41:59 <fluix> it's because you're trying to send like 5 lines at a time
2021-07-17 19:42:36 <fluix> although who knows, the bridge mangles that as well it seems (like the link is cut off for example)
2021-07-17 19:43:49 <Lorenzoas3ii[m]> (now I have to go, here it is evening, tomorrow I'll read the possible answers)
2021-07-17 21:24:02 <tdtrask> ijohnr is gone, but I'm guessing acf-openssh didn't show up after installing it because his session was cached in /tmp/. Remove the session or wait until it expires, and the new tab will appear.
2021-07-17 21:24:32 <ikke> tdtrask: good to know
2021-07-17 22:28:15 <Hello71> fluix: most likely a copy paste issue
2021-07-18 06:23:04 <Skaface> is bluetooth a2dp audio working?
2021-07-18 06:26:29 <atka> Skaface: I believe it does
2021-07-18 06:31:26 <atka> Skaface: you'll need pulseaudio-bluez, maybe add yourself to the bluetooth group, I don't remember exactly the steps on alpine
2021-07-18 06:32:09 <Skaface> yep im working through it, i needed pulseaudio-modules-bt from testing
2021-07-18 06:32:13 <Skaface> thankyou
2021-07-18 06:33:12 <atka> np, good luck!
2021-07-18 06:42:10 <atka> Skaface: I don't know the status of pipewire on alpine, but that will be how bluetooth audio will be supported in the future, the pulseaudio-modules-bt is now deprecated, should still work fine though for now.
2021-07-18 07:37:31 <fucksmokers> Are you the maintainer for darktable on alpine, ikke? Can we get 3.6 in edge?
2021-07-18 07:37:46 <ikke> I can
2021-07-18 07:38:18 <fucksmokers> Thanks, comrade. I'm excited for some new features after I figure out the whole "install efi without grub upon first setup" situation
2021-07-18 07:38:49 <fucksmokers> Though I also have considered trying to figure out full disk encryption. How do LVM and btrfs interface?
2021-07-18 07:39:17 <ikke> as far as I understand, with btrfs you do not need something like lvm
2021-07-18 07:40:34 <atka> fucksmokers: btrfs is instead of lvm + fs
2021-07-18 07:41:16 <atka> fucksmokers: I've been diving into btrfs + alpine recently if you have any questions, also with luks
2021-07-18 08:09:00 <francescovolpini[m]> hi all!
2021-07-18 08:20:58 <fucksmokers> atka: interesting. how does the actual workflow for signin in work for you? how is it decrypted?
2021-07-18 08:24:16 <fucksmokers> And in the modern day, do people still recommend putting /home on a different part? Maybe I could just encrypt /home and have things be alright.
2021-07-18 08:32:42 <atka> fucksmokers: you can decrypt in grub if you want full disk encryption (minus efi partition if using uefi)
2021-07-18 08:33:33 <atka> fucksmokers: you can also just have encrypted root and unencrypted boot and boot/efi
2021-07-18 08:33:36 <fucksmokers> Damn. Why couldn't I use the initramfs to do that and then use the kernel as the efistub?
2021-07-18 08:34:18 <atka> I'm not sure about that, I use grub personally
2021-07-18 08:34:47 <fucksmokers> Any reason why?
2021-07-18 08:34:51 <atka> but if using btrfs you'll not have a separate home partition but different subvolumes
2021-07-18 08:35:07 <fucksmokers> I suppose you just prefer exploits like hitting backspace 28 times? ;P
2021-07-18 08:35:19 <fucksmokers> iiiinteresting
2021-07-18 08:35:32 <fucksmokers> so i'd just do efi and then "the rest," and maybe a swap partition? is a partition better than a file on an ssd?
2021-07-18 08:36:01 <atka> eh, I just use a swapfile, but you can do a swap subvolume on btrfs
2021-07-18 08:36:46 <fucksmokers> i'll have to do more review on how all of that then interfaces with the rest of setup
2021-07-18 08:36:58 <fucksmokers> feels weird interfacing with scripts like that sometimes
2021-07-18 08:37:34 <atka> also that grub exploit won't do anything if you are using luks
2021-07-18 08:37:48 <atka> and was patched anyway
2021-07-18 08:40:10 <atka> basically I do luks on my drive, then setup a btrfs filesystem on the luks container, then make all my subvolumes for / /boot /home etc
2021-07-18 08:41:02 <atka> umount the toplevel btrfs and then mount the / subvolume and do a sys install to mount, chroot in and finish up
2021-07-18 08:41:35 <fucksmokers> i was just bringing up how i can't see the software as that good or necessary when stuff like that sits around; it's part of a general ethos that if i don't need it, i don't install it
2021-07-18 08:41:50 <fucksmokers> huh; very interesting
2021-07-18 08:42:27 <atka> I understand, I'm usually very minimal in my setups but grub just works and I've been using FDE for quite awhile
2021-07-18 08:44:41 <fucksmokers> very interesting; i'll keep that in mind
2021-07-18 08:45:39 <atka> and evil maid attacks or someone with physical access that is knowlegeable about linux vulerabilities are very improbable
2021-07-18 08:45:58 <atka> for me anyway
2021-07-18 09:00:36 <atka> fucksmokers: I'm barely awake and headed to bed, have a good one.
2021-07-18 09:01:03 <fucksmokers> Same! I'm gonna try finishing this up tomorrow.
2021-07-18 09:02:19 <atka> fucksmokers: alright, I'll be around, good luck!
2021-07-18 15:54:12 <jaafarrc> Hey, I accidentally disconnected yesterday while asking a question.
2021-07-18 15:55:00 <jaafarrc> Basically: my Wacom tablet gets recognized (according to dmesg and /proc/bus/input/devices) but trying to move the mouse with the tablet doesn't do anything.
2021-07-18 15:55:27 <ikke> Are you member of the input group?
2021-07-18 16:08:00 <jaafarrc> yes.
2021-07-18 16:08:25 <jaafarrc> ikke: `$ groups` `input video wheel`
2021-07-18 16:09:02 <ikke> Not sure if required, but do you have xf86-input-wacom installed?
2021-07-18 16:09:50 <jaafarrc> I was told I was better off `xf86-input-libinput`, and I *did* have xf86-input-wacom, but it didn't change anything.
2021-07-18 16:09:57 <ikke> ok
2021-07-18 16:10:27 <jaafarrc> If I `find` my kernel's modules directory, I can see a Wacom module.
2021-07-18 16:11:15 <ikke> jaafarrc: does modprobe show the module being loaded?
2021-07-18 16:11:24 <jaafarrc> How do I check?
2021-07-18 16:11:28 <ikke> modprobe
2021-07-18 16:11:54 <jaafarrc> modprobe what?
2021-07-18 16:11:59 <ikke> sorry, lsmod
2021-07-18 16:12:03 <ikke> my bad
2021-07-18 16:13:52 <jaafarrc> ikke: Nothing comes up when I `doas lsmod | grep wacom`.
2021-07-18 16:27:51 <jaafarrc> By the way, I only have wacom_w8001, not just wacom.
2021-07-18 16:27:59 <jaafarrc> So, I think I need to compile this?
2021-07-18 16:38:12 <jaafarrc> Hmmm, input-wacom wiki says wacom_w8001.ko is for older stuff and wacom.ko is for newer stuff. I have no idea when my model was released.
2021-07-18 16:39:26 <ikke> did you try modprobe wacom?
2021-07-18 16:40:07 <jaafarrc> Yes, I modprobe'd wacom_w8001 (wacom doesn't exist, only wacom_w8001) does.
2021-07-18 16:40:19 <ikke> ok
2021-07-18 16:40:28 <ikke> Not sure then
2021-07-18 16:40:37 <jaafarrc> Hmmm, maybe I need xf86-input-libinput? I use Wayland, so, I dunno.
2021-07-18 16:49:13 <fucksmokers> If I don't use grub, how does the kernel get updated?
2021-07-18 16:49:39 <Hello71> if you use wayland then xf86-* is totally irrelevant
2021-07-18 16:49:42 <Hello71> note the xf86
2021-07-18 16:49:56 <Hello71> fucksmokers: grub doesn't "update the kernel"
2021-07-18 16:50:02 <fucksmokers> correct
2021-07-18 16:50:06 <fucksmokers> but if you update the kernel
2021-07-18 16:50:10 <jaafarrc> Hello71: I know.
2021-07-18 16:50:19 <fucksmokers> then grub would change its settings, right
2021-07-18 16:50:27 <Hello71> check libinput debug-events
2021-07-18 16:50:38 <jaafarrc> Where do I check those?
2021-07-18 16:50:46 <Hello71> fucksmokers: usually, unless you don't have grub
2021-07-18 16:50:49 <Hello71> or are using arch
2021-07-18 16:50:51 <Hello71> or a custom config
2021-07-18 16:50:51 <fucksmokers> and additionally, my curiosity becomes "if I put the kernel elsewhere (/boot/efi), how the hell does it get updated later"
2021-07-18 16:50:58 <Hello71> jaafarrc: run the command "libinput debug-events"
2021-07-18 16:51:04 <fucksmokers> since i plan to just efistub the kernel
2021-07-18 16:51:08 <jaafarrc> Oh.
2021-07-18 16:52:08 <jaafarrc> Huh, the command doesn't exist.
2021-07-18 16:52:59 <Hello71> install it
2021-07-18 16:53:08 <jaafarrc> Yeah, I did.
2021-07-18 16:53:09 <Hello71> it's for debugging
2021-07-18 16:53:19 <Hello71> not required for normal use
2021-07-18 16:54:29 <jaafarrc> Alright, I ran that command, I plugged in my tablet, and my tablrt pops up as event19.
2021-07-18 16:56:22 <Hello71> yes but do you get events
2021-07-18 16:56:32 <jaafarrc> What do you mean?
2021-07-18 17:08:51 <holgersson> fucksmokers: EFI -> EFI entry you can write with efibootmgr -> efistub or bootloader
2021-07-18 17:08:51 <holgersson> fucksmokers: Maybe this article helps you a bit with configs and understanding: https://wiki.archlinux.org/title/EFISTUB#Using_UEFI_directly
2021-07-18 17:10:01 <holgersson> fucksmokers: You put you kernel *somewhere* and then let your software point to that. There are some more or less canonical paths, but you can at least configure the EFI and the most bootloaders I know off to nearly any path in /boot or the ESP.
2021-07-18 17:13:28 <fucksmokers> As I'm thinking about this, if the kernel updates, there's likely a script that gets called as well--when the package updates, my worry is that I'm gonna be moving my kernel and initramfs into places the package doesn't install it to, so something might break or be held back
2021-07-18 17:13:52 <fucksmokers> I did it once before, but don't know how package management works enough to be certain about that.
2021-07-18 17:14:53 <ikke> There is a trigger in apk
2021-07-18 17:16:12 <ikke> btw, I just merged the darktable upgrade, it should be available soon in edge
2021-07-18 17:23:43 <fucksmokers> ah hell yeah
2021-07-18 17:23:44 <fucksmokers> thanks ikke
2021-07-18 17:23:46 <fucksmokers> Do you shoot?
2021-07-18 17:23:59 <ikke> zNo
2021-07-18 17:24:07 <fucksmokers> fucked up
2021-07-18 17:24:12 <fucksmokers> maybe one day
2021-07-18 17:24:28 <jaafarrc> Hello71: Sorry; got disconnected. And no, I do not get events.
2021-07-18 17:24:41 <Hello71> only device_added?
2021-07-18 17:24:54 <jaafarrc> Yeah.
2021-07-18 17:28:02 <Jaafarhehim[m]> I wonder if I can send a message from Matrix.
2021-07-18 17:28:08 <ikke> You can
2021-07-18 17:28:13 <Jaafarhehim[m]> Yes!
2021-07-18 17:34:30 <Hello71> jaafarrc: then the problem is with your kernel driver, not userspace
2021-07-18 17:34:41 <Hello71> check dmesg
2021-07-18 17:35:42 <jaafarrc> It detects Wacom.
2021-07-18 17:36:07 <jaafarrc> As in, my tablet.
2021-07-18 17:43:36 <Jaafarhehim[m]> Matrix me is still here.
2021-07-18 17:45:29 <lopid> i don't see it
2021-07-18 17:45:56 <Jaafarhehim[m]> What do you not see?
2021-07-18 17:46:11 <lopid> matrix you
2021-07-18 17:46:41 <Jaafarhehim[m]> Oh.
2021-07-18 17:46:45 <Jaafarhehim[m]> I'm over here.
2021-07-18 17:46:51 <lopid> ah
2021-07-18 17:47:01 <Jaafarhehim[m]> :)
2021-07-18 18:52:40 <JulianGro> Hello, was the lfm package removed or was it never in the repos? I thought I had it installed before but it doesn't seem to be on the repositories. lfm stands for "Last File Manager".
2021-07-18 18:56:03 <ikke> Seems to have never been in the repository
2021-07-18 18:57:57 <ikke> There is lf
2021-07-18 19:01:36 <JulianGro> Sad
2021-07-18 19:37:11 <donoban> aynone using fetchmail here? is there better alternatives?
2021-07-18 19:39:06 <donoban> I would like to run fetchmail or similar with their own uid but then have a maildir on my home folder, if possible also store my credentials in my home dir
2021-07-18 19:41:36 <ikke> donoban: what do you use it for
2021-07-18 19:41:50 <Habbie> 20 years ago i switched from fetchmail to getmail
2021-07-18 19:41:58 <Habbie> i'm not saying it's better; i'm just saying it also exists and worked well for me
2021-07-18 19:42:55 <donoban> ikke: nothing special, sync my riseup account with IMAPS
2021-07-18 19:43:14 <donoban> Habbie: I read about it but it seems not packaged in alpine
2021-07-18 19:43:48 <Habbie> donoban, ah, it's in pip
2021-07-18 19:47:18 <donoban> ok
2021-07-18 19:50:32 <kevint> I used isync for a while, it's packaged for alpine too
2021-07-18 19:51:22 <donoban> thanks I will take a look tomorrow
2021-07-18 20:05:22 <lopid> imapsync is in the repo, but i don't know what it is
2021-07-18 20:27:18 <Skaface> i used to, I just use mutts builtin imap client now
2021-07-18 20:29:08 <Skaface> in my opinion, unless you are running a mail server, you might as well just use the mail server you already have with your email address?
2021-07-18 21:05:00 <Jaafarhehim[m]> Hello71: I've discovered something interesting: everytime I reboot, wacom_8001.ko is not loaded. I have to enable it everytime I boot. (my tablet still doesn't work, though.)
2021-07-18 21:15:52 <atka> Jaafarhehim[m]: echo "wacom_8001" >> /etc/modules I believe
2021-07-18 21:16:09 <Jaafarhehim[m]> I'll try that.
2021-07-19 01:09:05 <kelvinfahrenheit43> i'm trying to get cwm to work on alpine, but i can't get xinit to work without using doas to log in as root
2021-07-19 01:11:12 <Hello71> add user to input and video group
2021-07-19 01:11:15 <Hello71> and use startx instead of xinit
2021-07-19 01:11:22 <Hello71> or better, use sx
2021-07-19 01:27:07 <kelvinfahrenheit43> startx doesn't work either
2021-07-19 01:27:21 <kelvinfahrenheit43> but i'll try adding the user to the input and video groups
2021-07-19 01:29:20 <kelvinfahrenheit43> Hello71, i should just edit /etc/group, right?
2021-07-19 01:29:41 <Hello71> addgroup is better
2021-07-19 01:29:51 <Hello71> i did say "and"
2021-07-19 01:41:27 <kelvinfahrenheit43> i can use startx as a user now
2021-07-19 01:41:48 <kelvinfahrenheit43> i can't use sx though, i set it to use my xinitrc 
2021-07-19 01:42:15 <kelvinfahrenheit43> which is just exec cwm & xsetroot -solid steelblue
2021-07-19 01:42:37 <kelvinfahrenheit43> xsetroot is installed so idk why it would throw an error
2021-07-19 01:43:10 <Hello71> i don't think you can exec &
2021-07-19 01:43:14 <Hello71> or at least it doesn't make sense
2021-07-19 01:43:42 <kelvinfahrenheit43> actually it's exec cwm on one line, and then xsetroot -solid steelblue
2021-07-19 02:24:24 <jimmy> hello, I'm a bit confused with alpine's three modes. I want to set up a server for a few things like email, web, and media. I'll have a separate hard drive for media and things outside of the OS. the speed of diskless mode sounds appealing, but are there any downsides in my scenario?
2021-07-19 02:53:00 <jimmy> also, i think there's a discrepancy in the documentation here https://wiki.alpinelinux.org/wiki/Installation#Data_Disk_Mode and here https://alpine-linux.readthedocs.io/en/latest/getting_started.html
2021-07-19 02:53:32 <jimmy> data and sys modes are swapped?
2021-07-19 03:02:28 <Skaface> hmm yeah they seem swapped on one site
2021-07-19 03:02:35 <Skaface> i think readthedocs.io is incorrect
2021-07-19 07:35:14 <LACampbell> I've always just done a traditional install with alpine. I can't be bothered syncing all the time
2021-07-19 08:53:50 <Guest1049> Hey, I've read about the news of plans on replacing openrc in alpine, is that in execution or planning or anything but abandoned?
2021-07-19 08:53:50 <Guest1049> Also, I've wondered why alpine uses busybox init instead of the openrc-init that is in the system anyway. Could trim a little size off by removing the init part of busybox.
2021-07-19 08:54:36 <ikke> https://skarnet.com/projects/service-manager.html
2021-07-19 08:54:39 <ikke> See the update at the top
2021-07-19 08:55:02 <ikke> I don't know the reasons for bb init stead of openrc-init
2021-07-19 08:55:11 <ikke> instead*
2021-07-19 08:55:16 <Guest1049> nice, thanks
2021-07-19 13:53:11 <Hello71> i think most of the reason is that openrc-init is fairly new
2021-07-19 13:53:26 <Hello71> and afaik doesn't have any major improvements
2021-07-19 13:53:39 <Guest1049> iirc openrc-init came after openrc yeah
2021-07-19 13:53:51 <Guest1049> but it's still in the system, so I'd think one has to go
2021-07-19 13:55:17 <Hello71> why
2021-07-19 13:55:41 <Guest1049> alpine has a history of minimizing size
2021-07-19 13:55:53 <Guest1049> going as far as to even use the `-Os` flag for packages
2021-07-19 13:55:53 <Hello71> but for 22 KB?
2021-07-19 13:56:00 <Hello71> -Os saves a lot more space than that
2021-07-19 13:56:19 <Hello71> you can make a MR to move openrc-init to a subpackage if you want
2021-07-19 13:57:35 <Guest1049> I don't know how much space busybox init takes
2021-07-19 14:00:05 <Hello71> https://git.busybox.net/busybox/tree/init/init.c says 10 kb
2021-07-19 14:01:01 <ikke> Moving it to a subpackage would most likely take up more space
2021-07-19 14:02:02 <Hello71> isn't the per-package overhead only /lib/apk/db/installed size?
2021-07-19 14:02:30 <Hello71> ~400 bytes/packages
2021-07-19 14:03:10 <ikke> I also take mirror storage into account
2021-07-19 14:15:43 <Guest1049> thanks for the info
2021-07-19 19:52:10 <Gues__________________________> Hey I'm working with packer and cloud-init to create an alpine linux template for my homelab. It's working great but I enable udhcpd/c during setup which helps me get packages which is perfect. Now I have cloud-init on the server I no longer need udhcpd to run. Not sure if I'm blind or not but I'm struggling to find a way to stop it running at boot.
2021-07-19 19:53:00 <Gues__________________________> Could someone give me a bit of assistance?
2021-07-19 19:53:35 <danieli> does `rc-update del udhcpc` work?
2021-07-19 19:54:52 <danieli> i checked an alpine box i have, i just see the 'networking' service on boot, that's the only seemingly relevant one
2021-07-19 19:55:15 <lopid> ridiculous nick
2021-07-19 19:55:47 <Gues__________________________> It's not a part of the runlevel so that doesn't seem to do it
2021-07-19 19:56:32 <lopid> does it get added?
2021-07-19 19:58:04 <Gues__________________________> No, you can't add it yourself even
2021-07-19 19:58:13 <lopid> danieli
2021-07-19 21:22:42 <JackThomasAnder> i'm back
2021-07-19 21:24:57 <JackThomasAnder> i have a question about mounting a network drive having spaces in the name
2021-07-19 22:03:18 <nitsky> Hi, I have a prebuilt binary for aarch64. I would like to build an APK for it, but I am on a x86_64 machine. This fails with the error: "Package not available for the target architecture (x86_64). Aborting.".
2021-07-19 22:05:40 <nitsky> Is it possible to tell abuild that I am cross-building and to allow the package to not match the architecture of the computer it is built on?
2021-07-19 22:24:57 <spacejam> MORNING BOYS
2021-07-19 22:25:10 <spacejam> How do custom initramfses work with alpine?
2021-07-19 22:26:53 <spacejam> is atka around?
2021-07-19 22:29:24 <Jaafarhehim[m]> This is weird...
2021-07-19 22:30:47 <Jaafarhehim[m]> 
ACTION  < https://matrix.org/_matrix/media/r0/download/matrix.org/rTjLzMkDaoOCUJMzbgBLNDWn/message.txt >

2021-07-19 22:31:29 <Jaafarhehim[m]> I don't seem to be on the list of users who are part of wheel according to `/etc/group`, but `groups` tells me I am.
2021-07-19 22:42:24 <atka> spacejam: hello
2021-07-19 22:44:59 <spacejam> atka: so you're using btrfs, and no lvm on the disk you've encrypted, right?
2021-07-19 22:45:20 <atka> correct
2021-07-19 22:45:39 <spacejam> one thing i don't get about the linux boot process is that i'm unsure how the decryption of the full disk works. how does grub decrypt the drive and not take ten years to decrypt it
2021-07-19 22:45:56 <spacejam> and if grub's doing the unlocking, how does it pass info to the kernel to just do the decryption from there?
2021-07-19 22:46:34 <atka> as I understand grub asks for your key and passes it to the kernel which then decrypts it
2021-07-19 22:47:04 <spacejam> so it passes it in as a kernel param to the kernel code, which is where i'm confused: if you're encrypting your whole disk, how's the kernel available to grub
2021-07-19 22:48:03 <spacejam> and my next question is also regarding the linux boot process again: does the initramfs just chroot to the real fs? how's that even work?
2021-07-19 22:49:44 <atka> I don't know how the whole process works on a lower level, you'd have to dig into the grub documentation for that
2021-07-19 22:49:49 <spacejam> ah, interesting.
2021-07-19 22:49:51 <atka> I used this as a resource https://wiki.archlinux.org/title/GRUB#Encrypted_/boot
2021-07-19 22:50:04 <spacejam> and /boot is just a btrfs volume for you?
2021-07-19 22:50:17 <spacejam> do you have the device-mapper stuff at all when btrfs does the voluming?
2021-07-19 22:51:28 <atka> yes, /boot is a btrfs subvolume, yes you need the device mapper stuff as btrfs is on top of luks and does not yet have its own encryption
2021-07-19 22:52:31 <spacejam> when we say on top of luks, do we mean that we still have to do all of the cryptsetup stuff, and you just do the dm-crypt plain mode on the whole volume?
2021-07-19 22:53:43 <atka> yes, you have to create a luks container on your disk, unlock and mount said container, then create the btrfs filesystem within that container
2021-07-19 22:54:23 <atka> and then subvolumes within the btrfs filesystem
2021-07-19 22:54:24 <spacejam> ah, but that's SEPARATE from lvm
2021-07-19 22:54:31 <spacejam> i think i'm starting to hate acronyms again
2021-07-19 22:54:32 <atka> I don't use lvm at all
2021-07-19 22:54:55 <spacejam> yeah, i think i was merging the "container" and "volume" concepts in my head
2021-07-19 22:55:01 <spacejam> and then being like "wtf do i need a container for"
2021-07-19 22:55:31 <atka> yeah, luks container with filesystem and subvolumes inside
2021-07-19 22:55:40 <spacejam> sec
2021-07-19 23:30:39 <anoduck> Having the worst time setting up alpine as a router. 
2021-07-19 23:31:03 <anoduck> Switching to awall, has helped. 
2021-07-19 23:31:44 <anoduck> Still not passing packets though. Not forwarding. 
2021-07-19 23:32:12 <anoduck> Anyone have experience with awall and multiwans. 
2021-07-20 00:05:35 <danieli> what's your routing like? have you enabled IP forwarding in the kernel?
2021-07-20 00:20:56 <Jaafarhehim[m]> I don't seem to be on the list of users who are part of wheel according to `/etc/group`, but `groups` tells me I am.
2021-07-20 00:21:08 <Jaafarhehim[m]> (i bumped my question, if that's okay)
2021-07-20 00:30:05 <anoduck> danieli: ip forwarding is enabled at the kernel level. 
2021-07-20 00:33:00 <anoduck> danieli: I have dual wans, both are provided by other routers. dsl modem and a mifi modem. then I have one Lan. 
2021-07-20 00:34:55 <Hello71> anoduck: did you enable nat
2021-07-20 00:35:17 <Hello71> Jaafarhehim[m]: did you log out and in again
2021-07-20 00:35:43 <anoduck> Jaafarhehim[m]: run 'cat /etc/passwd | grep $user' are you a member of wheel there? 
2021-07-20 00:36:07 <anoduck> Hello71: via modprobe, yes. 
2021-07-20 00:36:16 <Hello71> what?
2021-07-20 00:36:49 <Hello71> actually neither of those make sense
2021-07-20 00:37:20 <Hello71> although it is possible to have wheel be your primary group, it would be a number
2021-07-20 00:39:07 <anoduck> Hello71: there is a kernel module for nat. I don't believe there is a configuration variable for nat, or it might not be needed. 
2021-07-20 00:40:22 <anoduck> Hello71: configuration variable via sysctl, for clarification. 
2021-07-20 00:41:17 <Hello71> there are ten kernel modules for nat: https://pkgs.alpinelinux.org/contents?file=*nat.ko&path=&name=&branch=edge&arch=x86_64
2021-07-20 00:41:41 <Hello71> if you exclude *conntrack*.ko, xt_state.ko, etc
2021-07-20 00:42:07 <Hello71> it would be insane if loading the nat module magically natted all traffic
2021-07-20 00:42:16 <Hello71> you need to configure it in the firewall
2021-07-20 00:43:22 <Hello71> alpine linux provides standard kernel and userspace facilities like netfilter, iptables, and nftables. awall is an optional added package to make certain configurations easier
2021-07-20 00:43:29 <Hello71> otherwise it is exactly identical to other linux distros
2021-07-20 00:43:53 <anoduck> Hello71: I enables iptables_nat, per awall setup instructions. 
2021-07-20 00:44:30 <Hello71> if you're reading anything on the wiki it's probably obsolete
2021-07-20 00:45:25 <anoduck> Hello71: well, that is irritating... 
2021-07-20 00:45:29 <Hello71> although actually specifically for iptables it might still be up-to-date-ish
2021-07-20 00:45:32 <Hello71> notwithstanding nftables
2021-07-20 00:45:53 <anoduck> I don't have nftables installed 
2021-07-20 00:48:54 <anoduck> I will enable xt_nat.ko since awall uses xtables for something. 
2021-07-20 00:49:36 <anoduck> I allow pass all from Lan to anywhere. 
2021-07-20 00:50:06 <anoduck> Drop from inet to inet
2021-07-20 00:50:21 <anoduck> Inet being wan
2021-07-20 00:51:18 <anoduck> Allow _fw to inet
2021-07-20 00:51:35 <anoduck> Allow ping across all 
2021-07-20 00:57:46 <anoduck> Ip6 is disabled 
2021-07-20 01:01:44 <anoduck> It should be working, I replaced all cables, bought a new switch, and tested the cables. The only thing I can think of is hardware issues. But if it were, then ethernet should be not working, not just solely failing to forward. 
2021-07-20 01:02:51 <anoduck> My routes might be botched. But none were set manually 
2021-07-20 01:09:37 <anoduck> nf_nat, xt_nat, iptables_nat, ipset are all enabled
2021-07-20 01:16:00 <anoduck> Kernel mod wise. 
2021-07-20 01:17:11 <anoduck> Working on enabling adp-router policy. Needs variable adp_lan_interfaces defined. 
2021-07-20 01:27:30 <Hello71> modprobe rarely enables anything in network
2021-07-20 03:45:58 <anoduck> The only module that does not appear to be loaded is xt_nat.ko. I will use insmod, to ensure that it is. After that and enabling the adp_router rule. I will be stretching for ideas of what could be the issue. 
2021-07-20 03:46:44 <anoduck> A rule in awall policies refers to adp_config. Does anyone know where that file is located? 
2021-07-20 04:10:59 <unas> Hello i was wondering how to open another session on Alpine linux Ish?
2021-07-20 04:29:32 <anoduck> I am going to start swapping out ethernet cards and see if it makes a difference. I had the router working on Ubuntu, then they discontinued support for x86. Debian still supports it, but I have enjoyed working with alpine for several years now. It should be a cinch. Lightning storm has turned into a nightmare. 
2021-07-20 04:30:09 <anoduck> Network has been down for nearly two weeks. 
2021-07-20 04:30:34 <anoduck> Ah, FML... 
2021-07-20 05:15:20 <atka> anoduck: sorry for your trouble, I don't have any experience with your particular setup, I have built a few alpine routers but they were pretty basic, iptables, dnsmasq etc
2021-07-20 05:16:39 <atka> maybe try a simple setup and see if that works?
2021-07-20 07:46:47 <nbastin> anyone know anything about openvpn client configuration on alpine re: openrc and such?
2021-07-20 07:50:11 <clandmeter> nbastin: you can add an openvpn.conf and start the service iirc.
2021-07-20 07:50:46 <nbastin> yeah I mean I have that part, although I guess if I want openvpn.<name> to work, I have to symlink the rc script myself?  (not super familiar with openrc yet)
2021-07-20 07:51:10 <ikke> yes
2021-07-20 07:51:15 <ikke> you need to create the symlink
2021-07-20 07:51:49 <nbastin> I was getting some odd behaviour, which seems to be related to PEER_DNS being set in the init script - is there some way I'm supposed to cleanly change this, or do I just edit it?
2021-07-20 07:52:41 <clandmeter> peerdns is set to yes per default
2021-07-20 07:52:59 <nbastin> yeah I'm just wondering how I change that
2021-07-20 07:53:04 <clandmeter> conf.d
2021-07-20 07:53:07 <nbastin> (aside from just editing the file...)
2021-07-20 07:53:19 <nbastin> I can't use conf.d because tap clients can't have peer dns, so the conf is invalid
2021-07-20 07:53:23 <ikke> /etc/conf.d/openvpn
2021-07-20 07:53:26 <nbastin> oh
2021-07-20 07:53:26 <clandmeter> add a conf.d file and set what you want
2021-07-20 07:53:31 <nbastin> that side, not the openvpn.conf
2021-07-20 07:53:31 <nbastin> ok
2021-07-20 07:53:33 <clandmeter> its sourced
2021-07-20 07:53:49 <clandmeter> it needs the same naming as the init.d
2021-07-20 07:53:52 <ikke> yea
2021-07-20 07:54:03 <nbastin> ah, thanks
2021-07-20 07:54:32 <clandmeter> np, we are the wiki service :)
2021-07-20 07:54:40 <nbastin> :-/
2021-07-20 07:54:44 <nbastin> I did look at the wiki, I promise.. :-)
2021-07-20 07:56:28 <nbastin> ok that seems happier, although it still reports that it isn't working even though it is...
2021-07-20 07:56:59 <nbastin>  * WARNING: openvpn.ntn has started, but is inactive
2021-07-20 07:57:02 <nbastin> I'm not sure what that's trying to tell me
2021-07-20 07:57:09 <nbastin> but it seems quite incorrect
2021-07-20 07:57:36 <ikke> Probably the process is behaving differently than the service is configured to
2021-07-20 07:58:55 <nbastin> the tap is attached to an ovs bridge, so who knows what kind of confusion I'm creating.. :-)
2021-07-20 07:59:24 <nbastin> I really just want the system to bring the vpn up automatically, which I need to test in a reboot, but seems to be working now
2021-07-20 07:59:27 <nbastin> thanks for your help
2021-07-20 07:59:28 <ikke> No, I had the same behavior
2021-07-20 07:59:34 <ikke> So it's not related to your usage
2021-07-20 07:59:37 <nbastin> ah
2021-07-20 07:59:41 <ikke> But never dug into why it happened
2021-07-20 08:02:00 <nbastin> also anyone know if there's a way to get interface configuration to wait for ovs to start?
2021-07-20 08:02:51 <nbastin> hrm, this script does say before net, so I'm not sure what the issue is...
2021-07-20 08:03:49 <ikke> https://github.com/OpenRC/openrc/blob/master/service-script-guide.md#be-wary-of-need-net-dependencies
2021-07-20 08:04:27 <nbastin> yeah, I'm familiar with that (in general, and I even read that page yesterday), but this is sortof the reverse
2021-07-20 08:05:03 <nbastin> I'm having a problem where it tries to configure an interface before it exists, because ovs isn't started
2021-07-20 08:05:10 <nbastin> or...finished starting, maybe, I'm not sure
2021-07-20 08:05:21 <nbastin> will have to look into it more
2021-07-20 09:50:23 <Guest1049> is ovs an openrc daemon as well?
2021-07-20 09:50:25 <Guest1049> depend ovs?
2021-07-20 10:26:09 <anoduck> atka: thanks for the response, sorry for the late reply. After testing out / removing hardware, and removing the secondary Wan, I still had no improvements. I actually lost the ability to ssh into the router after this test. So, I have decided to return to as close to the last working configuration as possible. Which means installing debian and configuring shorewall. It is already handling the additional Wan interface better. 
2021-07-20 10:26:09 <anoduck> So, I think it is a driver issue not supported/covered by alpine. 
2021-07-20 10:39:46 <anoduck> All of which is sad, because I have loved alpine since first boot. Preferred Linux for years, since 1.x.x.
2021-07-20 10:40:37 <donoban> uhM, anoduck what you try to do works fine on debian?
2021-07-20 10:40:58 <ikke> anoduck: maybe if you figure out what driver is missing, we could enable it
2021-07-20 11:52:36 <Jaafarhehim[m]> <anoduck "Jaafar (he/him): run 'cat /etc/p"> Doesn't seem to be listing any groups?
2021-07-20 11:56:03 <Jaafarhehim[m]> Oh, wait, it has `10` in the GID section, and `wheel`'s GID is `10`.
2021-07-20 12:06:14 <Soomppp> This is where I would ask for help yes?
2021-07-20 12:06:35 <lopid> about what?
2021-07-20 12:07:04 <Soomppp> Driver issue
2021-07-20 12:07:14 <lopid> in alpine linux?
2021-07-20 12:08:31 <Soomppp> amdgpu loads but the screen is garbled but vesa which did work no longer loads because "vesa: Refusing to run on UEFI"
2021-07-20 12:08:47 <Soomppp> Yes
2021-07-20 12:08:57 <lopid> yes
2021-07-20 12:09:27 <lopid> did you recently update?
2021-07-20 12:10:10 <Soomppp> I had to reinstall because kde sleep for some reason killed my system
2021-07-20 12:10:51 <Soomppp> But the driver that worked was vesa. And amdgpu does not.
2021-07-20 12:13:52 <Soomppp> Vesa also does not appear in /lib/modules
2021-07-20 12:15:01 <Soomppp> But amdgpu does
2021-07-20 12:15:27 <Soomppp> Despite being blacklisted and uninstalled
2021-07-20 12:17:28 <Hello71> i'm willing to bet $20 that there is nothing wrong with alpine and anoduck actually needs to read a linux networking tutorial and learn about iptables, tcpdump, ip route get, etc
2021-07-20 12:17:51 <Hello71> Soomppp: what does "screen is garbled" mean
2021-07-20 12:18:16 <Hello71> you also can't uninstall amdgpu kernel driver
2021-07-20 12:18:23 <Soomppp> It looks like vomit. Some parts black other parts in random colors
2021-07-20 12:18:31 <Hello71> most likely you are confusing it with xf86 driver
2021-07-20 12:18:46 <Hello71> what do you mean "amdgpu loads"
2021-07-20 12:20:04 <Soomppp> If I run starts it says it's using amdgpu. But amdgpu doesn't work. So if I blacklist it to try and use vesa which did work. Then there is nothing in law, lspci or dmesg saying a driver is in use.
2021-07-20 12:20:35 <Soomppp> But starts will give no errors about operation if I cancel out from tty
2021-07-20 12:21:00 <Soomppp> startx* (autocorrect)
2021-07-20 12:21:21 <Hello71> how did you blacklist
2021-07-20 12:21:29 <Hello71> do you have "screen is garbled" in tty or only in x
2021-07-20 12:21:34 <Hello71> more details.
2021-07-20 12:21:38 <Soomppp> In grub which is what I did before
2021-07-20 12:22:02 <Hello71> what *exactly* did you do and what *exactly* were the results
2021-07-20 12:22:10 <Soomppp> I have the garble in X only tty is fine
2021-07-20 12:22:47 <Soomppp> modprobe.blacklist=amdgpu and previously it would use vesa
2021-07-20 12:23:51 <Soomppp> And everything worked beford
2021-07-20 12:28:31 <ikke> Soomppp: did you check Xorg.0.log?
2021-07-20 12:30:28 <Soomppp> I have. The system loads vesa. Then unloads it and loads fb
2021-07-20 12:33:25 <Hello71> huh, apparently busybox does support modprobe.blacklist
2021-07-20 12:33:40 <Hello71> did you check that amdgpu is actually unloaded
2021-07-20 12:33:55 <Hello71> i bet something is wrong with your hardware
2021-07-20 12:38:07 <Soomppp> It wasn't listed in lshw or lspci when I would check lsmod
2021-07-20 12:39:14 <Soomppp> The thing that bothers me is that vesa does work. Its loaded at boot in the uefi (chromebook) but alpine can not load it because uefi
2021-07-20 12:54:36 <Hello71> probably you are confused by different uses of "vesa"
2021-07-20 12:54:43 <Hello71> same with "amdgpu"
2021-07-20 15:21:26 <raub> So I am running alping as docker container and found out if I get the kernel packages (linux-lts, linux-lts-dev), their version differs from that in the container (uname). How can I force it to match the container version?
2021-07-20 15:22:05 <donoban> uhM, you should get the host kernel version
2021-07-20 15:23:39 <raub> donoban: should can a funny word: https://paste.centos.org/view/6e4ba378
2021-07-20 15:25:16 <donoban> uhm
2021-07-20 15:25:56 <raub> Exactly
2021-07-20 15:26:00 <donoban> Linux 9d324dae2231 4.18.0-240.22.1.el8_3.x86_64 is a centos kernel?
2021-07-20 15:26:09 <raub> Yeah
2021-07-20 15:26:19 <donoban> the docker containers don't run their own kernel like a virtual machine
2021-07-20 15:26:26 <donoban> they share the host kernel
2021-07-20 15:28:17 <raub> I agree. So, is my issue due to this host kernel version not matching what alpine has available?
2021-07-20 15:28:28 <Hello71> xy problem
2021-07-20 15:28:48 <donoban> you don't need to worry about that
2021-07-20 15:29:00 <donoban> do you need some special hardware access for the container?
2021-07-20 15:29:09 <donoban> like gpu passtrhough
2021-07-20 15:30:39 <raub> donoban: building netmap
2021-07-20 15:32:13 <donoban> i don't know it, is it like a kernel module?
2021-07-20 15:34:53 <raub> Yep
2021-07-20 15:36:49 <donoban> uhM, I think that it's possible but will have to deal with the miss version problem and run the container on privileged mode
2021-07-20 15:38:47 <raub> Would you also know if the pf_packet module is available for alpine?
2021-07-20 15:40:44 <donoban> I don't see anything starting with pf_ https://pkgs.alpinelinux.org/contents?file=pf_*&path=&name=&branch=edge
2021-07-20 15:46:48 <raub> sorry; I meant pf_ring. Per https://packages.ntop.org/ they do not have it for alpine. So, I might have to change the OS in my container
2021-07-20 15:48:10 <programmerq> you'd need to install any kernel modules at the host level since the host is the thing running the kernel.
2021-07-20 15:49:32 <programmerq> since it looks like pf_ring is licensed proprietary software, the userspace application may not run in the alpine userspace if it's a glibc binary only.
2021-07-20 16:05:07 <Hello71> aiui the new ntop sucks
2021-07-20 16:42:06 <nbastin> you can build ntopng from source on alpine to use it in a container
2021-07-20 17:05:04 <anoduck> Hello71: Dude, I will take your bet. Your attitude is antagonistic. 
2021-07-20 17:06:31 <anoduck> Hello71: you were literally 0 help. 
2021-07-20 17:11:54 <anoduck> donoban: I suspect it has something to do with the iseego Mifi driver. Debian supported it before alpine, and I suspect it has to do with forwarding packets in the presence of another Wan. But, I could be wrong, have not finished setup yet, and a power surge blew out the computer it sits beside. 
2021-07-20 17:25:35 <danieli> I can't speak for alpine specifically as I haven't done any remotely complex routing with it, but the driver should not matter with regards to routing and switching if you're using generic hardware
2021-07-20 21:23:34 <anoduck> danieli: I don't know. Honestly, grabbing at straws at this point, because it should have worked. Like I said before, I built the router on Ubuntu, before Ubuntu stopped support of i386 architecture. 
2021-07-21 02:06:31 <geoman> clear
2021-07-21 02:06:48 <geoman> i'm using alpine linux x64 and recently installed netsurf
2021-07-21 02:08:22 <geoman> however whenever i open it text doesnt display
2021-07-21 02:08:26 <geoman> it's just a bunch of boxes
2021-07-21 02:08:41 <valerius> missing fonts
2021-07-21 02:08:51 <geoman> how could i fix that?
2021-07-21 02:08:58 <valerius> install more fonts
2021-07-21 02:09:10 <valerius> it's kind of a battle with Alpine honestly, but you only have to do it once
2021-07-21 02:10:49 <geoman> where does netsurf look for fonts?
2021-07-21 02:11:35 <valerius> search the package repo and install font packages
2021-07-21 02:11:46 <valerius> usually they will have the substring "ttf" in them or "font"
2021-07-21 02:13:22 <valerius> sometimes I still see the boxes, but typically it's from languages I can't read anyway
2021-07-21 02:14:10 <geoman> any recommendations?
2021-07-21 02:14:27 <geoman> i'm not very picky on fonts
2021-07-21 02:15:56 <valerius> https://termbin.com/xhl7
2021-07-21 02:15:58 <valerius> that's what I've got
2021-07-21 02:16:24 <valerius> that's just a simple grep for ttf and font, some of those are not fonts :p
2021-07-21 02:16:44 <geoman> thinking about installing the xfree86 font package
2021-07-21 02:23:45 <geoman> so text works, but it doesn't seem like netsurf can display some images
2021-07-21 02:23:56 <geoman> what do i need to install for that
2021-07-21 02:26:01 <valerius> good question
2021-07-21 02:26:16 <valerius> when you run netsurf from a terminal emulator, the error messages might indicate something
2021-07-21 02:26:31 <geoman> i just put it into xterm and no errors come back
2021-07-21 02:26:34 <valerius> missing libs, errors, etc
2021-07-21 02:28:13 <geoman> i don't get any error messages, but some images, like the ones shown on the netsurf guides, show fine
2021-07-21 02:28:31 <valerius> maybe netsurf just has a lot of limitations... never used it personally
2021-07-21 02:28:37 <valerius> looks cool though
2021-07-21 02:28:41 <geoman> this is the first time i've tried it
2021-07-21 02:28:47 <geoman> it might be because i'm using cwm
2021-07-21 02:28:53 <geoman> that might add some additional errors
2021-07-21 02:29:01 <geoman> idk
2021-07-21 02:29:17 <valerius> probably not
2021-07-21 02:31:06 <valerius> https://www.netsurf-browser.org/about/screenshots/images/gtk2-bbc.png
2021-07-21 02:31:11 <valerius> oi bruv, it displays the beeb
2021-07-21 02:35:51 <geoman> it's weird how some images work and others don't
2021-07-21 02:36:16 <geoman> the link you gave worked fine
2021-07-21 02:41:28 <geoman> i'd use surf but i don't really want to compile webkit
2021-07-21 02:43:11 <valerius> all browsers suck some serious ass
2021-07-21 02:43:24 <valerius> please let me know if you find a solution to this problem :p
2021-07-21 02:43:35 <geoman> except lynx
2021-07-21 02:43:54 <geoman> i'd use lynx over everything else if it had image support
2021-07-21 02:44:12 <valerius> aren't there some notcurses browsers out there?
2021-07-21 02:45:37 <geoman> idk
2021-07-21 02:45:46 <geoman> this is the first time i've heard of notcurses actually
2021-07-21 02:46:13 <valerius> so maybe I'm wrong but the way I see this problem is that it takes way too much effort to maintain a browser, and that means money is involved, and that means anyone making a browser is not truly respecting your privacy
2021-07-21 02:46:19 <valerius> or there are other weird caveats
2021-07-21 02:46:50 <geoman> idk there's ungoogled-chromium and icecat
2021-07-21 02:55:50 <yjftsjthsd> Has anyone done User Mode Linux on Alpine? There's a libvirt-uml package, which implies that it can be a thing, but there's no actual package for uml and trying to build myself fails to compile with "undefined reference to `__sprintf_chk'" (https://gitlab.com/yjftsjthsd-g/docker_uml/-/jobs/1439938458#L1418)
2021-07-21 03:00:29 <geoman> i'm having issues with SDL2's audio backend, it keeps saying there's no audio device
2021-07-21 03:00:48 <geoman> SDLTexture rendering doesn't work as well
2021-07-21 03:07:26 <jjuran> Hi, I'm trying to copy a disk partition from one Mac to another.  The recipient is booted from an Alpine installer on a flash drive, and the donor is in Thunderbolt target disk mode.  I can see a new entry under /sys/bus/thunderbolt/devices, but nothing new in /dev/sd*.  How do I get device files for the target drive?
2021-07-21 03:09:30 <jjuran> The recipient's drive is larger than the donor's, so the recipient drive will be partitioned, and I can install Alpine on the spare partition if necessary.
2021-07-21 04:42:23 <yjftsjthsd> jjuran does it maybe show up under /dev/disk/* ? I don't have TB, but I wouldn't expect it to show up as /dev/sd* unless it's actually SATA-like
2021-07-21 05:35:52 <jjuran> yjftsjthsd: I thought maybe the trick might be to boot with Thunderbolt already connected, and to my surprise the Mac booted into the system on the target disk.  I then booted into the Recovery partition instead, and I'm using dd in Terminal to copy the image onto a second internal drive in the recipient machine.  So I can't answer your question yet.
2021-07-21 05:42:46 <jjuran> It should have been done (or nearly so) by now, but apparently the recovery OS puts the machine to sleep when idle, where "idle" includes a dd process actively writing to disk.
2021-07-21 05:50:34 <yjftsjthsd> Ew. Darwin's an... interesting OS...
2021-07-21 05:50:40 <yjftsjthsd> Good luck:)
2021-07-21 06:05:15 <jjuran> Thanks
2021-07-21 07:09:13 <jjuran> yjftsjthsd: Nope, no /dev/disk*
2021-07-21 09:11:42 <donoban> anoduck: do you can test that the hardware just works before dealing with routing/forwarding?
2021-07-21 09:42:55 <wyk72> hellp ppl
2021-07-21 09:43:12 <jjuran> t
2021-07-21 09:43:24 <wyk72> I have a question: the setup-disk script, which version of  kernel tried to install?
2021-07-21 09:43:52 <wyk72> I ran it today and installed kernel 5.10.51
2021-07-21 09:44:02 <wyk72> last week was 5.10.50
2021-07-21 09:44:09 <wyk72> the week before 5.10.43
2021-07-21 09:44:44 <wyk72> this screws up completely my setup because mkinitfs fails to "compile" properly with proper modules tree
2021-07-21 09:46:21 <wyk72> if I try to boot from a zfs volume, the zfs modules don't make it into the initram, because they are not in sync with the kernel, if it changes every now and then, randomly
2021-07-21 09:46:49 <wyk72> it's very comfusiong to say the least
2021-07-21 09:56:54 <Sheila> if you want stability, stick to a point release (e.g. 3.14) rather than edge.
2021-07-21 09:57:45 <wyk72> that's what I did
2021-07-21 09:58:00 <wyk72> but the setup installs always a different kernel
2021-07-21 09:58:04 <wyk72> my apk point to 3.14
2021-07-21 09:58:21 <wyk72> not edge
2021-07-21 09:58:47 <Sheila> are you looking at /etc/apk/repositories?
2021-07-21 09:58:55 <wyk72> yup
2021-07-21 09:59:17 <wyk72> I stick to 3.14 and LTS kernels/zfs 
2021-07-21 09:59:34 <Sheila> my next recommendation, anyway, would be generating a disk image to work from instead of running `setup-alpine` every time you need to install, and using that image as a base.
2021-07-21 10:00:31 <ikke> We do not promise fixed point releases
2021-07-21 10:00:41 <wyk72> that's what I did: i downloaded the ISO, installed it, but when generating the initram, it gets a different kernel!
2021-07-21 10:00:45 <wyk72> to me this is nonsense
2021-07-21 10:01:11 <wyk72> it seems the setup script downloads a different kernel from the apk during install
2021-07-21 10:01:18 <wyk72> it gets messy
2021-07-21 10:01:18 <ikke> Most people expect an up-to-date kernel within the LTS release
2021-07-21 10:02:14 <wyk72> I can agree, but what if I need a mkinitfs with other features into it , like ZFS or nvme or other stuff
2021-07-21 10:02:38 <wyk72> the mkinitfs.conf is ratyher standrd, and lacks most of the features
2021-07-21 10:04:44 <wyk72> during install, the vmlinuz image is a 5.10.34 kernel, but into the mkinitfs you get 5.10.50, with wrong zfs modules and the rest, so the system gets screwed at boot
2021-07-21 10:05:04 <wyk72> it's a total mess
2021-07-21 10:05:48 <artok> adding features asap to mkinitfs.conf ?
2021-07-21 10:30:58 <clandmeter> setup-disk has some checks if nvme or others are needed and add them to mkinitfs.conf automatically
2021-07-21 10:31:33 <clandmeter> wyk72: if you upgrade your system, zfs should follow the kernel.
2021-07-21 10:34:56 <wyk72> I need something more "manual" I guess
2021-07-21 10:40:41 <Sheila> you could use a rootfs image…
2021-07-21 10:54:47 <clandmeter> wyk72: do you use aports zfs package?
2021-07-21 10:56:43 <wyk72> no I just use the standard .iso (3.14 x86-64)
2021-07-21 10:56:52 <wyk72> I am trying to use extended
2021-07-21 10:56:58 <wyk72> to get zfs out of the box
2021-07-21 10:57:22 <wyk72> but still, when it installs, it gets differente kernel/modules into mkinitfs
2021-07-21 10:57:29 <wyk72> and I find it very very strange
2021-07-21 10:57:50 <wyk72> and anso
2021-07-21 10:57:56 <wyk72> the package gummiboot moved to edge
2021-07-21 10:58:06 <wyk72> for whatever reason
2021-07-21 10:58:26 <wyk72> I find it difficult to track packages going bach and forth from repositories
2021-07-21 10:58:32 <wyk72> it gets tricky
2021-07-21 10:58:57 <wyk72> sometimes edge sometimes testing sometimes community
2021-07-21 10:58:59 <wyk72> omg
2021-07-21 10:59:04 <wyk72> it's so weird
2021-07-21 11:11:41 <xsteadfastx> im playing around with alpine and qemu. i tried the netboot variant to install something... but alpine-setup doesnt find any harddrives i added to qemu. no problem with the standard iso and qemu. lacks the virt kernel moduldes for harddrives?
2021-07-21 11:47:30 <lopid> alpine-virt was fine for me
2021-07-21 12:05:01 <xsteadfastx> lopid: even when using setup-alpine and harddrives?
2021-07-21 12:06:01 <lopid> yes: -device virtio-blk-pci,drive=disk0,iothread=io1 -drive if=none,id=disk0,file="$image",aio=threads,cache.direct=on
2021-07-21 12:06:44 <lopid> -object iothread,id=io1
2021-07-21 14:35:23 <bovis> Using wpa_cli, I'm seeing my network as connected, but I can't ping. Done add_network and enable_network, the ssid is correct, it's open so key_mgmt is set to none. Reassociating didn't help. What can I do next?
2021-07-21 14:37:30 <ikke> bovis: do you have an IP? Do you have a default route?
2021-07-21 14:41:19 <bovis> Ikke: IP is shown as 192.168.1.6. Unsure how to answer your second question.
2021-07-21 14:41:47 <zcrayfish> ip route show
2021-07-21 14:43:01 <lopid> or "ip r" for short
2021-07-21 14:43:06 <bovis> Yes 192.168.1.1
2021-07-21 14:43:32 <ikke> Can you ping that IP?
2021-07-21 14:44:53 <bovis> Ikke: To 192.168.1.1 there's 100 percent packet loss
2021-07-21 14:45:19 <danieli> then your problem is on L2 or L1
2021-07-21 14:45:22 <danieli> what does your arp table show?
2021-07-21 14:45:40 <danieli> can you see your router in the output of "arp -a"?
2021-07-21 14:46:08 <lopid> (or "ip neigh")
2021-07-21 14:47:28 <danieli> ip neigh / ip n works too
2021-07-21 14:48:07 <danieli> i think the full name of that one is "neighbour" fwiw
2021-07-21 14:48:14 <lopid> 
ACTION nods

2021-07-21 14:48:38 <bovis> arp -a shows: ? (192.168.1.1) at <incomplete> on wlan0
2021-07-21 14:49:06 <danieli> seems like your computer doesn't know your router's MAC address in that case, suggesting something is wrong with the physical link or on the ethernet layer
2021-07-21 14:49:16 <danieli> what does "ip l" say about the link itself?
2021-07-21 14:49:18 <ikke> danieli: this is wireless
2021-07-21 14:49:38 <danieli> ikke: oh right, i forgot for a second
2021-07-21 14:51:11 <bovis> Lopid: ip neigh: FAILED
2021-07-21 14:51:59 <bovis> danieleli the link to wlan0 is up
2021-07-21 14:57:42 <lopid> anything useful on the router?
2021-07-21 15:00:37 <bovis> lopid I don't have access to any equipment. This computer successfully connects to my home router. I'm in a public space at the moment.
2021-07-21 15:02:04 <lopid> do you need to go to a special web page to be let on the network, or something?
2021-07-21 15:04:33 <bovis> lopid Not here. Library with an open network and no splash page.
2021-07-21 15:42:23 <bovis> Ikke lopid Daniele: 'udhcpc -i wlan0' made things work. Why was that step necessary?
2021-07-21 15:43:06 <ikke> bovis: didn't you have an IP on the wlan0 interface yet?
2021-07-21 15:44:40 <ikke> wpa_supplicant only manages the wireless part, it does not request IP addresses
2021-07-21 15:45:25 <bovis> ikke: 'ip addr show' showed an ip for my device previously. Looks much different now.  10.26... instead of 192.168...
2021-07-21 15:45:50 <ikke> So apparently you still had on old (static?) address on there
2021-07-21 15:46:53 <bovis> Must have been. Crap I forgot about the static IP.
2021-07-21 15:49:10 <bovis> Thanks. I appreciate the guidance.
2021-07-21 16:44:39 <onelegend> what kernel hardening measures does alpine linux provide nowadays? I believe in the past grsec and PaX were used, and I was wondering if there was something similar being used instead
2021-07-21 16:55:40 <xordspar0> onelegend: I believe it's just a standard mainline kernel these days. https://git.alpinelinux.org/aports/tree/main/linux-lts?h=master
2021-07-21 16:55:56 <ikke> it is
2021-07-21 16:55:57 <programmerq> yeah, there's a few patches but that's it.
2021-07-21 17:04:11 <donoban> well, building it with muslc and some protection gcc flags could be considered "hardening"?
2021-07-21 17:04:28 <ikke> kernel does nothing with musl
2021-07-21 17:04:38 <donoban> ahh lol
2021-07-21 17:04:49 <donoban> I didn't know
2021-07-21 17:05:05 <invoked> a lot of hardening only makes sense if alpine is the host, anyway
2021-07-21 17:05:16 <donoban> muslc/glibc are like a layer upper linux?
2021-07-21 17:07:22 <Hello71> it's also not muslc
2021-07-21 17:10:47 <programmerq> donoban the C library is the lowest level user space library for C applications in userspace.
2021-07-21 17:11:51 <programmerq> the kernel has zero library dependencies and is fully self-contained.
2021-07-21 17:12:18 <donoban> ok I see
2021-07-21 17:12:40 <donoban> the syscalls are exposed by the kernel with kernel headers?
2021-07-21 17:13:04 <tiotags> syscalls are exposed by interrupts and the amd syscall_enter instruction
2021-07-21 17:13:50 <programmerq> yeah, there's no C stuff that is required for an application to use a syscall. non-C applications, for example, run fine on a linux kernel. They can make syscalls too.
2021-07-21 17:14:40 <tiotags> I'm not sure if syscall numbers are exposed in headers, are they ?
2021-07-21 17:14:42 <programmerq> you could have a non-C kernel on an x86_64 system that implements syscalls and its userspace applications could be in any language and use those syscalls
2021-07-21 17:15:45 <programmerq> I think the posix standard may define what syscalls are and both the kernel and userspace implementations follow that.
2021-07-21 17:16:23 <invoked> man 2 syscalls
2021-07-21 17:16:23 <programmerq> of course, linux has some differences from vanilla posix, but the linux userspace application compilers would be aware of those
2021-07-21 17:16:40 <programmerq> that's my understanding, but I'm also remembering back to my very few computer science classes years ago.
2021-07-21 17:17:10 <programmerq> but a userspace application doesn't interact with the kernel via C library linking at all.
2021-07-21 17:17:12 <donoban> hehe thank you all
2021-07-21 17:17:32 <programmerq> the kernel headers are really only needed if you want to compile something for that particular kernel, like a third party module.
2021-07-21 17:18:20 <donoban> so both kernel and C lib have same syscalls defined in their own headers
2021-07-21 17:18:26 <tiotags> lists syscalls but not numbers
2021-07-21 17:18:48 <programmerq> it may be the compiler and not the C library. I'm not sure where that actually is.
2021-07-21 17:18:53 <tiotags> that's suspicious
2021-07-21 17:20:38 <tiotags> donoban good way to put it
2021-07-21 17:21:06 <programmerq> yup, it's the C library.
2021-07-21 17:21:09 <programmerq> neat
2021-07-21 17:21:23 <programmerq> this is why I like IRC questions. I like having something that compels me to think. :)
2021-07-21 17:22:52 <donoban> hehe I felt a little idiot thinkg that kernel could be build with glib or musl but It's nice too :D
2021-07-21 17:24:51 <tiotags> donoban there's nothing that stops a kernel from being linked with a userspace library, just that it needs to be a static lib
2021-07-21 17:25:35 <donoban> uhM, but in theory it doesn't use all self-included code?
2021-07-21 17:25:58 <donoban> do you mean that you could add some lib if you want?
2021-07-21 17:26:12 <tiotags> the kernel does use printf, time and others
2021-07-21 17:27:30 <tiotags> yes you could ... in theory ... but the kernel doesn't call it's own functions as printf/time/etc
2021-07-21 17:27:47 <tiotags> but kprintf/kwhatever
2021-07-21 17:29:03 <donoban> https://kernelnewbies.org/FAQ/LibraryFunctionsInKernel
2021-07-21 17:30:15 <donoban> t0d0 add "read kernelnewbies faq" :D
2021-07-21 17:31:54 <tiotags> yes, but see that they mentioned that you can 'adapt' libraries for kernel use
2021-07-21 17:32:29 <tiotags> it's just like they said 'error prone'
2021-07-21 17:33:35 <donoban> gotcha
2021-07-21 18:57:13 <Hello71> glib isn't the same thing as glibc either
2021-07-21 18:58:24 <snickers> hi, i have problem with dns. In my configuration there is 2 nameserver in resolv.conf, first resolve my internal domain and second don't. On any other image everythigh works, but on alpine curl can't resolv internal domain in this configuration.
2021-07-21 18:58:47 <snickers> someone know how to solve this?
2021-07-21 18:59:15 <ikke> snickers: https://www.openwall.com/lists/musl/2015/09/15/2
2021-07-21 19:00:48 <ikke> summary: you should not rely on the order of nameservers in resolv.conf
2021-07-21 19:03:49 <snickers> ok thanks
2021-07-21 19:04:20 <ikke> more examples: https://github.com/kubernetes/kubernetes/issues/78138
2021-07-21 21:04:37 <snickers> have one more problem, I've got some server that return serverfial for all domain that don't have AAAA record, can I prefer ipv4 over ipv6 in alpine?
2021-07-21 21:07:33 <lopid> don't know about prefer, but you can disable ipv6 at the kernel level
2021-07-21 21:09:24 <snickers> i try but it don't help. 3.12 works without prblem but 3.13 can't resolve A record for domain that return serverfail for AAAA record
2021-07-21 21:10:25 <ikke> snickers: ask #musl
2021-07-21 21:10:40 <invoked> if you're using unbound you can set 'do-ip6: no' under 'server:'
2021-07-21 21:10:58 <ikke> But these are known issues, dns servers should not return SRVFAIL
2021-07-21 21:11:01 <ikke> in those cases
2021-07-21 21:11:17 <ikke> but sadly there is a lot of broken software
2021-07-21 21:11:53 <snickers> i can't chaneg DNS configuration so try to solve this on alpine level
2021-07-21 21:14:01 <invoked> i think in either case (what i suggested, or with ipv6 disabled in the kernel) that doesn't stop ip6 dns lookups
2021-07-21 21:14:12 <invoked> from a dns record pov
2021-07-21 21:15:34 <snickers> I wonder if dnsmasq can filter that behavior
2021-07-21 21:17:00 <invoked> probably.
2021-07-21 21:23:00 <invoked> i searched and found https://gist.github.com/noseka1/6b7cbeca199db9c4823324bc31214bb2
2021-07-21 21:23:48 <invoked> there's probably a few ways of solving it, but that's one way i guess.
2021-07-21 21:24:43 <snickers> I'll try
2021-07-22 00:38:42 <skintimidation> What's up? Looks like if I want to 1) not use grub and 2) set up btrfs, I shouldn't use setup-disk. Is this a reasonable assumption? I'm looking through the script and seeing only grub-efi integration for USE_EFI.
2021-07-22 00:39:08 <skintimidation> I suppose I could just manually modify setup-disk to do what I need, now that I think about it. That's the nice part about it being bash.
2021-07-22 01:38:41 <asdflkj> I'm super late but GNU unifont is a good font to have, it has glyphs for every printable code point in the Unicode Basic Multilingual Plane (BMP), but as tiny bitmaps so the .ttf is only 12 Mb (and other formats 1 Mb)
2021-07-22 02:44:13 <Skaface> Hey does anyone know how mdef determines which order to assign the network card devices?
2021-07-22 02:44:25 <Skaface> mdev
2021-07-22 03:05:36 <onelegend> what's the most secure operating system
2021-07-22 03:05:38 <onelegend> alpine linux
2021-07-22 03:05:40 <onelegend> ez
2021-07-22 03:06:27 <Skaface> openbsd, just ask them
2021-07-22 03:18:36 <Hello71> i don't think mdev adjusts network interface names at all
2021-07-22 03:19:24 <Hello71> skintimidation: you can pass a mounted path instead of a device
2021-07-22 03:19:47 <Hello71> basically using only half of setup-disk
2021-07-22 03:19:58 <Hello71> i kind of feel setup-disk should be split into two scripts to make it more clear that you can use one half and not the other
2021-07-22 03:20:12 <skintimidation> Hello71: Looks like that'll still run grub-install for me
2021-07-22 03:20:36 <skintimidation> When I can use efibootmgr alone instead; which I should just modify to not use setup_grub()...
2021-07-22 03:20:43 <Hello71> hm.
2021-07-22 03:21:11 <Hello71> i think you might be right
2021-07-22 03:21:15 <Hello71> make it three scripts then :p
2021-07-22 03:21:20 <skintimidation> Looks like I could just make a $BOOTLOADER case with my specific settings
2021-07-22 03:21:28 <Hello71> yes
2021-07-22 03:21:40 <Hello71> but i think efibootmgr is too system-specific to make a generic script
2021-07-22 03:21:43 <skintimidation> Yeah, in my head, I wish I knew more fundamentally, at the high level, what the linux installation process was all about
2021-07-22 03:22:02 <Hello71> https://wiki.archlinux.org/title/installation_guide
2021-07-22 03:22:32 <skintimidation> Ah, will this be a little more along the lines of alpine's compared to the gentoo wiki?
2021-07-22 03:22:39 <skintimidation> Gentoo is really informative, but goddamn
2021-07-22 03:22:50 <Hello71> haven't read the gentoo handbook in a long time
2021-07-22 03:23:44 <skintimidation> next in mind is how i make sure btrfs can actually run; is that where i chroot and install the obviously required packages?
2021-07-22 03:24:14 <Hello71> the gentoo instructions are also relevant but many sections can be skipped. for example you can skip the section on compiling your own kernel and use a distro kernel instead
2021-07-22 03:25:15 <Hello71> alpine is interesting in that it normally uses neither the arch method of $pkgmgr --root install base nor the gentoo method of unpacking a prebuilt tarball, and instead semi-clones the existing system
2021-07-22 03:25:37 <Hello71> it would probably be nice if alpine had an arch-chroot equivalent
2021-07-22 03:25:55 <Hello71> or if util-linux got a linux-chroot or something
2021-07-22 12:39:31 <skintimidation> Another big question: Am I reading this correctly? It seems like if you pass a mounted root to setup-disk, it overrides your choice of bootloader to syslinux.
2021-07-22 12:40:32 <skintimidation> nope, it doesn't; it uses the test builtin to check for syslinux; my bad
2021-07-22 13:15:37 <skintimidation> If I'm not using syslinux, update-extlinux.conf isn't necessary, right?
2021-07-22 13:31:05 <Hello71> it's used for grub config generation
2021-07-22 13:31:09 <Hello71> because ugh
2021-07-22 13:31:11 <skintimidation> I'm also curious how mkinitfs.conf gets pulled into the mix, as I can't find it being used anywhere in the script
2021-07-22 13:31:28 <skintimidation> Hello71: lol, so if i'm not planning on using grub, i'll just avoid that
2021-07-22 13:31:51 <skintimidation> it's unfortunate that now it's all abstracted behind grub-install
2021-07-22 13:33:38 <Hello71> initramfs is totally separate
2021-07-22 13:34:10 <Hello71> other distros also have initramfs
2021-07-22 13:34:25 <Hello71> except default gentoo i guess
2021-07-22 13:35:02 <skintimidation> right--i'm just writing my own simplified script to run instead of setup-disk, so i'm working through it and understanding what exactly puts stuff in the order needed
2021-07-22 13:35:17 <skintimidation> and it looks like grub_install or something similar does a lot of stuff that isn't detailed in the script
2021-07-22 13:35:32 <skintimidation> er, grub-install
2021-07-22 13:36:13 <skintimidation> additionally on linux, if you've got a swapfile instead of a partition, do you just put that in the fstab?
2021-07-22 13:38:00 <skintimidation> i don't see a part in the script [inside of install_mounted_root()] that calls swapon, so i'm not sure how it konws where the swap is at all
2021-07-22 13:38:13 <skintimidation> knows, even; i suppose i can just configure that after first boot
2021-07-22 13:39:15 <skintimidation> other docs say it's called on boot, but what script calls that, so i can modify it?
2021-07-22 13:57:36 <invoked> coreos had a thing called ignition which allows you to manipulate disks during initramfs
2021-07-22 13:57:45 <invoked> not sure what happened to that since redhat bought them
2021-07-22 13:58:22 <invoked> but that might be interesting for you?  i dunno.
2021-07-22 13:59:14 <skintimidation> i don't think so; how are swapfiles used on other systems? does swapon do a persistent change to some config somewhere?
2021-07-22 13:59:29 <skintimidation> what normally sets swap during a standard boot?
2021-07-22 14:00:42 <invoked> that's normally after initramfs. the point of initramfs is to mount the root filesystem
2021-07-22 14:01:04 <invoked> anyway i was just suggesting ignition if you were doing something weird during boot related to disks
2021-07-22 14:01:20 <invoked> i didn't carefully read what you wrote
2021-07-22 14:04:10 <skintimidation> ah, i see.
2021-07-22 14:04:41 <skintimidation> but no one types in "swapon /dev/sda3" or whatever on boot, so what script in a normal alpine install actually calls that? or does it need to?
2021-07-22 14:07:27 <invoked> that's handled by init, which in your case is probably openrc, which means the localmount service
2021-07-22 14:07:56 <invoked> that will read fstab and if you have a line for swap in there, it mounts it at that time
2021-07-22 14:08:30 <invoked> bios -> bootloader -> kernel -> initramfs -> init
2021-07-22 14:08:52 <invoked> mounting swap is part of init (regardless of whatever init system you're using)
2021-07-22 14:09:24 <skintimidation> oh, sweet
2021-07-22 14:09:38 <skintimidation> so swapfiles are just files mounted as swap instead of partitions?
2021-07-22 14:09:49 <skintimidation> do they need to have special settings like loopback or whatever, then?
2021-07-22 14:10:04 <invoked> they can be whatever you want
2021-07-22 14:10:10 <skintimidation> (i'll have to generate my own fstab instead of using setup-disks)
2021-07-22 14:10:13 <skintimidation> huh, wild
2021-07-22 14:11:00 <invoked> if you're using btrfs, it's a file on a subvolume made for swap specifically
2021-07-22 14:11:32 <invoked> (caveats apply)
2021-07-22 14:11:38 <invoked> it just depends on how you want to arrange your disks
2021-07-22 14:11:49 <skintimidation> that was another part of where i was confused; can i use btrfs to make the volumes FROM the live environment, without it getting confused on my current livedisk root as well?
2021-07-22 14:12:25 <invoked> you can make subvolumes at any time
2021-07-22 14:12:35 <invoked> assuming there's a btrfs fs there
2021-07-22 14:13:08 <skintimidation> huh, interesting. in that case, i'll just make the gpt disk, an esp, and then a big partition
2021-07-22 14:13:29 <skintimidation> and then btrfs can have the last partition, right? i can't think of another partition i need to care about
2021-07-22 14:13:34 <invoked> yeah but doing swap on a subvolume in btrfs has to be done a specific way
2021-07-22 14:13:36 <invoked> so read up on it
2021-07-22 14:13:43 <invoked> if that's what you're going to do
2021-07-22 14:13:52 <skintimidation> do modern SSDs still need to have their partitions undersized somewhat so they can recover from wear?
2021-07-22 14:14:27 <nickersonm> skintimidation: SSDs typically have disk-managed excess space
2021-07-22 14:14:34 <skintimidation> yeah, i definitely have to do swap on the subvolume if i do it the way i mentioned, as the swapfile existing on my main subvolumes would disable some other utilities (snapshotting, etc)
2021-07-22 14:15:14 <skintimidation> oh, ok, nickersonm. and they won't take any unallocated stuff for themselves afterward?
2021-07-22 14:16:36 <nickersonm> skintimidation: they tend to do better with more free space, but the various full-write tests I've seen didn't impact performance terribly with modern SSDs.  Always model-dependent, though.  Unless you're going to use the disk at 90% capacity, don't bother downsizing the partitions.  It's free space that they care about, not partitioned space.
2021-07-22 14:17:14 <skintimidation> Ah, sweet.
2021-07-22 14:17:28 <nickersonm> skintimidation: also, don't forget to disable CoW (chattr +C) on any swapfiles you make on btrfs (or mount nodatacow, which it sounds like you might be doing).  Easiest to just have a swap partition imo, or just use a small zram swap if you have plenty of real RAM.
2021-07-22 14:17:33 <skintimidation> Alright, quick nap before I come back. Thanks for the help so far.
2021-07-22 14:18:20 <onelegend> what's a zram swap?
2021-07-22 14:18:27 <onelegend> like with real RAM?
2021-07-22 14:19:05 <nickersonm> onelegend: yea, it's essentially a compressed ramdisk used as swap
2021-07-22 14:19:05 <invoked> the swap-on-btrfs thing is nice to have if you're running luks on raw disk. (i know, corner case, but interesting.)
2021-07-22 14:19:31 <nickersonm> onelegend: https://fedoraproject.org/wiki/Changes/SwapOnZRAM#zram_Basic_function
2021-07-22 14:19:59 <onelegend> ah
2021-07-22 14:22:06 <skintimidation> invoked: i settled on using ATA security for my fde; i can set that up at any time, luckily
2021-07-22 14:22:19 <skintimidation> it just relies on the research i did for this particular drive being accurate
2021-07-22 14:22:21 <skintimidation> lol
2021-07-22 14:22:23 <nickersonm> invoked: oh yeah, unpartitioned btrfs [on unpartitioned luks]
2021-07-22 14:22:47 <invoked> for paranoia reasons :)
2021-07-22 14:23:26 <invoked> but then if you're going to go that far, you'll want a removable bootloader
2021-07-22 14:23:39 <invoked> so on and so forth
2021-07-22 14:24:06 <invoked> evil maids exist!
2021-07-22 14:24:28 <onelegend> https://wiki.alpinelinux.org/wiki/Gaming_on_Alpine
2021-07-22 14:24:42 <onelegend> I love how the person who wrote the article is like "Gaming on Alpine is a thing"
2021-07-22 14:25:20 <skintimidation> yeah, i was thinking about that quite earlier, which is why i settled on ata security--the drive doesn't even present itself without the password, and you can disable the master pass, and modern controllers properly tie the password to the encryption key cryptographically, lmao
2021-07-22 14:25:43 <invoked> i don't trust on-device security
2021-07-22 14:25:58 <invoked> but that's me.  too many years of garbage to even bother giving it more chances
2021-07-22 14:28:13 <skintimidation> invoked: i think i can agree in a lot of ways; i'm worried that eventually something in the firmware will explode and lock me out of the data forever unless i get an xray machine or some shit, lol
2021-07-22 14:28:16 <skintimidation> but that'll just encourage me to do proper backups, right? RIGHT?
2021-07-22 14:28:19 <skintimidation> then again, if crucial would just publish their firmware source, i'd be able to check if it was using some real crypto for the drive, as well
2021-07-22 14:28:34 <skintimidation> it's unfortunate, since the diagram in my head of "CRYPTO ON CRYPTO" seems so wasteful compared to "good crypto once"
2021-07-22 14:29:10 <invoked> well. the thing is, most of the time, the device manufacturers don't take security seriously. it's just marketing
2021-07-22 14:29:36 <invoked> it's a thing many of them just tack on so they can charge more for it
2021-07-22 14:30:38 <invoked> and by the time anyone in the security field has an opportunity to inspect the device and find it's crap, everyone has moved on
2021-07-22 14:31:37 <invoked> so imo, stick with what works across everything, and gets regular attention: software
2021-07-22 15:29:25 <dekedro[m]> 
ACTION  < https://matrix.org/_matrix/media/r0/download/matrix.org/xRLRZntgOfwBLvJkEaHIdtPE/message.txt >

2021-07-22 15:30:08 <ikke> dekedro[m]: https://tpaste.us/r6eL 
2021-07-22 15:31:12 <lopid> dekedro[m]: http://vpaste.net/vZImh
2021-07-22 15:31:19 <ikke> https://tpaste.us/PrMx 
2021-07-22 15:32:44 <ikke> regarding your problem, yes, I get it as well
2021-07-22 15:35:04 <dekedro[m]> Sorry, my bad, I will try to not use multiline next time. About krita: Looks like a bug, I will open issue on gitlab
2021-07-22 15:35:30 <ikke> It probably needs to be rebuilt
2021-07-22 15:36:25 <ikke> quazip-1.1-r0 provides:
2021-07-22 15:36:27 <ikke> so:libquazip1-qt5.so.1.0.0=1.1
2021-07-22 15:36:51 <dekedro[m]> Doesn't look like rebuild will help, no?
2021-07-22 15:42:16 <Hello71> invoked: "part of init" is not really correct
2021-07-22 15:42:41 <Hello71> alpine uses busybox init, not openrc-init. busybox init invokes openrc, but not as pid1
2021-07-22 15:43:22 <Hello71> skintimidation: i don't think ssds ever actually required undersized partitions
2021-07-22 15:43:47 <Hello71> if that were true, then making a too-big partition would permanently damage the ssd
2021-07-22 15:44:01 <Hello71> > modern controllers properly tie the password to the encryption key cryptographically
2021-07-22 15:44:57 <Hello71> no they don't. most drives don't even advertise that the ATA password is used for encryption, and even the ones that do usually suck balls: https://www.ieee-security.org/TC/SP2019/papers/310.pdf
2021-07-22 16:01:49 <invoked> Hello71: fair. i'm on runit, but i was attempting to explain the overall boot process
2021-07-22 16:02:18 <invoked> more specifically to their question about how swap gets handled
2021-07-22 18:03:10 <skintimidation> I distinctly remember krita not working at all on my laptop ages (last year) ago.
2021-07-22 18:05:19 <skintimidation> any article on ata security.
2021-07-22 18:05:19 <skintimidation> Hello71: Just for the ultimate dox, I actually read that paper and emboldened my feelings about it. The Crucial MX100 failed to cryptographically secure the key with the password. The MX200 actually did this, and the only way the researchers could get an "easy gotcha" exploit on the MX200 was if you didn't clear the master-pass, and they say "it's unlikely to be cleared," which is true if you aren't actually serious about reading
2021-07-22 18:06:13 <skintimidation> I'm using the MX500, and I do realize that it's possible that the controller being a later generation might have rolled back the security, but I'm hopeful enough that the two-datapoint trend I saw made it through the rest of development. ;D
2021-07-22 18:06:39 <Hello71> or use luks and have actual security
2021-07-22 18:10:15 <onelegend> LUKS
2021-07-22 18:10:26 <minimal> skintimidation: in the past several vendors implementation of drive encryption have been shown to be woeful
2021-07-22 18:16:54 <skintimidation> Hello71: The problem again with luks is that real full-disk encryption involves me doing EVEN MORE crazy shit to avoid grub
2021-07-22 18:17:27 <skintimidation> minimal: Right--though it's a case by case thing, and that paper seemed to not analyze the underlying encryption much, sadly
2021-07-22 18:18:15 <skintimidation> my bad, they verified that it was the mx300 that was good. they progressively got better over three drives
2021-07-22 18:18:24 <skintimidation> which i think is sufficient to project going forward
2021-07-22 18:18:33 <minimal> minimal: what's your reason(s) for avoiding Grub?
2021-07-22 18:18:42 <minimal> s/minimal/skintimidation/
2021-07-22 18:18:42 <alpine-meetbot> minimal meant to say: skintimidation: what's your reason(s) for avoiding Grub?
2021-07-22 18:19:00 <skintimidation> heh; you answered it in your typo, minimal. ;P
2021-07-22 18:19:17 <skintimidation> if i can just use the kernel to boot, then i'll just use the kernel to boot
2021-07-22 18:19:54 <minimal> I assume you want to use drive (hardware) encryption to handle the fact that the ESP partition can't be encrypted
2021-07-22 18:20:46 <minimal> so instead you'll be at the mercy of the drive vendor's (in)competence
2021-07-22 18:20:57 <skintimidation> right, and then if you use grub i learned that it can't decrypt luks2, and what the hell
2021-07-22 18:22:21 <minimal> skintimidation: re: luks2, you mean on Alpine? yeah there's something wrong with grub-mkconfig on Grub 2.06 regarding luks2, I started investigating it - basically its not putting "insmod luks2" and "cryptmount -u ....." into the generated grub.cfg. I need to find some more time to develop a fix for it
2021-07-22 18:22:24 <skintimidation> i'm probably safer saying it here, but being at the behest of gnu (in)competence is also part of the grub problem--i know it works, but then you get ludicrous bugs like the 28-backspace auth bypass
2021-07-22 18:23:58 <skintimidation> and at the entire end of all of it, there's just no need to use grub when i don't change the options very often and the kernel can boot its own damn self; i hear you can also just put in an initramfs into your kernel to do all of the decryption, etc, etc, and i was considering moving things off of a disk, but i'm much closer to actually getting my system up and running with the model i have in my head
2021-07-22 18:24:15 <skintimidation> because i guarantee i'm going to fuck it up given the resources i've read and how sparse they are for modern implementations
2021-07-22 18:25:02 <skintimidation> the fact that there is no single resource that does anything regarding encryption with UEFI boot that doesn't involve grub, for example, makes it just hella impossible to parse how i'd go about tying together two guides, since there's no documentation telling me at a higher level what needs to be where
2021-07-22 18:25:15 <minimal> yes initramfs in Alpine handles decryption - in a FDE conifg that's *after* Grub itself has decrypted /boot to get at the kernel and initramfs files
2021-07-22 18:25:40 <skintimidation> (not accessed well by someone who's been using google for apparently too long before they started making it impossible to find anything)
2021-07-22 18:25:41 <minimal> or in the situation of "everything but /boot encrypted" then the initramfs also takes care of it
2021-07-22 18:25:50 <skintimidation> i can't even find in the alpine scripts where the initramfs is created
2021-07-22 18:26:33 <minimal> skintimidation: I've been doing Alpine encryption with Syslinux as well as Grub, including UEFI. Obviously with Syslinux you can't encrypt /boot
2021-07-22 18:27:03 <minimal> BTE I recently submitted a mkinitfs MR for adding SSH-based remote encryption unlock at boot time
2021-07-22 18:27:08 <minimal> s/BTE/BTW/
2021-07-22 18:27:08 <alpine-meetbot> minimal meant to say: BTW I recently submitted a mkinitfs MR for adding SSH-based remote encryption unlock at boot time
2021-07-22 18:27:33 <minimal> initramfs is created by the mkinitfs tool
2021-07-22 18:32:48 <minimal> skintimidation: you might be interested in the script I'm working on to see how I set up Grub and Syslinux including for encryption
2021-07-22 18:33:13 <donoban> I have some alpine server with ssh unlocking but I need to look careful when upgrading kernel
2021-07-22 18:34:05 <minimal> donoban: your own custom code? I submitted this 1 week ago: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/86
2021-07-22 18:35:03 <donoban> well I configured it using some example, maybe the wiki, but I remeber that std upgrade breaks it
2021-07-22 18:36:07 <minimal> donoban: breaks it how exactly? the initramfs should be auto re-generated when kernel updates. Are you using run-from-ram or data mode? I'm only using "sys" based installs
2021-07-22 18:36:42 <skintimidation> how would remote encryption unlock work anyway?
2021-07-22 18:36:50 <skintimidation> some pxe server running a minimal ssh situation?
2021-07-22 18:36:59 <donoban> I think that it just ignores it, the standard initramfs generation ignores that module/freature... so I needed to run a custom mkinitfs command
2021-07-22 18:37:15 <donoban> I didn't boot it in months hehe
2021-07-22 18:37:32 <donoban> I tried to do some routing experiment
2021-07-22 18:37:48 <minimal> skintimidation: nope, initramfs start dropbear and waits for a key-connection, then prompts you are usual for passphrase, then unlocks LUKS as usual, kills dropbear, and boot continues as usual
2021-07-22 18:38:12 <donoban> I guess that your solution is properly integrated with kernel upgrades
2021-07-22 18:39:27 <skintimidation> how do you do any initramfs modifications like that?
2021-07-22 18:39:47 <minimal> skintimidation: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/86
2021-07-22 18:40:23 <minimal> though its been a week with no comments/feedback - mkinitfs development seems to be infrequent so it might be weeks or months before this gets merged (if at all)
2021-07-22 18:41:21 <donoban> I think that also a LUKS handling option for setup-disk would be nice
2021-07-22 18:42:19 <minimal> donoban: I thought setup-disk handled LUKS? Haven't used it in more than 1 year, I create my own (encryped) disk images using my own script
2021-07-22 18:44:28 <donoban> I think that it doesn't handle LUKS setup
2021-07-22 18:44:54 <donoban> you have to luksFormat, luksOpen... and then call setup-disk 
2021-07-22 18:45:07 <minimal> donoban: maybe I could interest you in my script then :-)
2021-07-22 18:46:10 <donoban> what is your script?
2021-07-22 18:46:38 <minimal> https://github.com/dermotbradley/create-alpine-disk-image
2021-07-22 18:47:01 <minimal> need to push an update to it shortly - am finishing up a major restructuring of it at present
2021-07-22 18:47:33 <minimal> it designed to create server-oriented disk images that you then write to SSD or HDD using "dd"
2021-07-22 18:48:02 <donoban> uhM
2021-07-22 18:49:04 <nickersonm> skintimidation: for a grub alternative, I know systemd-boot does FDE just fine; you may want to look in to gummiboot on alpine.  Not sure what support or capability there is, but there is a package.
2021-07-22 18:49:26 <skintimidation> i can just use the kernel
2021-07-22 18:49:27 <skintimidation> i don't see why that's a bad idea
2021-07-22 18:50:14 <Hello71> gummiboot is also unmaintained and also doesn't support encrypted /boot
2021-07-22 18:50:43 <donoban> it looks nice minimal 
2021-07-22 18:50:49 <minimal> skintimidation: that would work as long as you pass command options like "cryptroot=UUID" etc to the initramfs
2021-07-22 18:50:51 <Hello71> imo encrypted /boot is useless in 99.99% of cases
2021-07-22 18:51:37 <minimal> donoban: thanks. The rework will make things somewhat simpler/run less code with root privs
2021-07-22 18:53:40 <minimal> Hello71: encrypted /boot might have made more sense in the past (with BIOS) but as UEFI needs an encrypted ESP partition it makes less sense now
2021-07-22 18:53:53 <minimal> s/an encrypted/an unencrypted/
2021-07-22 18:53:53 <alpine-meetbot> minimal meant to say: Hello71: encrypted /boot might have made more sense in the past (with BIOS) but as UEFI needs an unencrypted ESP partition it makes less sense now
2021-07-22 18:54:03 <Hello71> what is the threat model
2021-07-22 18:55:25 <minimal> worry that kernel and/or initramfs could be modified - these days SecureBoot is intended to "take care" of that without needing encryption
2021-07-22 18:56:47 <nickersonm> skintimidation: you may find this informative: https://www.cogitri.dev/posts/04-secure-boot-with-unified-kernel-image/
2021-07-22 19:06:50 <minimal> nickersonm: 1 thing in that article that isn't clear/doesn't make sense to me is regarding "--loader "\alpine.efi" the location of the Unified Kernel Image, in relation to the EFI partition, since we used /boot/alpine.efi then we set it to \alpine.efi"
2021-07-22 19:07:50 <minimal> as I'd expect that the contents of the ESP partition with be a "EFI" dir with at least a "boot" subdir of that - so unless EFI/boot is mounted onto /boot then /boot/alpine.efi would not appear as EFI/boot/alpine.efi in the ESP partition
2021-07-22 19:08:41 <nickersonm> minimal: looks like they mounted sda1 as /boot, and generated the kernel image on /boot/alpine.efi
2021-07-22 19:08:54 <invoked> imo easier to just boot off usb on a keyring than mess with any of that stuff, unless you're paranoid about usb and put nail polish on all the ports. but you'd still need to know (somehow) that your bios wasn't tampered with.
2021-07-22 19:09:11 <invoked> which is close to impossible
2021-07-22 19:09:29 <minimal> nickersonm: right, but to be UEFI-compliant sda1 would need to contain "EFI/boot" directory hierarchy
2021-07-22 19:10:24 <nickersonm> minimal: No need for an EFI subdirectory explicitly afaik?  They're essentially just passing "/dev/sda1/alpine.efi" as the EFI boot option.  Perhaps it's noncompliant with the spec, but works with most implementations?
2021-07-22 19:10:29 <minimal> invoked: bios? I thought we were talking about UEFI
2021-07-22 19:11:33 <minimal> nickersonm: it certainly wouldn't be interchangable with any other bootloader. Not sure about the spec - think I looked at it once a couple of years ago and almost lost the will to live ;-)
2021-07-22 19:12:54 <minimal> nickersonm: I've even seen some discussion of uppercase/lowercase/case-independant issues with the naming of "EFI/BOOT" vs "EFI/boot" paths in some vendors UEFI implementations
2021-07-22 19:13:22 <nickersonm> minimal: Hah.  I've not dared that.  Re. other bootloaders, yeah - the author cleaned up everything else and only has the unified kernel image
2021-07-22 19:14:56 <minimal> nickersonm: I haven't (yet) tried creating ESP partitions smaller than 128MiB as apparently that's the smallest the standard permits
2021-07-22 19:18:11 <nickersonm> minimal: I didn't realize that.  I use 100MB for my EFI VMs, from following standard BIOS partitioning and transitioning to EFI instead
2021-07-22 19:20:21 <minimal> nickersonm: actually from a quick search it seems that 100MiB is a Microsoft recommendation for Windows, not sure where I got 128MiB from - again I guess to be sure of what the min is involves looking at the spec :-(
2021-07-22 19:22:09 <nickersonm> In any case, my EFI VM has 288k actual usage on /boot/efi and 20.8MB on /boot (not a separate partition), so any MB seems like more than enough
2021-07-22 19:30:41 <minimal> nickersonm, yeah I wanted to have the ESP partition, um, minimal in size :-) Have a brief look at UEFI 2.5 spec and it doesn't define the layout of the ESP filesystem in much detail: "For more information about the EFI file system and file image format, visit the web site from which
2021-07-22 19:30:41 <minimal> this document was obtained". Doh!
2021-07-22 19:32:54 <nickersonm> minimal: hah.  Depending on your bootloader, you might be able to get away with a 1MB EFI partition - it sounds like UEFI is ill-defined or at least ill-implemented enough that the best test is to try it and see if it works.
2021-07-22 19:34:22 <minimal> nickersonm: more details relevant to the dir structure: "An EFI system partition that is present on a hard disk must contain an EFI defined directory in the
2021-07-22 19:34:22 <minimal> required to be stored in any specific location in the EFI system partition."
2021-07-22 19:34:22 <minimal> subdirectories below EFI. Applications that are loaded by other applications or drivers are not
2021-07-22 19:34:22 <minimal> root directory. This directory is named EFI. All OS loaders and applications will be stored in
2021-07-22 19:35:20 <minimal> so a "EFI" top-level directory is required (for a hard disk) with bootloader inside that but seems that files loaded by a bootloader don't themselves have to be inside "EFI"
2021-07-22 19:35:52 <nickersonm> minimal: Thanks for looking it up; sounds like the article from earlier should have used /EFI/alpine.efi
2021-07-22 19:38:16 <minimal> nickersonm: I'm interested in getting my understanding of this straight as I had Syslinux booting via UEFI in the recent past with a ESP partition and a /boot partition where I had to copy the kernel/initramfs across to the ESP (as Syslinux UEFI only supports loading files in the same directory as itself). I then decided to instead have just a ESP partition and to bind-mount /efi/boot/ as "/boot" to avoid the need to copy kernel/initramfs 
2021-07-22 19:38:16 <minimal> across after every update but its not currently working...
2021-07-22 19:41:45 <nickersonm> minimal: yeah, I don't see why the ESP can't just be mounted as /boot; I've seen that in some guides.
2021-07-22 19:44:18 <minimal> nickersonm: its a Syslinux-specific issue - Syslinux's own files need to be in EFI/boot/ in the ESP partition and (due to Syslinux) kernel/initramfs need to be in the same directory - so EFI/boot on partition needs to be (bind)mounted at /boot for Alpine to place/update kernel/initramfs files
2021-07-22 19:44:54 <minimal> if Grub or something else is used its not an issue
2021-07-22 19:46:16 <nickersonm> minimal: ah, I see.  I wasn't able to get EFI working with syslinux after a bit of effort, likely related to that.  grub worked just fine, so I didn't bother investigating in detail.
2021-07-22 19:47:31 <minimal> yeah I had it working in my script (with seperate ESP and /boot parts) and then changed it to use the simgle ESP plus bind-mount menthod but haven't had the time to track down what exactly broke with that change
2021-07-22 19:48:19 <minimal> There's no install script for Syslinux UEFI - you have to copy the right files across manually
2021-07-22 19:50:59 <nickersonm> Yeah, I was following https://wiki.alpinelinux.org/wiki/Bootloaders#EFI and https://wiki.archlinux.org/title/syslinux#UEFI_Systems
2021-07-22 19:55:42 <minimal> nickersonm: me too :-) In the Arch box in the "UEFI Systems" the 2nd bullet point is both right (1st sentence) and sort-of-wrong/misleading (2nd sentence) as if you mount ESP part as /boot/ then you'd have to copy kernel/initramfs from /boot to /boot/EFI/syslinux/ everytime they change (hence my bind-mount idea)
2021-07-22 19:57:10 <minimal> s/box/article/
2021-07-22 19:57:11 <alpine-meetbot> minimal meant to say: nickersonm: me too :-) In the Arch article in the "UEFI Systems" the 2nd bullet point is both right (1st sentence) and sort-of-wrong/misleading (2nd sentence) as if you mount ESP part as /boot/ then you'd have to copy kernel/initramfs from /boot to /boot/EFI/syslinux/ everytime they change (hence my bind-mount idea)
2021-07-22 20:06:11 <skintimidation> I'm pretty certain I don't need a unified kernel image for my setup, luckily.
2021-07-22 20:08:57 <skintimidation> It's really crazy how much this is revealing to me that the only method of doing anything with machines anymore is patching together a series of guides. Any search I come up with is just inundated with guides that do x, when all I want to do is y.
2021-07-22 20:08:59 <skintimidation> It would be so nice if there was some more central resource that contained the variations that each distro did for how linux is supposed to work, since I'm pretty sure linux IS supposed to work (like that). lol
2021-07-22 20:13:58 <invoked> that's what linux fram scratch is for
2021-07-22 20:14:47 <skintimidation> Do UUIDs refer to drives or partitions?
2021-07-22 20:14:50 <skintimidation> Oh, there are PARTUUIDs
2021-07-22 20:28:51 <skintimidation> minimal: so would this script you're working on let you LUKS encrypt your own colocated server, or something similar?
2021-07-22 20:29:17 <skintimidation> How does the key get safely delivered? Is it just through the fact that ssh is secure, or is there another layer I'm not sure of?
2021-07-22 20:34:31 <Hello71> you need to authenticate the host key
2021-07-22 20:34:49 <Hello71> in that case it's roughly as secure as entering it at the console
2021-07-22 20:34:50 <skintimidation> Which device is the host in that case?
2021-07-22 20:34:54 <Hello71> both subject to evil maid attacks
2021-07-22 20:35:38 <skintimidation> How would the evil maid get the deets when you're remoting in? You're saying they'd still be able to set up a new firmware that reads your authkey, I suppose?
2021-07-22 20:36:29 <skintimidation> So you'd just have to also secure-boot your image in that case, right? One thing I don't get is how secure boot can't just be overwritten by the evil maid--isn't it in the firmware of the booting device? Or is it verified on the encrypted /root?
2021-07-22 20:36:39 <skintimidation> er, encrypted root. encrypted /. those are the terms i meant.
2021-07-22 20:39:42 <Hello71> aiui you can store key (or partial key) on tpm
2021-07-22 20:40:22 <Hello71> in theory if evil maid comes to reset your keys, the tpm will also reset, which you can detect from os
2021-07-22 20:40:29 <Hello71> although i'm not sure exactly now
2021-07-22 20:40:30 <Hello71> how
2021-07-22 20:42:43 <skintimidation> very interesting; i was thinking maybe they'd replicate your setup on a new board that doesn't have those things, or something like that
2021-07-22 20:43:07 <skintimidation> i better start checking my keyboard for hardware keyloggers more often, lmao
2021-07-22 20:43:19 <skintimidation> I really hope I can cobble this together today. I've been without a main rig for like a week.
2021-07-22 20:51:37 <skintimidation> OH, was thinking about this in the shower: I don't know at all what's going on here: https://github.com/itoffshore/alpine-linux-scripts/blob/master/setup-disk#L284
2021-07-22 20:52:31 <skintimidation> does this 1) package up the current livedisk 2) put it in /tmp/ovlfiles and then 3) later unpack the live system into the new root?
2021-07-22 20:53:33 <skintimidation> because i only see it do that ONE time, and i don't know where you'd later use the ovlfiles
2021-07-22 20:54:13 <skintimidation> 		$repoflags $pkgs <$ovlfiles
2021-07-22 20:54:13 <skintimidation> oh, from here? apk add --root "$mnt" $apkflags --overlay-from-stdin \
2021-07-22 21:02:29 <minimal> skintimidation: You're talking about the remote unlocking? For a "normal" machine you need to be at the console/have access to the console to unlock a LUKS partition at boot time, with the remote unlock feature when a machine boots it then during the early start of the initramfs it pauses booting whilst waiting for an SSH connection to be used to give the LUKS passphrase. So it could be used for a headless server (e.g. with no screen or 
2021-07-22 21:02:29 <minimal> keyboard) or could be used with a VM/Cloud server where you have no console access. The ssh connection used SSH-keys only (the initramfs has an authorized_keys file that you configure)
2021-07-22 21:03:11 <skintimidation> kinda fun.
2021-07-22 21:03:32 <skintimidation> good work on that script; wish i wasn't going a completely different direction, lmao
2021-07-22 21:03:59 <skintimidation> i'll keep it in mind for future installs outside of me house, though, because that's exactly the kind of thing i wouldn't think about that i should more often
2021-07-22 21:05:49 <minimal> Hello71: a TPM is like a HSM (Hardware Security Module) but without all the physical protection. Its more akin to a credit card, debit card, SIM card where its "hard" (but not impossible) to extract keys from it
2021-07-22 21:06:34 <onelegend> awww shit that thing
2021-07-22 21:07:05 <onelegend> you're not supposed to store your private keys in there
2021-07-22 21:07:11 <onelegend> well you are
2021-07-22 21:07:14 <skintimidation> I'm starting to think I might not even use btrfs
2021-07-22 21:07:21 <onelegend> but you shouldn't store your disk encryption keys in the TPM
2021-07-22 21:07:22 <minimal> skintimidation: I see notice someone else has combined remote unlock with wireguard VPN for ssh-over-VPN unlocking
2021-07-22 21:07:45 <skintimidation> but if i don't choose btrfs from the start, conversion later might not work--I guess snapshotting can be useful for what I wanted, where I could fuck around and then undo stuff
2021-07-22 21:08:30 <minimal> onelegend: On Windows I believe Bitlocker disk encryption relies on the TPM.
2021-07-22 21:08:45 <onelegend> winblowz
2021-07-22 21:09:00 <skintimidation> it does indeed rely on that
2021-07-22 21:10:32 <minimal> onelegend: using a TPM is just a way to try and "lock" a disk to a specific machine so if someone steal the physical disk, rather than the whole computer, then they can't decrypt it
2021-07-22 21:10:56 <invoked> tpm is useful for knowing if something that shouldn't have changed has changed
2021-07-22 21:11:19 <onelegend> you can use it for secure boot
2021-07-22 21:11:43 <minimal> yeah, I'm just talking about it in the context of disk encryption
2021-07-22 21:12:07 <onelegend> minimal: would you use a TPM to store the key?
2021-07-22 21:13:02 <minimal> haven't used a TPM but I believe you could use it as part of the key derivation
2021-07-22 21:13:51 <invoked> look up matthew garrett's talks on tpm (@mjg59 on twitter)
2021-07-22 21:14:11 <onelegend> no thanks lol. I
2021-07-22 21:14:25 <onelegend> I appreciate the help tho invoked
2021-07-22 21:14:41 <minimal> invoked: yes I've read some of his articles in the past
2021-07-22 21:15:25 <minimal> https://github.com/shpedoikal/tpm-luks
2021-07-22 21:15:30 <onelegend> nice repo
2021-07-22 21:16:11 <minimal> its an example of how to store a LUKS key on a TPM
2021-07-22 21:16:56 <onelegend> you can write to the TPM?
2021-07-22 21:17:26 <minimal> onelegend: did you look at the README?
2021-07-22 21:18:12 <c705> onelegend: can i help you?
2021-07-22 21:18:30 <onelegend> c705: you're the same boi from libera right?
2021-07-22 21:18:39 <c705> sure
2021-07-22 21:21:11 <onelegend> invoked: https://lca2020.linux.org.au/schedule/presentation/64/ is it this one?
2021-07-22 21:21:11 <skintimidation> oh yeah? sign this message:
2021-07-22 21:21:16 <skintimidation> (kidding)
2021-07-22 21:21:57 <c706> YOOOO
2021-07-22 21:22:05 <invoked> i dunno... i'm not on a machine that i can look at that link. but he's been doing tpm and other security topics for a long time.
2021-07-22 21:22:20 <invoked> power management also.
2021-07-22 21:22:25 <onelegend> thanks
2021-07-22 21:27:45 <[[OneLegend]]> ok
2021-07-22 21:27:55 <[[OneLegend]]> the group function works differently in oftc
2021-07-22 21:28:59 <skintimidation> where does alpine usually put the kernel and stuff when i first install? is it provided from the apkoverlay? just curious so when i copy it and do the efibootmgr stuff i'm not doing anything goofy
2021-07-22 21:35:38 <skintimidation> Oh, this might be a useful question, too: I see in setup-disk that grsec is changed for max chroots, but then they don't chroot at all--they just mount /proc and /dev
2021-07-22 21:35:44 <skintimidation> so is that necessary?
2021-07-22 21:36:06 <OneLegend> is grsec still in Alpine?
2021-07-22 21:36:46 <skintimidation> that's something else that i figured wasn't actually true
2021-07-22 21:42:41 <Hello71> [[OneLegend]]: no, and please cut that out. if you want to group a bunch of shit, open another client
2021-07-22 21:42:57 <[[OneLegend]]> ?
2021-07-22 21:43:09 <[[OneLegend]]> umm no thanks
2021-07-22 21:46:12 <skintimidation> The biggest thing I don't get now is how you can setup grub BEFORE you install all the packages with apk--how do they get the kernel and stuff? Is it just the one already on the livedisk?
2021-07-22 21:46:57 <[[OneLegend]]> the livedisk
2021-07-22 21:47:42 <[[OneLegend]]> you're operating from a livedisk anyways, so that shouldn't be a problem
2021-07-22 21:47:47 <skintimidation> okay, so I'll add that to my script
2021-07-22 21:47:50 <skintimidation> things are COMING TOGETHER, BOYS.
2021-07-22 21:47:58 <[[OneLegend]]> :)
2021-07-22 21:49:37 <skintimidation> i think the part i don't understand is if efibootmgr's gonna do the copying for me, or if it just does the config and i need to do the copying--i'll probably just do both, lmao
2021-07-22 22:37:33 <skintimidation> Nope, I'm convinced i do the copying
2021-07-23 01:52:01 <geoman> how do you install opengl? i tried using apk list but didn't find anything
2021-07-23 01:56:16 <geoman> i figure i'll just use glfw
2021-07-23 01:56:51 <geoman> that could have been solved with a wiki search
2021-07-23 01:56:55 <raven523> I think you want mesa
2021-07-23 01:57:26 <raven523> specifically mesa-gl
2021-07-23 02:01:38 <geoman> i was trying to compile CSE2, which allows you to use GLFW3 or SDL2 as a backend platform for rendering
2021-07-23 02:01:52 <drgibbon> hello, just trying to install Alpine on a Raspberry Pi, what is the correct way to go with the disk selection part?
2021-07-23 02:02:20 <drgibbon> (the docs here https://wiki.alpinelinux.org/wiki/Raspberry_Pi don't cover setup-alpine at all, just says to run it).
2021-07-23 02:04:46 <drgibbon> oh, well I don't know what I did before but it's working now.
2021-07-23 02:04:50 <geoman> k
2021-07-23 02:05:07 <drgibbon> no question then :)
2021-07-23 02:05:21 <geoman> setup-alpine does pretty much everything and gives you decent control over everything except disk partitioning
2021-07-23 02:05:48 <geoman> i guess that's what setup-disk is for
2021-07-23 02:23:41 <drgibbon> diskless mode seems nice on Raspberry Pi, but how hard would it be to set up a persistent /home ?
2021-07-23 02:54:47 <LACampbell> I wish raspberry PI had an SSD bay or something. I know people will say that defeats the purpose but I don't care. I want a tiny minimal linux computer :P
2021-07-23 03:03:57 <Hello71> LACampbell: buy a different computer?
2021-07-23 03:11:31 <LACampbell> suggestions? :p
2021-07-23 03:12:03 <LACampbell> small cheap desktop computer with SSD that we know linux will run well on. I dunno how many there are
2021-07-23 03:13:58 <Sheila> if you can, try to find an AMD-based version of Intel's compute sticks.
2021-07-23 03:15:15 <Sheila> I used one for working on Adélie for about a year, then Meltdown happened.
2021-07-23 03:23:51 <LACampbell> those little things have proper SSDs in them!?
2021-07-23 03:24:03 <Sheila> no, just eMMC
2021-07-23 03:24:41 <Sheila> I'm sure you could find a somewhat bulkier version of the concept that has an m.2 slot, though.
2021-07-23 03:30:07 <LACampbell> I'm actually not too clued up on the big difference between eMMC and 'proper' SSDs
2021-07-23 03:32:13 <Sheila> well. SD cards are just MMC in a proprietary package, so eMMC is almost invariably a soldered-on SD card.
2021-07-23 03:45:32 <LACampbell> ah
2021-07-23 03:45:39 <LACampbell> that clarifies it
2021-07-23 04:07:12 <skintimidation> Does btrfs save all of its volume stuff inside the filesystem itself?
2021-07-23 04:07:19 <skintimidation> How does it all get loaded up at boot?
2021-07-23 04:20:21 <skintimidation> hmm; is gptfdisk not in the repos?
2021-07-23 04:20:41 <skintimidation> ah, looks like it's in edge? interesting
2021-07-23 04:43:40 <skintimidation> If I move to edge, though, doesn't that change which kernel I'm running and everything else? Can I change to edge and run apk update/upgrade to get that moved over without fucking up my install?
2021-07-23 04:45:22 <ikke> no, you should not combine different versions
2021-07-23 04:46:14 <ikke> gptfdisk is in stable releases as well
2021-07-23 04:46:18 <ikke> https://pkgs.alpinelinux.org/packages?name=gptfdisk&branch=v3.14
2021-07-23 04:59:14 <skintimidation> ikke: running "apk add gptdisk" says there's no such packge
2021-07-23 04:59:24 <skintimidation> package, even.
2021-07-23 04:59:40 <ikke> you mentioned gptfdisk, not gptdisk
2021-07-23 05:23:46 <skintimidation> my bad; gptfdisk*
2021-07-23 05:29:04 <ikke> what repos do you use?
2021-07-23 05:35:44 <skintimidation> ikke: just the livedisk
2021-07-23 05:36:25 <ikke> ah, so I guess a standard image without network connection?
2021-07-23 05:39:35 <skintimidation> ikke: it's got net, luckily
2021-07-23 05:39:49 <skintimidation> oh wtf, now it doesn't
2021-07-23 05:40:30 <ikke> Make sure there is an online repository enabled in /etc/apk/repositories
2021-07-23 05:41:22 <skintimidation> yeah lemme run setup-alpine and then not run setup-disk
2021-07-23 05:41:31 <skintimidation> i can avoid that just through saying "none" at the disk prompt, right?
2021-07-23 05:41:42 <ikke> yes
2021-07-23 05:42:34 <skintimidation> will do; glad you're online, ikke. did you write the setup scripts?
2021-07-23 05:42:42 <ikke> no, I did not
2021-07-23 05:50:07 <skintimidation> ah, damn. 
2021-07-23 05:50:40 <skintimidation> but you're so familiar with it by now i assume that you're able to tell me if my script is gonna asplode my computer
2021-07-23 05:51:01 <skintimidation> do you use btrfs?
2021-07-23 05:51:04 <ikke> nope
2021-07-23 05:54:19 <skintimidation> shiiii.
2021-07-23 05:54:34 <skintimidation> well, that shouldn't matter too much if i go from there
2021-07-23 05:54:41 <skintimidation> are you still using grub/syslinux and bios boot?
2021-07-23 05:54:59 <skintimidation> i can't remember if you use alpine at all on desktop
2021-07-23 05:55:34 <ikke> Not on my desktops atm, I do have alpine vms
2021-07-23 05:58:57 <skintimidation> ah, interesting.
2021-07-23 05:58:59 <skintimidation> what's the stuff about "where to store configs"?
2021-07-23 06:00:01 <skintimidation> weird; i don't see that in my copy of setup-disks
2021-07-23 06:01:00 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-lbu.in
2021-07-23 06:03:49 <skintimidation> ah, i likely don't need that for a sys install?
2021-07-23 06:03:53 <skintimidation> son of a bitch, apk cache directory mount was a big fail
2021-07-23 06:03:56 <skintimidation> resource busy
2021-07-23 06:04:24 <ikke> no, you only need it on a run-from-ram system
2021-07-23 06:04:53 <ikke> the apk cache directory is optional
2021-07-23 06:06:41 <skintimidation> ah, schweet. good to know.
2021-07-23 06:07:16 <ikke> and you can always set it up later
2021-07-23 06:07:29 <ikke> it's just a symlink from /etc/apk/cache to the location where you want to store it
2021-07-23 06:11:52 <skintimidation> ah, hell yeah.
2021-07-23 06:12:19 <skintimidation> what does this line in setup-disk do? lbu package - | tar -C /mnt -zxv > "$ovlfiles"
2021-07-23 12:10:15 <Hello71> magic
2021-07-23 13:37:45 <skintimidation> man tar hand hook tar door
2021-07-23 13:38:22 <skintimidation> oh wait, okay
2021-07-23 13:38:41 <skintimidation> actually no, i still don't know the last part
2021-07-23 13:41:34 <skintimidation> if $ovlfiles is an empty string, does > put its value in the variable somehow, or does that get expanded first? if you pass an empty string to the file destination for >, wtf happens
2021-07-23 13:42:53 <skintimidation> wat is this "nonexistent directory" shit
2021-07-23 13:52:43 <lopid> if ovlfiles is empty, you will get an error when writing to such a filename
2021-07-23 13:57:26 <skintimidation> yep, just realized from the script it's set right before
2021-07-23 13:57:28 <skintimidation> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-disk.in#L503
2021-07-23 13:57:35 <skintimidation> i'm just "can't read city" or something
2021-07-23 13:57:42 <skintimidation> which means i'll brb after a nap
2021-07-23 16:33:02 <lopid> am i the only one getting warnings when using git-bash-completion? one day i'll look into it
2021-07-23 19:34:50 <thepian> I'm trying to create bootable media with RaspberryPi. I boot on USB and create bootable media on SD Card. Should that work? I get Resource busy for /dev/sda1 when I run setup-bootable
2021-07-23 19:36:27 <programmerq> are you booting from usb with the sdcard slot unpopulated? I believe some pis can do that if you first boot from sd card and then configure it to do so.
2021-07-23 19:36:54 <programmerq> a workaround might be to boot, and then plug an sd card -> usb adapter into a usb port and then work with the card on that block device
2021-07-23 19:38:13 <thepian> Thank you. Will try
2021-07-23 19:50:15 <greatwielder[m]> does anyone know if it's possible to get v9fs (plan 9 fs over network) to work properly as a /home? some applications misbehave because they fail to create special files such as pipes at their desired locations in /home.
2021-07-23 19:59:09 <invoked> greatwielder[m]: i keep all my 9 stuff separate, but if you're feeling resilient, you can ask in #cat-v
2021-07-23 19:59:32 <invoked> someone there might be using it that way.
2021-07-23 20:01:46 <invoked> or email 9fans, i guess...
2021-07-23 20:02:18 <invoked> 
ACTION is not advocating anything

2021-07-24 01:10:00 <ajin> my vps only has the standard iso; is it possible to "crossgrade" to virtual? (or might I not want to?)
2021-07-24 01:15:01 <Skaface> i think its just a more stripped down kernel
2021-07-24 01:15:13 <Skaface> if it works fine i wouldn't bother
2021-07-24 01:22:52 <ajin> oh, it was as easy as apk add linux-virt and apk del linux-vanilla
2021-07-24 01:26:52 <ajin> my next question was going to be if I could jump multiple versions, but that seems to have worked 
2021-07-24 04:11:01 <LACampbell> ajin: which VPS offers alpine?
2021-07-24 04:18:12 <fluix> at least Linode does
2021-07-24 04:21:00 <LACampbell> nice
2021-07-24 10:11:39 <donoban> anyone here using tinydm?
2021-07-24 10:12:15 <ikke> scaleway does as well
2021-07-24 10:12:22 <ikke> LACampbell: ^
2021-07-24 10:13:39 <LACampbell> thinking more, many VPSes allow you to just upload an ISO.
2021-07-24 10:14:21 <ikke> yes, or you can even install it over an existing OS
2021-07-24 10:16:22 <PureTryOut1> donoban: all UI's on postmarketOS use it
2021-07-24 10:17:46 <donoban> PureTryOut1: I installed it but it seems that fails to start without any error
2021-07-24 10:18:52 <donoban> when it works, does it have some kind of password prompt? Maybe I'm better using autologin, since I have full disk encription I think that it would be fine to get a desktop directly on startup
2021-07-24 10:19:54 <PureTryOut1> It does not no, the point is to auto login
2021-07-24 10:20:14 <PureTryOut1> Did you set a graphical session it should start?
2021-07-24 10:20:44 <donoban> no, I check the init file and /etc/conf, I didn't find where specify it
2021-07-24 10:20:52 <PureTryOut1> We do `tinydm-set-session -s /usr/share/wayland-sessions/plasma-mobile.desktop` for Plasma Mobile for example
2021-07-24 10:21:07 <PureTryOut1> If you don't use `tinydm-set-session` once then tinydm won't know what graphical session to start
2021-07-24 10:23:07 <donoban> ok thanks PureTryOut1 , let me try
2021-07-24 10:23:31 <donoban> it would be nice to add some man-doc
2021-07-24 10:24:31 <donoban> tinydm: session set: /usr/share/wayland-sessions/sway.desktop
2021-07-24 10:24:36 <donoban> let's check next reboot :)
2021-07-24 10:29:55 <PureTryOut1> PR's for docs welcome 😉
2021-07-24 10:33:01 <donoban> yeah I would like to help, is this a good starting point? https://gitlab.com/postmarketOS/pmaports/-/issues/823 (skipping ligthdm comments)
2021-07-24 10:34:39 <PureTryOut1> Uh that issue isn't relevant anymore now we have tinydm
2021-07-24 10:34:56 <PureTryOut1> https://gitlab.com/postmarketOS/tinydm/ is now the place to be
2021-07-24 10:35:25 <PureTryOut1> I'm off for now, if you have more questions you can ask them in #devel:postmarketos.org (or #postmarketos-devel on OFTC)
2021-07-24 10:35:37 <donoban> it linked to the issue More information:
2021-07-24 10:35:39 <donoban> pmaports#823
2021-07-24 10:35:45 <donoban> lol
2021-07-24 10:36:38 <donoban> I will try to write a wiki page which is more eaiser to check and modify, and when it's ok try to create a tinydm-doc package
2021-07-24 10:36:57 <donoban> pmOS has his own wiki?
2021-07-24 10:39:22 <donoban> well, really it doesn't need too much info
2021-07-24 10:39:54 <donoban> just that you need to call tinydm-set-session
2021-07-24 11:14:09 <DItronics> hi
2021-07-24 11:14:37 <DItronics> any chance to build driver for rtl8814au?
2021-07-24 11:14:48 <ikke> Is that an in-tree driver?
2021-07-24 11:15:10 <DItronics> don't think so
2021-07-24 11:15:39 <DItronics> i'm trying to build it but not lucky
2021-07-24 11:15:47 <DItronics> 3.14
2021-07-24 11:18:35 <DItronics> make ARCH=x86_64 CROSS_COMPILE= -C /lib/modules/5.10.43-0-lts/build M=/srv/service1/drivers/rtl8812au  modules
2021-07-24 11:19:01 <DItronics> make[1]: *** /lib/modules/5.10.43-0-lts/build: No such file or directory.  Stop.
2021-07-24 11:19:42 <DItronics> when I try to install linux-lts-dev I've got an error
2021-07-24 11:21:25 <ikke> what error?
2021-07-24 11:22:02 <DItronics> ERROR: Failed to create lib/modules/5.10.52-0-lts/build: No such file or directory
2021-07-24 11:22:02 <DItronics> ERROR: linux-lts-dev-5.10.52-r0: failed to rename lib/modules/5.10.52-0-lts/.apk.7799b9831755df9b8d2bd7bcfbd82d8dc23a1898c8802be2 to lib/modules/5.10.52-0-lts/build.
2021-07-24 12:43:38 <donoban> PureTryOut1: thanks! it works :)
2021-07-24 14:50:19 <GreyDev> does anybody know how to enable predictable interface names over the traditional ethX naming scheme?
2021-07-24 14:50:43 <GreyDev> I'm not sure where to start looking - is this part of the kernel config? or is it something in userspace?
2021-07-24 14:52:04 <ikke> userspace
2021-07-24 14:59:04 <trylletre> afaik its udev that does it in other distros with a 80-net-name-slot.rules
2021-07-24 16:12:49 <minimal> GreyDev: predictable names should be enabled by default. Adding "net.ifnames=0" to kernel cmdline (i.e. via your bootloader's config) will *disable* predictable names and use the old ethX format
2021-07-24 16:31:05 <Hello71> GreyDev: apk add eudev-netifnames
2021-07-24 16:31:18 <Hello71> also setup-udev etc
2021-07-24 16:38:12 <GreyDev> thanks for your help both
2021-07-24 16:38:26 <GreyDev> haven't got it working yet but it's a start
2021-07-24 16:38:58 <Hello71> did you setup-udev
2021-07-24 16:39:29 <GreyDev> where might I find that script? I don't see it in /sbin/ with the other setup-* scripts
2021-07-24 16:40:36 <GreyDev> this is a fresh install (well... run from RAM) of a the regular iso
2021-07-24 16:41:08 <GreyDev> so clean boot and add eudev-netifnames, and that is where I am
2021-07-24 16:41:43 <GreyDev> AH never mind
2021-07-24 16:41:47 <GreyDev> just had to add eudev as well
2021-07-24 16:45:23 <mallory> Hello. Some files in the GitLab seem to consistently give error 500 when trying to view them, e.g. https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/radicale/radicale.pre-install and https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/radicale/radicale.initd -- I've also had this before with other files.
2021-07-24 16:48:22 <ikke> Seems to run into a timeout
2021-07-24 16:50:11 <Sheila> in the meantime, use e.g. https://git.alpinelinux.org/aports/tree/community/radicale/radicale.pre-install
2021-07-24 16:50:18 <ikke> Gitlab::Git::CommandTimedOut
2021-07-24 16:59:11 <minimal> GreyDev: seems mdev doesn't handle persistent names "out-of-the-box" but you could add a rule to /etc/mdev.conf to call nameif and then create a /etc/mactab to map MAC addresses to the names you want
2021-07-24 16:59:27 <minimal> that's if you don't want to use udev/eudev
2021-07-24 17:22:27 <mallory> Sheila: thanks, didn't know there's another web frontend
2021-07-24 17:58:50 <ikke> Yeah, we keep it because file browsing in other front-ends is not really performant, or even limits the amount of entries
2021-07-24 18:29:52 <PureTryOut1> donoban: no problem, have fun with it!
2021-07-24 19:27:52 <Guest2066> hi all! I'm new to alpine, and wondering what the best way to get security updates is - would it be https://lists.alpinelinux.org/lists/~alpine? In other words, what would be the best way of finding out things like the kernel patch status for CVE-2021-33909? Thank u !!!
2021-07-24 19:28:28 <ikke> Well, the kernel is kind of an exception
2021-07-24 19:29:03 <invoked> https://lists.alpinelinux.org/lists/~alpine
2021-07-24 19:29:15 <invoked> see: security-announcements
2021-07-24 19:29:24 <invoked> not very active though
2021-07-24 19:29:25 <ikke> Yes, but we do not make a lot of anouncements there
2021-07-24 19:29:40 <ikke> In general, here you can view the patch status of packages: https://security.alpinelinux.org/
2021-07-24 19:29:52 <ikke> But we do not track CVEs for the kernel
2021-07-24 19:33:09 <invoked> secfixes-tracker will eventually post to security-announcements, right
2021-07-24 19:33:45 <invoked> eventually(tm)
2021-07-24 19:33:48 <ikke> I think the goal was to create issues in gitlab
2021-07-24 19:35:40 <Guest2066> That's all very helpful! Really appreciate all the info, and keep up the excellent work
2021-07-24 19:36:20 <invoked> i'm guessing that most people who would sub to security-announcements is already getting oss-security (et al) so on some level it's just duplicative
2021-07-24 19:36:53 <invoked> i don't use arch much anymore, but i still get a blast of ASAs every couple of weeks which are completely redundant.
2021-07-24 21:14:59 <jeffreybenzos[m]> i totally fooking thought that i was on the oftc one.
2021-07-24 21:21:51 <jeffreybenzos[m]> How's everyone been?
2021-07-24 21:25:40 <ikke> Doing alright
2021-07-24 21:26:23 <jeffreybenzos[m]> hell yeah
2021-07-24 21:26:42 <jeffreybenzos[m]> I got everything working so far, ikke; had to chroot to fix forgetting my root passwd because it had been a few days
2021-07-24 21:26:48 <jeffreybenzos[m]> but getting my stuff going with efibootmgr
2021-07-24 21:27:30 <jeffreybenzos[m]> WITHOUT first installing grub
2021-07-24 21:27:33 <jeffreybenzos[m]> and it worked, by golly
2021-07-24 21:28:01 <jeffreybenzos[m]> Dunno if I can generalize the system setup enough to help you guys out, though, by contributing back to the build scripts
2021-07-24 21:28:09 <jeffreybenzos[m]> because it was, overall, pretty simple.
2021-07-24 21:43:51 <jeffreybenzos[m]> weird unexpected behavior: after installing, only one of my monitors shows anything (the other doesn't mirror), and it's only like 1/10th of the screen space
2021-07-24 21:44:03 <jeffreybenzos[m]> oh shit let's see if it mounted boot
2021-07-24 21:56:48 <jeffreybenzos[m]> Interesting. Does the "upgrade to edge" guide on the wiki recommend rebooting? I needed to reboot (or maybe do something else) before running mkinitfs again
2021-07-24 22:27:22 <Skaface> so im battling with a problem... I have a PXE network boot working, but there doesnt seem to be a way for it to install programs from the apk cache, rather than a repository
2021-07-24 22:46:54 <jeffreybenzos[m]> i'm totally unsure how to fix that, skaface, but i feel for you
2021-07-24 22:47:07 <jeffreybenzos[m]> always crazy when there's a solution that you feel like should be simple, but can't quite hack
2021-07-24 23:02:56 <Skaface> the only option i can think of is somehow creating a local mirror of only the packages i have installed
2021-07-24 23:03:17 <Skaface> like a apk cache sync but make an actual package mirror
2021-07-24 23:04:09 <Skaface> im not sure if thats easily accomplished though
2021-07-24 23:08:45 <jeffreybenzos[m]> also, skaface, if you setup-apkcache and then reboot, shouldn't it be installing those programs directly from the cache on boot?
2021-07-24 23:09:42 <Skaface> yeah it should, but i think the initramfs only auto mounts usb type media, it wont automount and use a nfs mount
2021-07-24 23:18:47 <jeffreybenzos[m]> oh, is that where the cache would be, skaface?
2021-07-24 23:20:28 <jeffreybenzos[m]> who was that cool guy whose name started with an m-- was it mla?
2021-07-24 23:20:46 <jeffreybenzos[m]> there's one dude who was a great guy and helped me set up my laptop ages ago; i just can't remember what's necessary to get x to start bspwm, etc etc
2021-07-24 23:21:33 <Skaface> yes, ive been working around the problem by including the local cache directory in the apkovl file, but that would keep the cache in ram unless i clear it
2021-07-24 23:21:58 <Skaface> i guess i could make it empty the cache in a startup script
2021-07-24 23:22:05 <Skaface> that will work
2021-07-24 23:24:50 <jeffreybenzos[m]> i believe in you.
2021-07-24 23:30:36 <Hello71> have you looked at how the standard media installs from local repo in the first place
2021-07-24 23:32:15 <jeffreybenzos[m]> hello71, wassaaaaap
2021-07-24 23:32:20 <Hello71> pls no
2021-07-24 23:32:41 <jeffreybenzos[m]> lmao
2021-07-24 23:32:45 <jeffreybenzos[m]> no, i got it working; didn't you see?
2021-07-24 23:32:53 <jeffreybenzos[m]> the only thing i need to do now is figure out x
2021-07-24 23:33:05 <jeffreybenzos[m]> which means just mounting and copying my dotfiles i'm p sure
2021-07-24 23:39:05 <jeffreybenzos[m]> Do you guys use doas or sudo?
2021-07-24 23:45:22 <asdflkj> doas, sudo's bloat makes it too insecure
2021-07-24 23:47:31 <asdflkj> but it would be nicer if apk had separate commands (not subcommands) for searching, etc. so some they could be nopass and apk del &c not
2021-07-24 23:48:36 <invoked> it depends. i'm not advocate for either one, but doas does not cover all the sudo use cases.
2021-07-24 23:51:48 <ajin> LACampbell: virmach.com  I just wanted someone who hosted out of the data center right by my house. it's a lot cheaper than $60/month for 1U
2021-07-24 23:52:51 <ajin> (the question was what vps offers alpine)
2021-07-24 23:53:10 <asdflkj> that must have awesomely low lag
2021-07-24 23:53:35 <asdflkj> lovely low lag*
2021-07-24 23:53:58 <ajin> yahoo pings less than 2ms. they must be in the building
2021-07-24 23:56:01 <LACampbell> I'm in NZ, options are limited, so my server is about 4 hours away by plane :) it's fine with mosh though
2021-07-24 23:56:14 <LACampbell> (typing on it right now)
2021-07-24 23:56:18 <asdflkj> 2ms pings must be great for timing attacks :>
2021-07-24 23:56:34 <ajin> pings from my house are < 40ms so good but within usual internet infrastructure tolerances. I'm hosting a podcast for locals, that's why I figured I might as well keep it close
2021-07-24 23:57:31 <ajin> it's kvm so I'm sure it depends on whatever else happens to be going on
2021-07-24 23:57:49 <ajin> you can pay extra for a second network port
2021-07-25 00:01:24 <LACampbell> ajin: those are crazy ping response times :) practically accessing your hard disk!
2021-07-25 00:01:59 <LACampbell> maybe you can run a long cable out your window, see if they mind
2021-07-25 00:02:47 <ajin> this is a great episode of On The Metal podcast: https://oxide.computer/podcasts/kenneth-finnegan
2021-07-25 00:13:51 <jeffreybenzos[m]> What are the sudo usecases you're thinking of, invoked?
2021-07-25 00:17:18 <LACampbell> jeffreybenzos[m]: I use doas. no strong reason, smaller config file & binary size just feels right (also a big reason why I use alpine)
2021-07-25 00:19:13 <invoked> sudo does a lot, also has a plugin system for complicated enterprise configs.
2021-07-25 00:19:44 <invoked> doas is better for people who just have an install where it's just their workstation or something.
2021-07-25 00:20:09 <jeffreybenzos[m]> Oh, cool.
2021-07-25 00:20:16 <jeffreybenzos[m]> Do I need chsh, or can i just modify passwd
2021-07-25 00:21:13 <invoked> you can always just modify passwd.
2021-07-25 00:24:42 <jeffreybenzos[m]> turns out i can't, invoked. i'm incompetent and have to figure out how to use doas other than just adding my user to wheel
2021-07-25 00:24:43 <jeffreybenzos[m]> let's goooo
2021-07-25 00:28:24 <asdflkj> echo 'permit :wheel as root' >> /etc/doas.conf
2021-07-25 00:30:06 <asdflkj> on OpenBSD there are nice, commented example configs in /etc/examples, eg. /etc/examples/doas.conf, but unfortunately I don't think those are packaged for alpine (even in *-doc, where I think they should be)
2021-07-25 00:31:31 <LACampbell> openBSD is really well thought out
2021-07-25 00:33:44 <asdflkj> if only libertybsd wasn't dead it would be practically perfect in every way (tm)
2021-07-25 00:34:38 <invoked> *cough* plan9
2021-07-25 00:35:18 <asdflkj> iveqy: nah, too much rat wagging
2021-07-25 00:35:40 <asdflkj> invoked:*
2021-07-25 00:38:25 <asdflkj> someday I'll make a WildeBSD with pure free software, more copyleft and no clang, and it will be perfect
2021-07-25 00:45:09 <LACampbell> asdflkj: when I become a billionaire I'll hire openBSD devs to write man pages for alpine and everything will be OK
2021-07-25 00:46:08 <asdflkj> no, linux (as in Linux(tm)) sucks
2021-07-25 00:47:11 <LACampbell> meh, alpine made me realise I never disliked linux, I disliked GNU :D
2021-07-25 00:47:38 <zcrayfish> same here.
2021-07-25 00:47:47 <asdflkj> opposite for me
2021-07-25 00:47:54 <jeffreybenzos[m]> Do I need xf86 ati if I've got amdgpu?
2021-07-25 00:48:02 <raven523> no
2021-07-25 00:48:14 <jeffreybenzos[m]> startx is mad and looking for ati
2021-07-25 00:48:25 <jeffreybenzos[m]> dunno why it's asking for it, and then it gives me an error because it's like "bro ast and ati bro"
2021-07-25 00:48:34 <raven523> check your /etc/X11/xorg.conf and /etc/X11/xorg.conf.d
2021-07-25 00:48:47 <jeffreybenzos[m]> can I just remove those lines? included vesa?
2021-07-25 00:48:49 <raven523> are you sure your GPU works with amdgpu?
2021-07-25 00:49:16 <jeffreybenzos[m]> rx 480?
2021-07-25 00:49:29 <raven523> in general you should not have /etc/X11/xorg.conf. If it doesn't exist, xorg will try to configure itself automatically when it starts and will read all of the files in /etc/X11/xorg.conf.d
2021-07-25 00:49:35 <raven523> rx 480 definitely works with amdgpu
2021-07-25 00:49:52 <jeffreybenzos[m]> lemme just kill it real quick
2021-07-25 00:49:53 <jeffreybenzos[m]> or rename it
2021-07-25 00:50:17 <jeffreybenzos[m]> uh oh, it's not even in there.
2021-07-25 00:50:26 <jeffreybenzos[m]> just xinit in there
2021-07-25 00:50:56 <jeffreybenzos[m]> (a directory)
2021-07-25 00:51:05 <raven523> that's probably fine. I forget whether you need (e)udev running for xorg to detect the GPU correctly
2021-07-25 00:51:52 <raven523> is xf86-video-amdgpu installed?
2021-07-25 00:52:01 <jeffreybenzos[m]> i swear it was. how can i double-verify
2021-07-25 00:52:15 <raven523> apk info xf86-video-amdgpu
2021-07-25 00:52:35 <jeffreybenzos[m]> It gives me a description
2021-07-25 00:52:48 <jeffreybenzos[m]> Says installed size: 180 KiB; is that always in there?
2021-07-25 00:53:36 <jeffreybenzos[m]> yeah, no clue if this actually tells me i fit's installed, since i just ran `apk info bash` and `apk info john` (one installed, one not-installed package) and they look the same
2021-07-25 00:53:43 <invoked> jeffreybenzos[m]: unless you have a reason to run xorg maybe consider wayland
2021-07-25 00:54:11 <jeffreybenzos[m]> mostly to avoid wayland, lmao
2021-07-25 00:54:22 <jeffreybenzos[m]> i just hate drew devault, and my senses of hating specific devs has done wonders so far
2021-07-25 00:54:57 <invoked> hate is a strong word
2021-07-25 00:55:46 <invoked> anyway wayland is not drew and vice versa.
2021-07-25 00:56:14 <jeffreybenzos[m]> he's definitely a big force in it, though, but i also just don't like wayland's restructuring where there's been some big Ls done
2021-07-25 00:58:50 <jeffreybenzos[m]> some protocol stuff that i can't even remember because of how it all happened like three years ago
2021-07-25 01:01:29 <asdflkj> I don't understand how he can rant about rust unportability and be blind to all of wayland's #include <linux/input.h>
2021-07-25 01:01:51 <jeffreybenzos[m]> lol
2021-07-25 01:01:58 <jeffreybenzos[m]> i'm saving that one for my next tirate
2021-07-25 01:02:01 <jeffreybenzos[m]> tirade, even
2021-07-25 01:02:19 <jeffreybenzos[m]> brb, fun, but then i'll be back and wondering how
2021-07-25 01:02:31 <jeffreybenzos[m]> raven523: I verified it was installed, by the way. Unsure what else
2021-07-25 01:02:33 <jeffreybenzos[m]> but i'll be back
2021-07-25 01:02:51 <asdflkj> this was on funny imo: https://news.ycombinator.com/item?id=27253430
2021-07-25 01:03:21 <asdflkj> s/on//
2021-07-25 01:03:21 <alpine-meetbot> asdflkj meant to say: this was  funny imo: https://news.ycombinator.com/item?id=27253430
2021-07-25 01:10:56 <invoked> i can't wait for people to blame whatever happens with xorg on wayland for taking away developer interest.
2021-07-25 01:11:19 <invoked> the culty people on either side need to find new hobbies
2021-07-25 01:11:23 <jeffreybenzos[m]> xorg is in legacy anyway now, right?
2021-07-25 01:11:42 <jeffreybenzos[m]> i thought i read about how it's not going to receive updates any further, or something
2021-07-25 01:12:02 <invoked> it'll get patched if vulns get found, but beyond that, nobody is doing the work
2021-07-25 01:12:08 <asdflkj> xorg is basically complete
2021-07-25 01:12:18 <asdflkj> xenocara still gets updates afaik
2021-07-25 01:13:43 <jeffreybenzos[m]> Oh, interesting. Yeah, I didn't think there was much to update. I'd use some replacement that was done in rust or something. :^)
2021-07-25 01:14:57 <jeffreybenzos[m]> Kinda crazy that the server almost instantly errors out. Hmm
2021-07-25 01:15:04 <asdflkj> xenocara has more focus on security, simplicity, and correctness, like any OpenBSD stuff
2021-07-25 01:15:47 <jeffreybenzos[m]> Does it work as some sort of dropin, or are there few programs that support it?
2021-07-25 01:16:16 <asdflkj> I think it's fairly drop in because x11 is a standard
2021-07-25 01:17:25 <invoked> maintaining portability to linux/et al is not a goal of xenocara.
2021-07-25 01:18:46 <jeffreybenzos[m]> https://paste.ee/p/JQSyJ here's my xorg log
2021-07-25 01:19:03 <asdflkj> fortunately gnu portability to bsd is a goal of hyperbola, may it reach prod long before hurd (*crosses fingers*)
2021-07-25 01:21:13 <asdflkj> I had about the same problem with xorg on alpine a couple weeks ago, now I'm vacation and have given up for the time being
2021-07-25 01:22:46 <jeffreybenzos[m]> lmao asdflkj; so you weren't using it?
2021-07-25 01:22:52 <asdflkj> iirc using vesa and adding startx to a bunch of groups made x start, but not respond to any input
2021-07-25 01:23:06 <jeffreybenzos[m]> fun...?
2021-07-25 01:23:17 <jeffreybenzos[m]> wish the error was more descriptive
2021-07-25 01:23:56 <asdflkj> yeah, I'm more productive when I can't watch videos or even see comment indentation on the orange site
2021-07-25 01:24:35 <jeffreybenzos[m]> super true, though i intend to use this as the main machine, which means memes are a must
2021-07-25 01:24:40 <jeffreybenzos[m]> unless i can run firefox in a framebuffer...
2021-07-25 01:24:46 <jeffreybenzos[m]> (kidding; i'd like to game etc etc)
2021-07-25 01:25:07 <asdflkj> you can run links2 in a fb iirc :)
2021-07-25 01:25:52 <asdflkj> actually I wonder if netsurf works there, that could be awesome
2021-07-25 01:26:33 <jeffreybenzos[m]> lol
2021-07-25 01:27:08 <asdflkj> but goshdarnit, now I realize I can probably read the orange site without X (with links2), now this channel owes me $$$ in productivity loss
2021-07-25 01:30:07 <invoked> i fully expect xenocara to have a reverse-xwayland (wayxorg?) at some point
2021-07-25 01:30:29 <invoked> if it hasn't already been at least discussed
2021-07-25 01:31:28 <invoked> not sure how that would work without seams, tho
2021-07-25 01:31:58 <asdflkj> that'll be nice if any good software ever gets made wayland only, but that's kinda oxymoronic imo
2021-07-25 01:32:50 <jeffreybenzos[m]> Oh, another curiosity: At some point, linux stopped using both of my screens and mirroring them
2021-07-25 01:33:05 <jeffreybenzos[m]> Even on the live boot disk, it's now only using one of my screens for the console, and not the entire screen
2021-07-25 01:34:35 <invoked> xorg won't go away.  it'll probably be a slow burn of declining interest, and before there's any real consensus about this stuff, we'll just be trying to not go the way of the dinosaurs.
2021-07-25 01:35:48 <rcf> invoked: the problem with wayland applications is that they are inherently more tied to DE/compositor because the core protocol is so anemic. You can emulate a generic X WM; Wayland, less so. 
2021-07-25 01:37:16 <rcf> And the unpleasant reality is that Xorg is likely being retained by people (myself included) because it provides functionality that Wayland can't or won't ever see implemented. 
2021-07-25 01:37:27 <asdflkj> you'd probably emulate wlroots bc it's designed to work for more than one de
2021-07-25 01:38:23 <asdflkj> xorg'll probably be similar to sysv init
2021-07-25 01:39:36 <invoked> rcf: sure, if eg network transparency is something you want/need
2021-07-25 01:39:54 <invoked> i'm not arguing for or against either one. i am arguing against cults, tho
2021-07-25 01:40:55 <asdflkj> it'll shrink to a tiny niche, but still have a decent and fairly similar obsd implementation (xenocara/openrc) as an alternative to the nasty newish freedesktop thing (wayland/systemd)
2021-07-25 01:47:07 <jeffreybenzos[m]> Yep, sigabort still happening after double-checking I had the firmware installed and configured x to only look for amdgpu
2021-07-25 01:47:19 <jeffreybenzos[m]> still some random sigabort
2021-07-25 01:48:20 <jeffreybenzos[m]> This may get me to try wayland, invoked. ;D
2021-07-25 01:52:23 <asdflkj> is your user added to groups like video?
2021-07-25 01:53:08 <jeffreybenzos[m]> let's double-check
2021-07-25 01:53:35 <jeffreybenzos[m]> yeah, we're in there, asdflkj
2021-07-25 02:00:47 <asdflkj> jeffreybenzos[m]: rereading the irc logs, I think what finally made my xorg start was installing f86-video-vesa
2021-07-25 02:01:35 <asdflkj> IIUC that's unnaccelerated but may be less buggy
2021-07-25 02:02:19 <jeffreybenzos[m]> very interesting; would i still be able to use my gpu for other stuff?
2021-07-25 02:02:33 <jeffreybenzos[m]> or will that just default to the lower perf for everything
2021-07-25 02:02:59 <asdflkj> everything I (not an expert) think
2021-07-25 02:04:21 <asdflkj> s/ I/, I/
2021-07-25 02:04:21 <alpine-meetbot> asdflkj meant to say: everything, I (not an expert) think
2021-07-25 02:07:39 <jeffreybenzos[m]> yep, sadly no big fix there
2021-07-25 02:52:12 <Skaface> asdflkj: have you tried the modesetting driver?
2021-07-25 02:52:26 <asdflkj> I don't remember
2021-07-25 02:52:40 <asdflkj> I'm away from the computer I wanted alpine on right now
2021-07-25 02:53:07 <Skaface> ah ok, well give that a go next time
2021-07-25 02:53:48 <asdflkj> ok, thanks for the tip
2021-07-25 03:00:48 <jeffreybenzos[m]> skaface: modesetting driver?
2021-07-25 03:01:35 <asdflkj> pkg search modesetting, or kms?
2021-07-25 03:01:44 <asdflkj> s/pkg/apk/ :P
2021-07-25 03:01:44 <alpine-meetbot> asdflkj meant to say: apk search modesetting, or kms?
2021-07-25 03:05:06 <jeffreybenzos[m]> looks like it's in xf86-video-modesetting, asdflkj 
2021-07-25 03:05:22 <asdflkj> probably
2021-07-25 03:16:25 <jeffreybenzos[m]> Thanks for the slant advice, skaface; seems like that works
2021-07-25 03:17:44 <jeffreybenzos[m]> but now i have to wonder if amdgpu is just borked
2021-07-25 03:17:50 <jeffreybenzos[m]> which is so sad
2021-07-25 03:58:34 <jeffreybenzos[m]> invoked: Is there a way to just use everything I've previously used, but with xwayland in the middle of it?
2021-07-25 04:20:19 <Hello71> asdflkj: why do you need to run apk search as root
2021-07-25 04:20:55 <asdflkj> you don't, good point
2021-07-25 04:21:18 <Hello71> and this is exactly why complex sudo policies are bad, because if you allow anybody to apk search *, that's arbitrary privesc via apk search -X
2021-07-25 04:21:50 <asdflkj> or why complex software is bad
2021-07-25 04:22:21 <Hello71> xf86-video-modesetting doesn't contain anything
2021-07-25 04:22:30 <Hello71> the package description is "dummy package for compat - can safelu be removed"
2021-07-25 04:22:40 <asdflkj> maybe it has a useful hook?
2021-07-25 04:23:08 <Hello71> no
2021-07-25 04:23:24 <Hello71> hooks kinda suck, i wish we had less of them
2021-07-25 04:25:52 <jeffreybenzos[m]> oh, interesting, hello71
2021-07-25 04:26:24 <asdflkj> it doesn't even install some package that's been renamed?
2021-07-25 04:28:37 <Hello71> it's very clear, it's dummy package for compat
2021-07-25 04:28:42 <jeffreybenzos[m]> lmao
2021-07-25 04:28:43 <Hello71> modesetting driver was moved to xorg-server
2021-07-25 04:29:00 <jeffreybenzos[m]> it wasn't clear when i read it, as i thought it meant a "dummy" that just did a small set of things
2021-07-25 04:29:11 <jeffreybenzos[m]> since the word "dummy" is super abused now, and doesn't really mean "utterly empty"
2021-07-25 04:29:21 <Hello71> notwithstanding the typo, "can safelu be removed" is also clear
2021-07-25 04:29:27 <jeffreybenzos[m]> but that's from someone who hasn't read it in other packages, of course
2021-07-25 04:29:47 <Hello71> in theory you could use xf86-video-vesa and still have opengl, but in practice if modesetting doesn't work then your opengl is probably totally busted
2021-07-25 04:30:07 <Hello71> except maybe nvidia?
2021-07-25 04:31:08 <jeffreybenzos[m]> then i've gotta change my modules around and remake initfs, right?
2021-07-25 04:31:12 <jeffreybenzos[m]> can try, at least
2021-07-25 04:31:17 <Hello71> no
2021-07-25 04:31:24 <Hello71> xorg ddx is totally unrelated to kernel modules
2021-07-25 04:31:31 <jeffreybenzos[m]> also, to clarify, modesetting worked for me
2021-07-25 04:31:39 <jeffreybenzos[m]> if you were just talking to asdflkj though, no clue on his side
2021-07-25 04:31:52 <jeffreybenzos[m]> i just wanna be able to run darktable with opencl
2021-07-25 04:32:46 <asdflkj> I won't have access to the pc I was trying to install alpine on until after the vacation I'm on right now
2021-07-25 04:33:02 <greatwielder[m]> when you set swapiness to 200, does this mean anon pages get swapped in preference to cache?
2021-07-25 04:33:02 <jeffreybenzos[m]> just gimme your address asdflkj
2021-07-25 04:33:05 <jeffreybenzos[m]> i'll go figure it out
2021-07-25 04:33:18 <asdflkj> jeffreybenzos[m]: 127.0.0.1
2021-07-25 04:33:25 <Skaface> would the nomodeset kernel option prevent X from using modesetting?
2021-07-25 04:33:38 <jeffreybenzos[m]> wtf our graphics cards look the same!!
2021-07-25 04:33:59 <asdflkj> hmm, try 0.0.0.0
2021-07-25 04:34:01 <jeffreybenzos[m]> skaface: everywhere i've read said it does.
2021-07-25 04:34:22 <jeffreybenzos[m]> asdflkj: i'd /part normally for the joke, but i'm on matrix so it's a pain
2021-07-25 04:34:24 <Skaface> yeah i would assume so
2021-07-25 04:34:49 <jeffreybenzos[m]> i'm gonna call amd and demand them to fix it
2021-07-25 04:35:16 <Hello71> opencl is a totally different beast
2021-07-25 04:35:27 <jeffreybenzos[m]> orly? oh boy
2021-07-25 04:35:28 <Hello71> and mesa opencl sucks and in particular doesn't work for darktable
2021-07-25 04:35:40 <Hello71> something jpeg something
2021-07-25 04:35:56 <Hello71> Skaface: nomodeset was removed from default config
2021-07-25 04:35:59 <Hello71> but yes
2021-07-25 04:36:14 <Hello71> greatwielder[m]: could be
2021-07-25 04:37:11 <Hello71> you need rocm for proper opencl support, and i doubt anyone's dared to try it on musl
2021-07-25 04:37:27 <Hello71> theoretically it should work but the code quality is shit so i doubt it
2021-07-25 04:37:49 <jeffreybenzos[m]> lol
2021-07-25 04:38:17 <greatwielder[m]> when you use swap priority, what happens when the highest priority swap device gets filled? will the oldest memory in it get moved to the lower priority swap device to make room?
2021-07-25 04:39:56 <greatwielder[m]> I found that with zram you can make a ridiculously good trade for some cpu to more than double your ram, I wanna be greedy and set up another zram tier and use zstd for it.
2021-07-25 04:41:45 <greatwielder[m]> server with 512 gb ram is juggling a dbs thats ttaking 1.3 tb ram, latency barely increased.
2021-07-25 04:53:25 <Skaface> interesting
2021-07-25 05:16:31 <jeffreybenzos[m]> switched over to trying to figure out setting ata security, accidentally set user security on a different drive that won't unfreeze now, and the drive i set to maximum security isn't prompting for a pass
2021-07-25 05:16:32 <jeffreybenzos[m]> lmao
2021-07-25 07:00:25 <jeffreybenzos[m]> ikke: do you get a segfault with the newest darktable?
2021-07-25 07:04:09 <jeffreybenzos[m]> Totally could be because I'm in the modesetting driver and thus lack glamor
2021-07-25 07:07:18 <jeffreybenzos[m]> nevermind, modesetting supposedly supports glamor, but yeah, both alacritty and darktable aren't loading with that setup
2021-07-25 07:07:25 <jeffreybenzos[m]> i will away to shleep
2021-07-25 07:07:39 <ikke> I'll test it later
2021-07-25 07:21:24 <jeffreybenzos[m]> thanks a bunch, ikke
2021-07-25 07:21:32 <jeffreybenzos[m]> but i'm p sure it's just me and my broken amdgpu situation
2021-07-25 07:22:21 <jeffreybenzos[m]> oh cool, someone already flagged veracrypt because it's old as hell
2021-07-25 07:30:31 <Skaface> ahh ive been messing with this all afternoon.. when my monitor goes into sleep mode, every minute the backlight turns on, displays "no input" then turns off again
2021-07-25 07:30:43 <Skaface> at exact 1 minute intervals
2021-07-25 07:31:31 <jeffreybenzos[m]> lmao, it just wants to shake hands as often as possible
2021-07-25 07:39:05 <Skaface> not when im trying to sleep!
2021-07-25 12:01:23 <Hello71> greatwielder[m]: i don't think swap priority is used that much
2021-07-25 12:01:48 <Hello71> usually if you have a real swap device you should use zswap
2021-07-25 12:02:17 <Hello71> jeffreybenzos[m]: modesetting *uses* glamor
2021-07-25 12:02:53 <Hello71> x programs don't interact with glamor at all, they use xrender or opengl
2021-07-25 12:03:40 <Hello71> on linux if you need to open a tcrypt container just use cryptsetup
2021-07-25 12:42:05 <jeffreybenzos[m]> cryptsetup can handle veracrypt? that's good to know; i'll look into it
2021-07-25 12:42:26 <jeffreybenzos[m]> especially now that i have to consider my options now that the drive is like "heehee no security options detected AT ALL BRO"
2021-07-25 15:14:39 <graywolf> When writing openrc service file, how can I log stdout/err? By program (gunicorn) should just log to stdout(err?), but I do not see anything in /var/log/message; Is there some extra variable I need to set?
2021-07-25 15:17:13 <ikke> openrc has no system for capuring stdin/stdout
2021-07-25 15:17:21 <ikke> stdout/stderr*
2021-07-25 15:20:22 <graywolf> Ah, so I want to use output_log and error_log to set the log file if I read the manpage correctly?
2021-07-25 16:07:37 <ikke> graywolf: yes, I suppose so
2021-07-25 17:15:23 <nanopier> hey everyone
2021-07-25 17:15:39 <nanopier> quite some people around here, awesome!
2021-07-25 17:16:07 <nanopier> wanted to ask whether anyone has practical experience of aarch64 installations on anything else than raspberry pi
2021-07-25 17:16:40 <nanopier> im struggling with a nanopi r2s in particular, which i see has a dtb in the official alpine release, but no u-boot... :(
2021-07-25 17:16:45 <caskd> the postmarketos peeps might be able to help you since it runs aarch64
2021-07-25 17:17:03 <nanopier> that's a good idea!
2021-07-25 17:17:06 <caskd> pinephone postmarketos*
2021-07-25 17:17:18 <nanopier> do they also have IRC? how could i contact them?
2021-07-25 17:19:38 <caskd> some are here, they probably will reply as soon as they can
2021-07-25 17:26:10 <geoman> how do i compile a browser that requires the header file sys/cdefs.h
2021-07-25 17:26:40 <geoman> i have libc6-compat and the glibc compat package installed
2021-07-25 17:26:57 <caskd> you need bsd-compat-headers
2021-07-25 17:27:11 <caskd> https://pkgs.alpinelinux.org/contents?file=cdefs.h&path=&name=&branch=edge
2021-07-25 17:27:34 <nanopier> do you guys have any kind of bounty program by the way?
2021-07-25 17:27:51 <ikke> no, we have not
2021-07-25 17:28:01 <nanopier> i mean where people could post issues / questions they would love to see resolved / answered
2021-07-25 17:28:17 <caskd> that would probably be gitlab issues
2021-07-25 17:28:19 <nanopier> and offer a contribution to the "alpine foundation" if there is any :)
2021-07-25 17:29:22 <nanopier> yes i know about the gitlab issues option
2021-07-25 17:29:32 <ikke> We do not have any system for that
2021-07-25 17:30:29 <nanopier> is anyone interested in pulling something like that off?
2021-07-25 17:30:52 <nanopier> i love alpine linux myself, and we heavily use it at the company where i work as well
2021-07-25 17:31:10 <nanopier> could pull some strings, no promises though. just an idea
2021-07-25 17:31:13 <ikke> We do not have a foundation or other legal entity yet which can receive donations
2021-07-25 17:31:33 <ikke> We are looking if we could set something up, but nothing concrete yet
2021-07-25 17:31:49 <nanopier> sounds good
2021-07-25 17:32:22 <nanopier> is there anyone who could be contacted regarding this? i think my guys at the firm would love it
2021-07-25 17:34:12 <ikke> Any one of the council members, ncopa, clandmeter or me
2021-07-25 17:40:32 <geoman> what's the difference between PTHREAD_ADAPTIVE_MUTEX_INITIALIZER_NP and PTHREAD_MUTEX_INITALIZER?
2021-07-25 17:41:53 <geoman> i'm trying to compile the browser i mentioned earlier and one of the c files failed to compile bc PTHREAD_ADAPTIVE_MUTEX_INITIALIZER_NP and PTHREAD_MUTEX_ADAPTIVE_NP were unrecognized
2021-07-25 17:42:32 <geoman> PTHREAD_MUTEX_INITIALIZER and PTHREAD_MUTEX_STALLED were suggested instead of the unrecognized functions
2021-07-25 17:43:05 <nanopier> awesome, thank you @ikke, will get back to you on this!
2021-07-25 17:44:41 <geoman> i tried searching the web but didn't find anything on the adaptive initializer
2021-07-25 17:48:06 <geoman> update: i did find something, but it says that the adaptive mutexs are used for something called SMP
2021-07-25 17:49:55 <geoman> as in they're used under SMP, not for it
2021-07-25 17:59:32 <geoman> i'm not really concerned about the mutex initializers now as PTHREAD_MUTEX_INTIALIZER is in the code
2021-07-25 17:59:55 <geoman> i'm more concerned about PTHREAD_MUTEX_STALLED vs PTHREAD_MUTEX_ADAPTIVE_Np
2021-07-25 19:34:26 <danieli> is there a legal entity representing alpine these days or still none?
2021-07-25 19:34:35 <danieli> at least a TSC of sorts is a step up
2021-07-25 19:34:52 <danieli> ah you just said that, i missed it, my bad
2021-07-25 19:34:59 <danieli> nevermind me :)
2021-07-25 19:38:25 <ikke> :)
2021-07-25 19:38:42 <ikke> https://gitlab.alpinelinux.org/alpine/council/-/issues/6
2021-07-25 19:39:10 <danieli> hopefully it'll allow for bigger orgs to justify donating to alpine, and that it in turn will make it easier for some of the core devs to justify spending time on alpine
2021-07-25 19:39:23 <danieli> consulting as a side gig isn't necessarily super lucrative
2021-07-25 20:23:47 <jeffreybenzos[m]> Today I will get distracted and not debug the amd gpu.
2021-07-25 21:05:44 <jeffreybenzos[m]> Whoops; meant to ask here. Can I downgrade all of my packages from the versions in edge to the versions in 3.14?
2021-07-25 21:07:25 <ikke> change the repository in /etc/apk/repositories, and then run apk upgrade -U --available
2021-07-25 21:07:48 <jeffreybenzos[m]> Oh, --available included. Got it.
2021-07-25 21:08:06 <ikke> yes, otherwise it would not downgrade any packages
2021-07-25 21:08:26 <jeffreybenzos[m]> Makes sense. Looks like polkit is mad
2021-07-25 21:08:27 <jeffreybenzos[m]> surprise
2021-07-25 21:09:00 <jeffreybenzos[m]> and then, if i read this correctly, ikke, I can select certain packages from the edge repo when I want to use those?
2021-07-25 21:10:27 <ikke> technically you could, but it's not really supported
2021-07-25 21:10:38 <ikke> mostly due to different versions in dependencies
2021-07-25 21:12:22 <jeffreybenzos[m]> Ah, interesting, ikke
2021-07-25 21:12:42 <jeffreybenzos[m]> Turns out downgrading to 3.14 didn't fix my amdgpu issue. All out of ideas until the looking glass discord tells me the secrets
2021-07-25 21:18:33 <caskd> jeffreybenzos[m]: what kind of issue do you have?
2021-07-25 21:18:40 <caskd> i am using the amdgpu driver as well
2021-07-25 21:19:36 <jeffreybenzos[m]> caskd: launching x gives a completely unhelpful sigabort error
2021-07-25 21:19:49 <jeffreybenzos[m]> works with modesetting driver instead, but then "lol no GL context" so half my applications don't work
2021-07-25 21:19:59 <caskd> i don't think that's necessarly a amdgpu issue
2021-07-25 21:20:17 <caskd> do you have any logs and do you use startx?
2021-07-25 21:20:23 <jeffreybenzos[m]> it's the only time that the issue crops up (when removing amdgpu, modesetting kicks in and x actually starts)
2021-07-25 21:20:26 <jeffreybenzos[m]> yeah, i do, caskd
2021-07-25 21:21:45 <jeffreybenzos[m]> https://paste.ee/p/K8vET for your viewing pleasure, caskd
2021-07-25 21:21:58 <caskd> thanks
2021-07-25 21:22:03 <jeffreybenzos[m]> obviously setting the 20-amdgpu.conf with the mandatory driver selection for amdgpu cleans that up a bit
2021-07-25 21:23:49 <caskd> that's for sure not a amdgpu issue
2021-07-25 21:23:58 <caskd> it is already loaded by the time it aborts
2021-07-25 21:24:28 <jeffreybenzos[m]> you don't think so? hmm. to clarify, the reason why i say it's an amdgpu issue is because i can uninstall amdgpu and the problem goes "away" (modesetting works)
2021-07-25 21:24:37 <caskd> that's odd
2021-07-25 21:24:50 <caskd> it seems like the glamoregl causes it
2021-07-25 21:24:51 <jeffreybenzos[m]> and by that I obviously mean that modesetting "works"
2021-07-25 21:25:11 <jeffreybenzos[m]> if it's all synchronous, then i agree with you
2021-07-25 21:25:34 <caskd> afaik, yes, but it could be a underlying issue that's not logged
2021-07-25 21:26:07 <jeffreybenzos[m]> absolutely. i'm sorry to cut off good help, but i've got a remote obligation
2021-07-25 21:26:23 <jeffreybenzos[m]> that's kind of inspiring, though i don't know if i can choose another egl provider to test this hypothesis
2021-07-25 21:26:28 <jeffreybenzos[m]> i assume that's exactly configurable
2021-07-25 21:26:51 <caskd> on my host, when that loads the driver enables the X accel with glamor in my logs
2021-07-25 21:27:01 <caskd> so i suspect that's the problem
2021-07-25 21:27:26 <caskd> loads, the driver
2021-07-25 22:00:35 <minimal> caskd: re: your mailing list comment about Syslinux not supporting UEFI - it does but you have to set it all up manually
2021-07-25 22:01:31 <caskd> minimal: i mean... that's one more reason to not make it bootloader specific
2021-07-25 22:01:43 <caskd> that's exactly why i argued against it
2021-07-25 22:03:10 <minimal> caskd: ah ok, if was you use of the word "if" when you said "If syslinux supported UEFI" that I was picking up on
2021-07-25 22:03:33 <caskd> thanks for letting me know
2021-07-25 22:04:01 <caskd> i didn't think syslinux is uefi capable
2021-07-25 22:06:40 <minimal> it is but it has limitations that make it fiddly
2021-07-25 23:26:59 <Skaface> i wish that the alpine_dev= would mount that device to a standard mount point, no matter the media.. like /media/alpine or something. It seems pretty difficult to put a copy of a pre-configured system onto a different type of media
2021-07-25 23:41:04 <Skaface> why on earth would they make the routine that mounts devices on startup a compiled C program rather than a shell script?
2021-07-25 23:42:37 <jeffreybenzos[m]> what's the apk command for "why is this package installed"
2021-07-25 23:42:50 <Hello71> apk del
2021-07-25 23:42:53 <Hello71> ... kind of
2021-07-25 23:43:05 <jeffreybenzos[m]> lmao
2021-07-25 23:44:04 <Hello71> Skaface: as i recall it only prints the device
2021-07-25 23:49:15 <jeffreybenzos[m]> interesting; wayland-libs-{four packages} is used by a bunch of stuff? why is the wayland section req'd when i'm only using xorg
2021-07-25 23:51:00 <jeffreybenzos[m]> Kinda looks like I can use mesa-egl instead? Is that right?
2021-07-25 23:52:42 <jeffreybenzos[m]> Oh, interesting? xorg-server owns libglamoregl.so, and glamor-egl tries to provide its own
2021-07-25 23:53:24 <jeffreybenzos[m]> The dependency goes mesa-egl->glamor-egl, so I assume that's not gonna work to just rip out glamor, though it's also interesting to see that it has co-ownership
2021-07-26 02:14:24 <Sheila> wayland-libs-* is used by a bunch of stuff because a lot of stuff supports wayland and instead of writing wayland-specific versions they just shoved it on to their existing codebases.
2021-07-26 03:01:05 <jeffreybenzos[m]> lmfao, you guys are not going to believe this shit.
2021-07-26 03:01:14 <jeffreybenzos[m]> So I install all the stuff to try sway. Try running sway. Doesn't work.
2021-07-26 03:01:19 <jeffreybenzos[m]> Now startx is working fine.
2021-07-26 03:01:26 <jeffreybenzos[m]> mind is fucking reeling at what I'm missing
2021-07-26 03:03:09 <jeffreybenzos[m]> To clarify, it's working WITH amdgpu, now
2021-07-26 03:24:19 <invoked> would it surprise you if i did believe that shit
2021-07-26 03:24:43 <danieli> invoked: i am absolutely shooketh
2021-07-26 03:26:31 <danieli> but yeah, i had bad luck on some of the hardware i tried it on too, didn't care to mess about with it enough to fix it
2021-07-26 03:27:54 <jeffreybenzos[m]> lmao, invoked
2021-07-26 03:28:26 <jeffreybenzos[m]> i guess i can try paring it down later. one thing i'm curious about now is if i can actually try out custom-install stuff by making a progressively bigger script--where i just boot the live image and then see if i can boot shit from there
2021-07-26 03:28:46 <jeffreybenzos[m]> so that i can finally one day have a clean ACTUALLY MINIMAL install that i get everything working on, while still being able to boot my main mostly-working cruftbox
2021-07-26 03:41:06 <jeffreybenzos[m]> however, ikke, darktable still crashes
2021-07-26 03:56:19 <jeffreybenzos[m]> Anyone here using pipewire? ALSA itself was fine for me in the past, but I'm curious.
2021-07-26 04:18:37 <jeffreybenzos[m]> Looks like alsa's missing a particular "/usr/share/alsa/ucm2/ucm.conf"
2021-07-26 04:32:19 <ikke> jeffreybenzos[m]: can you open an issue on gitlab.alpinelinux.org/alpine/aports?
2021-07-26 04:54:51 <jeffreybenzos[m]> Yeah, for darktable? I can try.
2021-07-26 04:55:02 <jeffreybenzos[m]> Also, interesting. The alsa situation is not giving me audio despite my user being in the audio group
2021-07-26 06:16:11 <jeffreybenzos[m]> Was a typical situation where alsa trolled me and reordered the cards.
2021-07-26 08:57:50 <Skaface> i moved to pipewire, and had all sorts of problem with bluetooth audio
2021-07-26 08:58:01 <Skaface> in arch, not alpine
2021-07-26 08:59:02 <Skaface> audio was stuttering, 500ms delay
2021-07-26 08:59:16 <Skaface> messing with it for ages, ended up going back to pulseaudio
2021-07-26 08:59:51 <Skaface> my other pc, it works ok, but my speakers thump whenever it starts playing audio
2021-07-26 09:02:51 <Skaface> i found a post in their dev forums about that issue, but it wasnt resolved because the developers were just arguing with each other about the "right" way to do it
2021-07-26 09:02:53 <Skaface> so im not sold
2021-07-26 09:26:34 <jeffreybenzos[m]> looool, skaface
2021-07-26 09:26:43 <jeffreybenzos[m]> unfortunate to hear that it wasn't perfect
2021-07-26 09:26:55 <jeffreybenzos[m]> https://gitlab.alpinelinux.org/alpine/aports/-/issues/12869 no clue if i did it right, ikke,b ut here she goes
2021-07-26 09:26:57 <jeffreybenzos[m]> but**
2021-07-26 09:27:04 <ikke> Yes, I got the notification, thanks
2021-07-26 09:27:26 <PureTryOut1> jeffreybenzos: I've been using Pipewire as the main audio service on my laptop for quite a while, works fine
2021-07-26 09:27:26 <ikke> fyi, it's the -dbg package, if available, that normally has the debug symbols
2021-07-26 09:27:41 <PureTryOut1> I'm just experiencing one bug but the main dev can't reproduce it 😢
2021-07-26 09:28:11 <jeffreybenzos[m]> ikke: yeah, i've got that in there, but then darktable-dbg doesn't actually do it--i bet it's something like "darktable --mode=debug" or some garbo
2021-07-26 09:28:52 <jeffreybenzos[m]> it was darktable -d, if i recall, and lol, still no debug already in there
2021-07-26 10:45:53 <Aftershock> hey
2021-07-26 10:46:17 <Aftershock> does anyone have experience setting up btrfs as the root filesystem for alpine?
2021-07-26 10:46:54 <Aftershock> I've tried several different ways but grub refuses to boot after I setup the disk
2021-07-26 10:49:35 <Aftershock> I followed the wiki for alternate mount points and I exported the use efi=1 setting as well as packages mentioned to install to a custom partitioning scheme
2021-07-26 10:54:03 <caskd> Aftershock: i do use it
2021-07-26 10:54:23 <caskd> Do you use GPT or MBR?
2021-07-26 10:54:45 <Aftershock> gpt and efi
2021-07-26 10:55:10 <caskd> so grub just straight up doesn't load at all or is it that the initramfs cannot mount it?
2021-07-26 10:55:30 <Aftershock> grub will load
2021-07-26 10:55:44 <Aftershock> but it says unrecognised filesystem
2021-07-26 10:56:10 <Aftershock> and plonks into a recovery shell
2021-07-26 10:56:28 <Aftershock> and I can't ls the disks
2021-07-26 10:57:00 <caskd> do you have /boot on btrfs?
2021-07-26 10:57:27 <Aftershock> I was mounting a vfat partition for the bootloader
2021-07-26 10:57:37 <Aftershock> at /mnt/boot 
2021-07-26 10:57:57 <Aftershock> and then a btrfs root subvolume to /mnt
2021-07-26 10:58:04 <caskd> and grub was set to use that with --boot-directory ?
2021-07-26 10:58:17 <caskd> on grub-install?
2021-07-26 10:58:20 <Aftershock> well I did the install to /mnt
2021-07-26 10:58:28 <Aftershock> so that should be the root
2021-07-26 10:59:02 <caskd> right, but it seems like grub didn't set up the stage 1 bootloader properly so it cannot load configs and the rest from the vfat partition
2021-07-26 10:59:36 <Aftershock> hmmm
2021-07-26 10:59:54 <caskd> can you chroot into your installation and run grub-install manually with that argument set and everything mounted?
2021-07-26 11:00:03 <caskd> that should fix it
2021-07-26 11:00:27 <jeffreybenzos[m]> Aftershock: I set mine up that way
2021-07-26 11:00:29 <jeffreybenzos[m]> I also don't use grub at all.
2021-07-26 11:00:41 <Aftershock> refind?
2021-07-26 11:00:44 <jeffreybenzos[m]> Nope.
2021-07-26 11:00:50 <jeffreybenzos[m]> Fuck a bootloader. The kernel boots itself.
2021-07-26 11:00:58 <Aftershock> heh?
2021-07-26 11:01:13 <caskd> you just straight up load the kernel?
2021-07-26 11:01:17 <jeffreybenzos[m]> yes
2021-07-26 11:01:28 <Aftershock> can I run setup-disk to /mnt and then chroot and run the grubinstall
2021-07-26 11:01:32 <jeffreybenzos[m]> linux is the best
2021-07-26 11:01:46 <Aftershock> you need to do anything special for that?
2021-07-26 11:01:46 <jeffreybenzos[m]> did you already do anything significant to your install, aftershock?
2021-07-26 11:01:54 <caskd> Aftershock: you don't have to re-run it if you already installed it
2021-07-26 11:02:01 <jeffreybenzos[m]> hell no, other than giving up on grub
2021-07-26 11:02:06 <Aftershock> I think I reformatted
2021-07-26 11:02:16 <caskd> in that case, you do have to then
2021-07-26 11:02:41 <caskd> jeffreybenzos[m]: what format is your rootfs? is the fs driver built into the kernel?
2021-07-26 11:03:01 <caskd> booting without initramfs seems sketchy to say the least
2021-07-26 11:03:05 <Aftershock> should I be using fdisk or gptfdisk for the formatting
2021-07-26 11:03:14 <jeffreybenzos[m]> boot the live disk, add btrfs to your kernel modules, mkinitfs with esp mounted to /boot, and i think you're almost golden after installing and using efibootmgr
2021-07-26 11:03:20 <caskd> Aftershock: anything works as long as you can set up everything properly
2021-07-26 11:03:31 <jeffreybenzos[m]> gdisk lets you format gpt disks easily
2021-07-26 11:03:57 <caskd> jeffreybenzos[m]: oh, so you use the host bootloader
2021-07-26 11:04:00 <caskd> i see
2021-07-26 11:04:23 <caskd> that's not that portable though and if you need to switch the drive between devices it's going to be annoying
2021-07-26 11:04:41 <jeffreybenzos[m]> https://wiki.archlinux.org/title/EFISTUB#efibootmgr by the way, do not use partuuid, that's fuckin dumb and misleads you
2021-07-26 11:04:54 <jeffreybenzos[m]> use the UUID.
2021-07-26 11:05:04 <jeffreybenzos[m]> i got hung up on that shit for so long and it was having such a hard time
2021-07-26 11:05:05 <Aftershock> can I use a btrfs subvol for swap?
2021-07-26 11:05:11 <jeffreybenzos[m]> i think i even--let me grab my cmdline
2021-07-26 11:05:21 <jeffreybenzos[m]> yes
2021-07-26 11:05:24 <jeffreybenzos[m]> that's what i'm currently doing
2021-07-26 11:05:28 <caskd> Aftershock: yes you can
2021-07-26 11:05:40 <Aftershock> cool
2021-07-26 11:05:43 <jeffreybenzos[m]> see the wiki on what you have to do (set nocompression, chattr +C, etc)
2021-07-26 11:05:49 <jeffreybenzos[m]> i set all of that shit up in the live environment
2021-07-26 11:05:56 <caskd> jeffreybenzos[m]: afaik PARTUUID doesn't work on current mkinitfs, will look into it eventually
2021-07-26 11:05:57 <jeffreybenzos[m]> before rebooting
2021-07-26 11:06:02 <caskd> i just mount with LABELs
2021-07-26 11:06:05 <jeffreybenzos[m]> man fuck partuuid anyway
2021-07-26 11:06:11 <jeffreybenzos[m]> yeah i think i might have the label on there?
2021-07-26 11:06:24 <caskd> you can check with blkid
2021-07-26 11:06:46 <Aftershock> is it stabler than grub?
2021-07-26 11:07:08 <Aftershock>  /is it worth doing?
2021-07-26 11:07:11 <jeffreybenzos[m]> root=UUID=[[my rootuuid]] rw rootflags=subvol=volume/root modules=btrfs initrd=\intel-ucode.img initrd=\initramfs-lts
2021-07-26 11:07:18 <caskd> efibootmgr? Aftershock 
2021-07-26 11:07:23 <jeffreybenzos[m]> i think that actually just lets you do it without even modifying mkinitfs or whatever
2021-07-26 11:07:24 <Aftershock> yeah
2021-07-26 11:07:37 <jeffreybenzos[m]> aftershock: it boots quickly and alpine is good enough that you can fix stuff from live environments really easily
2021-07-26 11:07:44 <jeffreybenzos[m]> so i would recommend doing it, despite other people not doing it like i did
2021-07-26 11:07:52 <caskd> it's not portable though
2021-07-26 11:07:53 <jeffreybenzos[m]> i have yet to use luks to properly encrypt my disk
2021-07-26 11:07:59 <Aftershock> I'm using it as headless server 
2021-07-26 11:08:04 <jeffreybenzos[m]> but that would be the next option and then it's gonna be just fuckin RAW BTRFS ALL THE WAY
2021-07-26 11:08:05 <jeffreybenzos[m]> oh neat
2021-07-26 11:08:05 <caskd> so if you boot the drive on multiple systems, it's not going to work
2021-07-26 11:08:21 <Aftershock> and its a pain to connect a screen to
2021-07-26 11:08:22 <jeffreybenzos[m]> caskd: what do you mean?
2021-07-26 11:08:40 <caskd> you install the boot option in the local NVRAM
2021-07-26 11:08:51 <caskd> and if you go on anoter host, you won't have tht
2021-07-26 11:08:52 <jeffreybenzos[m]> psh, you just take your live disk with you
2021-07-26 11:08:58 <jeffreybenzos[m]> the media has grub :^)
2021-07-26 11:08:58 <Aftershock> yeah damn currently booting the ssd to my laptop
2021-07-26 11:09:03 <caskd> that's not what everyone wants to do
2021-07-26 11:09:11 <jeffreybenzos[m]> it's true
2021-07-26 11:09:19 <caskd> for example i have a portable HDD with my system on it
2021-07-26 11:09:25 <Aftershock> me too
2021-07-26 11:09:25 <caskd> so efibootmgr wouldn't work
2021-07-26 11:09:27 <jeffreybenzos[m]> but marrying your drive to your machine is an important security moment, for consideration here
2021-07-26 11:09:36 <jeffreybenzos[m]> a portable hdd? very interesting
2021-07-26 11:09:50 <caskd> yep, i can use anyone's hardware with my software
2021-07-26 11:09:51 <Aftershock> nvme ssd with enclosuer
2021-07-26 11:09:56 <caskd> as long as i can boot from it
2021-07-26 11:10:00 <caskd> Aftershock: exactly
2021-07-26 11:10:05 <caskd> that's what i plan to do next
2021-07-26 11:10:06 <jeffreybenzos[m]> the idea i've got is slightly more portable: putting all boot on a liveusb and then the disk is fully cray cray, but that'll be int he future
2021-07-26 11:10:14 <Aftershock> literally plug into anyones pc and yeet all their data
2021-07-26 11:10:18 <jeffreybenzos[m]> anyone's hardware? WRONK
2021-07-26 11:10:23 <jeffreybenzos[m]> anyone's x86_64, maybe
2021-07-26 11:10:38 <jeffreybenzos[m]> and even that's not true for certain things, like driver setups
2021-07-26 11:10:44 <caskd> well, ignoring arch and that kind of stuff of course
2021-07-26 11:10:47 <Aftershock> so funny when I show people that I can access all their stuff without even trying
2021-07-26 11:11:00 <jeffreybenzos[m]> that seems like a use case for just putting it on a flash drive to me
2021-07-26 11:11:04 <jeffreybenzos[m]> why a big hdd?
2021-07-26 11:11:04 <caskd> jeffreybenzos[m]: if you got the right things in the initramfs you're good to go on most hardware
2021-07-26 11:11:14 <Aftershock> ssd is quite small
2021-07-26 11:11:22 <jeffreybenzos[m]> i'm getting closer to just fucking off the initramfs altogether
2021-07-26 11:11:32 <jeffreybenzos[m]> but that's counter to the next step of just not having initramfs on the current situation
2021-07-26 11:11:37 <caskd> booting without initramfs is harder than you think
2021-07-26 11:11:48 <caskd> but go at it if you want
2021-07-26 11:11:52 <jeffreybenzos[m]> harder than cramming it all into the kernel?
2021-07-26 11:12:13 <Aftershock> wait will my SSDs UUID stay the same if I plug it into a different system
2021-07-26 11:12:23 <jeffreybenzos[m]> it's a universally unique identifier
2021-07-26 11:12:27 <caskd> yes
2021-07-26 11:12:32 <Aftershock> so it should work then?
2021-07-26 11:12:36 <caskd> yes
2021-07-26 11:12:38 <Aftershock> the efibootmgr
2021-07-26 11:12:40 <jeffreybenzos[m]> just in this universe
2021-07-26 11:12:45 <caskd> the efibootmgr?
2021-07-26 11:12:47 <Aftershock> fair
2021-07-26 11:13:02 <jeffreybenzos[m]> efibootmgr wouldn't work otherwise because efibootmgr is installing it to your local mobo's nvram with all the options involved
2021-07-26 11:13:16 <Aftershock> so I need an efi partition
2021-07-26 11:13:21 <caskd> yes, with a bootloader
2021-07-26 11:13:24 <Aftershock> and then the rest can be btrfs
2021-07-26 11:13:25 <caskd> for a portable setup
2021-07-26 11:13:27 <jeffreybenzos[m]> you will still need an efi partition with efibootmgr
2021-07-26 11:13:37 <caskd> jeffreybenzos[m]: efibootmgr is optional
2021-07-26 11:13:52 <caskd> efi can find efi partitions automatically and boot from them
2021-07-26 11:14:01 <jeffreybenzos[m]> to clarify, "WITH EFIBOOTMGR, YOU WILL STILL NEED AN EFI PARTITION"
2021-07-26 11:14:02 <caskd> just like bios
2021-07-26 11:14:06 <jeffreybenzos[m]> aye
2021-07-26 11:14:09 <Aftershock> yeah ok
2021-07-26 11:14:10 <caskd> ah okay
2021-07-26 11:14:26 <jeffreybenzos[m]> efibootmgr isn't something put into the esp, so it wouldn't make sense to read it the way you did anyway, lole
2021-07-26 11:14:31 <Aftershock> I don't run setup disk anymore then or?
2021-07-26 11:14:42 <jeffreybenzos[m]> i did not run setup-disk
2021-07-26 11:14:50 <jeffreybenzos[m]> but i did have a big ol script i hacked together instead
2021-07-26 11:15:06 <jeffreybenzos[m]> i think you can run setup-disk /mnt (after mounting your filesystems to /mnt)
2021-07-26 11:15:11 <jeffreybenzos[m]> but that does grub as the installer
2021-07-26 11:15:28 <jeffreybenzos[m]> so instead, i can go grab that shitty script that isn't even complete, if you want it
2021-07-26 11:15:49 <Aftershock> I just want to boot to alpine on the ssd lol
2021-07-26 11:16:10 <Aftershock> anything that can make that happen with btrfs is a godsend at this point
2021-07-26 11:16:51 <jeffreybenzos[m]> if that's all you care about, then you can, if i recall correctly, set everything up how you want it
2021-07-26 11:16:56 <jeffreybenzos[m]> then run setup-disk /mnt
2021-07-26 11:16:56 <caskd> setup-disks sets up the common install with ext4 root, swap and all and a bootloader
2021-07-26 11:17:05 <jeffreybenzos[m]> with everything you've made (subvols etc) on /mnt
2021-07-26 11:17:10 <caskd> so, for btrfs you will need a manual setup
2021-07-26 11:17:21 <jeffreybenzos[m]> caskd: not when it calls install_mount_whatever; i'm PRETTY SURE it supports btrfs
2021-07-26 11:17:27 <jeffreybenzos[m]> but that was a few days ago when i was reading through
2021-07-26 11:17:29 <caskd> it does?
2021-07-26 11:17:30 <jeffreybenzos[m]> to the script
2021-07-26 11:17:53 <caskd> i did it manually, so at least i can say the manual way works
2021-07-26 11:17:55 <caskd> ^^'
2021-07-26 11:17:58 <Aftershock> yeah it seemed to insall without complaining to a /mnt formatted to btrfs
2021-07-26 11:18:16 <jeffreybenzos[m]> did you mount esp to /boot
2021-07-26 11:18:24 <Aftershock> yeah
2021-07-26 11:18:32 <Aftershock> to /mnt/boot
2021-07-26 11:18:38 <jeffreybenzos[m]> what's your fstab look like
2021-07-26 11:18:44 <jeffreybenzos[m]> are you doing a nested subvol situation or a flat one
2021-07-26 11:18:45 <Aftershock> it seemed fine
2021-07-26 11:18:51 <Aftershock> mix
2021-07-26 11:19:00 <Aftershock> top level is flat
2021-07-26 11:19:00 <jeffreybenzos[m]> how dare you
2021-07-26 11:19:00 <jeffreybenzos[m]> alright
2021-07-26 11:19:08 <jeffreybenzos[m]> well boot the fuckin thing and tell me if it worked
2021-07-26 11:19:13 <Aftershock> but I hvae a root partition
2021-07-26 11:19:26 <Aftershock> and a top level home partition that mounts under that
2021-07-26 11:19:38 <caskd> top level flat is the easiest and allows very flexible changes
2021-07-26 11:19:39 <Aftershock> maybe thats just normal flat idk
2021-07-26 11:20:01 <caskd> i have both root and home as subvol of subvolid=5
2021-07-26 11:20:04 <Aftershock> but with a root subvol for the /
2021-07-26 11:20:20 <Aftershock> yeah
2021-07-26 11:21:23 <Aftershock> it did boot to grub btw
2021-07-26 11:21:38 <Aftershock> just grub was saying unable to read filesystem
2021-07-26 11:22:10 <jeffreybenzos[m]> alpine is now a btrfs-first distribution
2021-07-26 12:15:33 <jeffreybenzos[m]> hmm; fzf autocompletion in bash isn't working (the hooks for ctrl-t, alt-c, and the **<tab> situation aren't, at east).
2021-07-26 12:16:16 <jeffreybenzos[m]> does bash not source everything in /etc/profile.d?
2021-07-26 12:17:38 <jeffreybenzos[m]> oh, only when started as a login shell.
2021-07-26 12:42:16 <lopid> the only system file bash looks at is /etc/profile, and yes, only when as login
2021-07-26 13:43:13 <bl4ckb0ne> are the kernel panic logs stored somewhere?
2021-07-26 13:52:53 <Hello71> on the screen /s
2021-07-26 13:53:32 <caskd> or the RAM
2021-07-26 13:53:33 <yyp> They may be written to a serial port as well
2021-07-26 13:53:53 <caskd> i remember trying to recover kernel panic logs from the ring buffer in ram once
2021-07-26 13:53:56 <caskd> :P
2021-07-26 14:22:56 <Spontaneous> Hi guys
2021-07-26 14:43:59 <bl4ckb0ne> ended up reinstalling alpine
2021-07-26 14:47:07 <durrendal> that's a feeling bl4ckb0ne, there's nothing quite the same 
2021-07-26 14:47:42 <bl4ckb0ne> true, this one boots fine
2021-07-26 17:48:27 <Ditronics> hi guys
2021-07-26 17:48:37 <Ditronics> hope everyone is doing ok
2021-07-26 17:49:19 <Ditronics> is there a way to buil a custom module or to copy one already done to the right place in a diskless system?
2021-07-26 17:49:26 <Ditronics> v3.14
2021-07-26 17:50:15 <ikke> Ditronics: you can set up an overlay for modloop
2021-07-26 17:50:38 <Ditronics> i'm searching for that info for days
2021-07-26 17:50:49 <ikke> check /etc/conf.d/modloop
2021-07-26 17:51:05 <Ditronics> aha
2021-07-26 17:51:17 <Ditronics> any link with info avaiable
2021-07-26 17:51:19 <Ditronics> ?
2021-07-26 17:51:56 <ikke> No, that's all I have for you
2021-07-26 17:52:01 <ikke> Just set it to something non-zero
2021-07-26 17:52:18 <ikke> well, a particular size, ofcourse
2021-07-26 17:52:40 <Ditronics> ok, i'll try
2021-07-26 17:52:49 <Ditronics> and thanks !!
2021-07-26 17:53:08 <ikke> The comment in that file is a bit misleading, as it implicates that a size of 0 is valid
2021-07-26 17:54:02 <Ditronics> I see
2021-07-26 17:54:16 <Ditronics> now it is 0 already 
2021-07-26 17:54:31 <ikke> yes
2021-07-26 17:54:37 <Ditronics> the size must to be any number or a particular one?
2021-07-26 17:54:43 <Ditronics> the ram size ?
2021-07-26 17:55:13 <ikke> The size of the overlayfs, and as it's tmpfs, it will be taken from ram
2021-07-26 17:56:02 <Ditronics> aha
2021-07-26 17:56:46 <Ditronics> so I set it, then reboots and there will be a space to add the module?
2021-07-26 17:57:14 <ikke> yeah, you should be able to install it into /lib/module/*
2021-07-26 17:57:41 <Ditronics> it must be expressed on what, mb, kb, gb?
2021-07-26 17:58:01 <ikke> whatever mount -o size= accepts
2021-07-26 17:59:07 <Ditronics> I see, let me try !
2021-07-26 18:14:55 <ikke> jeffreybenzos[m]: just running darktable in my VM, no segfault
2021-07-26 18:22:03 <ikke> jeffreybenzos[m]: n/m I was not on edge, on edge it fails to start for me
2021-07-26 19:38:36 <gopher22> Hello. I was randomly searching for updates on one of my Alpine Linux containers. I happened to use traefik a reverse proxy, which I have installed unsing Alpine edge repo. There is no update for it at the moment as it seems, however I noticed that if doing a quick search for *traefik* on pkgs.alpinelinux.org yields no results. Has Traefik been
2021-07-26 19:38:36 <gopher22> removed from the edge repositories? 
2021-07-26 19:40:17 <ikke> Apparently it has been moved to unmaintained
2021-07-26 19:40:45 <ikke> hmm
2021-07-26 19:41:05 <iggy> was going to say the same https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/unmaintained/traefik
2021-07-26 19:42:17 <gopher22> hmm ok, that might answer that. how does "unmaintaned" status work for Alpine in brief? Is there a chance for a package landing back to edge/testing in the near future (due to reasons of course)?
2021-07-26 19:42:35 <ikke> gopher22: it can if someone is willing to maintain i
2021-07-26 19:42:37 <ikke> it
2021-07-26 19:42:44 <ikke> I'm considering it
2021-07-26 19:42:51 <gopher22> I see, thank you
2021-07-26 19:44:47 <gopher22> I think one of the main strenghts of Alpine was the abundance of the availability of "state of the art" software, like Traefik, for ex. But of course it is understandable that behind these privileges is all manpower....
2021-07-26 19:45:17 <ikke> Yup, someone has to do the work to maintain it
2021-07-26 19:46:32 <gopher22> nevertheless, I do hope that these packages won't go down /dev/null and someone would pick them up to mend them, if not the original developer. I wish I could help, but I'm not a dev either unfortunately
2021-07-26 19:47:13 <ikke> checking if it builds atm
2021-07-26 19:48:10 <iggy> gopher22: you don't have to be a dev to maintain help out with packages
2021-07-26 19:48:18 <iggy> just use it and submit MRs
2021-07-26 19:48:53 <gopher22> I know... but still, I think one should at least have a minimal dev 'mindset' affinity
2021-07-26 19:52:51 <gopher22> I recently started to "Alpinze" all my LXC containers, previously/currently using Debian/Buster, but moving on making these more lightweight. I don't really like Docker, LXCs are the way for me, at least. Alpine is the OS of choiche at the moment, not only for obvious reasons known :) but also for the wealth of available packages under edge/testing
2021-07-26 19:54:33 <ikke> ugh, anoying, the release tarball of traefik has all files in the root
2021-07-26 19:54:34 <invoked> many people have started out just helping with packages before they knew anything about dev mindset. you have to start somewhere :)
2021-07-26 19:54:59 <gopher22> invoked agreed :)
2021-07-26 19:55:42 <invoked> the only way you learn anyhow is to just submit things, get feedback on what you submitted, et al
2021-07-26 19:56:00 <gopher22> ikke should be under /etc/traefik for configuration at least. that is where traefik looks for the "root
2021-07-26 19:56:14 <ikke> I mean the source archive
2021-07-26 19:56:14 <gopher22> or yaml static config files
2021-07-26 19:56:59 <ikke> source archives usually have the files in a subdirectory
2021-07-26 19:57:07 <ikke> So when you unpack it, it does not clobber you CWD
2021-07-26 19:57:21 <ikke> PWD*
2021-07-26 19:57:31 <gopher22> ah yes yes, understood
2021-07-27 02:29:05 <geoman> hey i tried installing nvidia drivers through the alpine package and a lot of packages got purged, now wifi doesn't work
2021-07-27 02:29:28 <geoman> when i try "ip link set wlan0 up" it comes up as ioctl 0x8913 failed:no such device
2021-07-27 02:30:17 <geoman> how would i get the wifi drivers for my laptop, it was automatically installed over the network before
2021-07-27 02:33:02 <Hello71> >installing nvidia drivers through the alpine package
2021-07-27 02:33:29 <geoman> it was the linux-firmware-nvidia package
2021-07-27 02:39:58 <geoman> is there a way to install the iwlwifi firmware from apk
2021-07-27 02:43:50 <caskd> iwlwifi firmware is located in linux-firmware-other
2021-07-27 02:44:19 <caskd> but you should try to recover whatever package was deleted
2021-07-27 02:44:32 <caskd> you might be missing more than just wifi
2021-07-27 02:45:14 <geoman> how would i see the packages that were purged?
2021-07-27 02:45:26 <geoman> does pkgs.alpinelinux.org show that?
2021-07-27 02:45:34 <geoman> at least possible ones?
2021-07-27 02:46:06 <caskd> if you have only one package which you removed/installed, it would show whatever conflicts happened
2021-07-27 02:46:12 <caskd> but for more, not really
2021-07-27 02:49:42 <geoman> ok so dmesg says that only the regulatory and wifi drivers are missing
2021-07-27 02:50:45 <geoman> do i have to restart for the iwlwifi driver to load?
2021-07-27 02:51:49 <caskd> you can also modprobe and restart mdev
2021-07-27 03:00:51 <geoman> how exactly would i restart mdev? i tried looking at the gentoo wiki and the mdev help page, along with the openrc man page
2021-07-27 03:00:59 <geoman> and i don't see anything about restarting services
2021-07-27 03:02:01 <geoman> do i just do openrc shutdown mdev and then openrc init mdev?
2021-07-27 03:08:43 <c705> /etc/init.d/mdev stop/start/restart
2021-07-27 03:13:51 <geoman> that didn't help, so i just decided to clone the linux-firmware tree and placed it in /lib/firmware
2021-07-27 03:31:17 <Hello71> ...
2021-07-27 03:31:43 <Hello71> "eh, let's just download this random crap"
2021-07-27 03:31:51 <Hello71> "what are 'updates' anyways"
2021-07-27 03:50:53 <jeffreybenzos[m]> Anyone on here using steam on alpine?
2021-07-27 03:52:50 <jeffreybenzos[m]> luckily, lopid, it runs $script for everything in profile.d named .sh
2021-07-27 03:52:52 <jeffreybenzos[m]> so i'll fix that
2021-07-27 03:53:02 <jeffreybenzos[m]> ikke: yeah, same. thanks for the traceback; maybe I can bug their devs about it? no clue.
2021-07-27 04:10:45 <jeffreybenzos[m]> Looks like fzf-bash-completion isn't actually causing anything to be completed on my setup. Am I missing something I should be doing?
2021-07-27 04:21:21 <ikke> jeffreybenzos[m]: I've been trying to debug it, but for some reason, it does not crash when I built in from source from their repo
2021-07-27 04:22:05 <jeffreybenzos[m]> lmao
2021-07-27 04:22:21 <jeffreybenzos[m]> well, i guess i could do that for the time being; is there an easy way to reversibly set that up?
2021-07-27 04:22:25 <ikke> so trying to figure out what's going on
2021-07-27 04:22:27 <jeffreybenzos[m]> (i've just got some photos i'd like to throw through the grinder)
2021-07-27 04:23:04 <ikke> do you need 3.6.0?
2021-07-27 04:23:14 <ikke> What worked for me was installing 3.4.1.1 from v3.14
2021-07-27 04:23:17 <ikke> at least, for now
2021-07-27 04:25:21 <jeffreybenzos[m]> 3.6 has their new color balance rgb module, which is key to a workflow of me getting better faster
2021-07-27 04:25:27 <jeffreybenzos[m]> but i can always fall back
2021-07-27 04:25:59 <ikke> ok
2021-07-27 04:26:20 <jeffreybenzos[m]> there's no way for me to just pull in a single package from 3.14 like that, right?
2021-07-27 04:27:00 <ikke> apk -X https://dl-cdn.alpinelinux.org/alpine/v3.14/community add darktable=3.4.1.1-r0
2021-07-27 04:28:35 <jeffreybenzos[m]> oh, very interesting.
2021-07-27 04:30:57 <ikke> note that this is not really supported in the long run, as dependencies might change
2021-07-27 04:31:10 <ikke> but for now, edge and v3.14 are close enough
2021-07-27 04:32:07 <ikke> for building it from source, I cloned both aports and the darktable repository (the latter with --recursive to also get submodules)
2021-07-27 04:32:41 <ikke> apk add alpine-sdk, to get the needed tools
2021-07-27 04:33:16 <ikke> in aports/community/darktable, I ran abuild deps, so that all the required dependencies are installed
2021-07-27 04:34:06 <ikke> in that aports repo are also two patches that are required
2021-07-27 04:36:54 <jeffreybenzos[m]> mind is breaking open thinking that the build system just does all that for you
2021-07-27 04:48:16 <ikke> One thing I'm thinking is that it might have to do with some of the build flags that abuild sets that are not set when I build it manually
2021-07-27 04:48:37 <ikke> So I could try to repeat that
2021-07-27 05:36:23 <Aftershock> hey so I installed alpine on an intel cpu machine, is there anything I need to do when switching the cpu to amd?
2021-07-27 05:37:28 <Aftershock> <jeffreybenzos[m]> I finally got it to boot in btrfs
2021-07-27 05:38:07 <Aftershock> just can't boot when its in the machine
2021-07-27 05:58:54 <jeffreybenzos[m]> lmao; what?
2021-07-27 05:58:56 <jeffreybenzos[m]> and yes
2021-07-27 05:59:12 <jeffreybenzos[m]> you should use the amd-ucode.img instead of the intel-ucode.img in your kernel params
2021-07-27 05:59:22 <jeffreybenzos[m]> initrd=\amd-ucode.img
2021-07-27 06:14:58 <Aftershock> <jeffreybenzos[m]>I'm still using grub, is there somewhere I can change the micro code for grub?
2021-07-27 06:15:14 <Aftershock> I know you don't use it so
2021-07-27 06:15:21 <Aftershock> might not be able to help
2021-07-27 06:15:30 <jeffreybenzos[m]> it'd be changing the kernel line IN grub, but it might not even be in there
2021-07-27 06:16:01 <Aftershock> should it still boot if it was wrong?
2021-07-27 06:16:08 <Aftershock> or if it was missing?
2021-07-27 06:16:28 <Aftershock> I don't have a screen on the server I'm putting it in
2021-07-27 06:16:52 <jeffreybenzos[m]> you can check cat /proc/cmdline to see what your kernel stuff was
2021-07-27 06:17:01 <jeffreybenzos[m]> i think it should
2021-07-27 06:17:58 <Aftershock> does alpine automatically use eth0 when available if it wasn't there during install?
2021-07-27 06:19:08 <jeffreybenzos[m]> that was part of setup-interfaces
2021-07-27 06:19:17 <jeffreybenzos[m]> check /etc/networking/interfaces
2021-07-27 06:19:27 <jeffreybenzos[m]> my bad
2021-07-27 06:20:09 <Aftershock> it says:
2021-07-27 06:20:59 <Aftershock> auto lo; iface lo inet loopback ; ; auto wlan0 inet dhcp ;
2021-07-27 06:21:05 <Aftershock> ; is newline
2021-07-27 06:21:49 <Aftershock> so I'll need to setup an ethernet interface?
2021-07-27 06:22:27 <Aftershock> yeah cmdline doesn't list any microcode
2021-07-27 06:23:43 <jeffreybenzos[m]> yeah, you'd need to set up something simple like
2021-07-27 06:24:03 <jeffreybenzos[m]> same concept, auto wlan0 inet dhcp
2021-07-27 06:42:44 <lopid> no, it just runs /etc/profile
2021-07-27 06:43:28 <Aftershock> can I copy the wlan0 configuration for eth0 (replacing wlan0 with eth0)
2021-07-28 20:26:53 <captainchris> hi everybody 
2021-07-28 20:31:14 <lopid> www.nohello.com
2021-07-28 20:37:59 <atka> hellow
2021-07-28 20:38:34 <atka> s/hellow/hello
2021-07-28 20:38:34 <alpine-meetbot> atka meant to say: hello
2021-07-28 21:16:13 <jeffreybenzos[m]> lopid: shut up.
2021-07-28 21:16:25 <jeffreybenzos[m]> that made sense when channels were bustling
2021-07-28 21:16:30 <jeffreybenzos[m]> this is a completely different era
2021-07-28 21:16:38 <jeffreybenzos[m]> how's life, captainchris?
2021-07-28 21:17:27 <jeffreybenzos[m]> 2013 was such a different time on irc that posts like that increased productivity (stupidly, at the sacrifice of humanity), and resulted in what we have now.
2021-07-28 21:18:35 <trylletre> ..what we have now?
2021-07-28 21:19:05 <captainchris> Hey Jeffrey 
2021-07-28 21:19:48 <captainchris> What do you think of alpine linux 
2021-07-28 21:19:50 <captainchris> ??
2021-07-28 21:19:56 <onelegend> Alpine Linux is great
2021-07-28 21:20:32 <captainchris> I'm under void linux and I hesitate to go under alpine for minimal linux distribution 
2021-07-28 21:20:45 <invoked> if only there was a way to have a pandemic without anybody dying. i've never been more productive or had a better diet
2021-07-28 21:21:09 <atka> alpine is my go to distro for most things, server, laptop etc. I used to be on void but alpine is like a better void to me.
2021-07-28 21:23:53 <captainchris> I 'm loving minimal software like suckless tools.
2021-07-28 21:24:05 <captainchris> And i want minimal distribution 
2021-07-28 21:24:23 <captainchris> I hesistate between sta.li openbsd and alpine 
2021-07-28 21:30:35 <c705> did you have a question or what?
2021-07-28 21:32:53 <invoked> openbsd has been around longer and is arguably still more opinionated
2021-07-28 21:33:50 <invoked> there is no upstream for openbsd
2021-07-28 21:34:37 <invoked> 
ACTION stops polluting the channel with that

2021-07-28 21:38:20 <jeffreybenzos[m]> c705: just have a goddamned conversation, holy shit
2021-07-28 21:38:22 <jeffreybenzos[m]> if you don't want to participate then just fuck off already
2021-07-28 21:38:29 <jeffreybenzos[m]> my god
2021-07-28 21:38:34 <jeffreybenzos[m]> captainchris: sorry, i've been cooking some chili, so i was distracted
2021-07-28 21:38:42 <jeffreybenzos[m]> i love alpine, mostly for similar reasons
2021-07-28 21:40:16 <jeffreybenzos[m]> sta.li seems kind of interesting. one thing that i don't think suckless captures well is the orchestration of these tools into larger projects, and that's what I want to see more of; that's why places like alpine's chats are useful
2021-07-28 21:40:42 <jeffreybenzos[m]> chris: what hardware are you running?
2021-07-28 21:40:47 <jeffreybenzos[m]> openbsd is fun
2021-07-28 21:40:56 <jeffreybenzos[m]> do you care about using a browser or any gui tools?
2021-07-28 21:42:09 <captainchris> A old intel i5 proco with asrock amdgpu and 8 gb of ram 
2021-07-28 21:45:30 <jeffreybenzos[m]> openbsd might support the former, but i don't know about your graphics
2021-07-28 21:45:40 <jeffreybenzos[m]> i have amdgpu running fine on alpine, though i haven't tried opencl yet
2021-07-28 21:45:44 <jeffreybenzos[m]> i hear there's some process available
2021-07-28 21:46:19 <jeffreybenzos[m]> invoked: it's been nice until more of my friends became busier than they were even pre-pandemic.
2021-07-28 21:46:19 <jeffreybenzos[m]> work is insane for some
2021-07-28 21:46:26 <jeffreybenzos[m]> seems like the bosses are just running people to make up for lost time
2021-07-28 21:57:07 <invoked> i'll respond -> #alpine-offtopic 
2021-07-28 21:57:25 <jeffreybenzos[m]> that channel should be nuked
2021-07-28 21:57:30 <jeffreybenzos[m]> this is not 2012; there are only 300 people here
2021-07-28 21:57:43 <jeffreybenzos[m]> no one uses these logs incompetently enough to lose track of a conversation
2021-07-28 21:58:30 <invoked> you can stop using it anytime.
2021-07-28 21:59:10 <jeffreybenzos[m]> splitting these things up is ludicrous when there is no one active
2021-07-28 21:59:26 <jeffreybenzos[m]> concentrate conversation so it actually happens once in a while or you'll face the same burnout everyone else faces
2021-07-28 21:59:30 <onelegend> true
2021-07-28 22:00:47 <invoked> if i told you there are mailing lists, you'd probably complain those aren't active enough
2021-07-28 22:01:19 <jeffreybenzos[m]> nope; i'm complaining equivalently that you're splintering one mailing list into 80, and also complaining that mailing lists aren't at all equivalent to instant flippant chat
2021-07-28 22:01:34 <invoked> have you looked in the mirror lately
2021-07-28 22:09:42 <jeffreybenzos[m]> oh, i should be able to see good conversation with other meaningful human beings in the mirror?
2021-07-28 22:09:48 <jeffreybenzos[m]> cmon, man; how does that even compute
2021-07-28 22:10:01 <c705> jeffreybenzos[m]: you should log off and go outside or something
2021-07-28 22:11:04 <jeffreybenzos[m]> i've been between cooking chili and being outside this entire time; these reflexes of yours are showing heavily that you don't know how to connect with someone, and you just rattle off a list of support-like answers instead of engaging with them
2021-07-28 22:11:12 <jeffreybenzos[m]> literally, guys, come the fuck on, get it together
2021-07-28 22:11:52 <c705> read the scrollback and tell me you're not just describing yourself
2021-07-28 22:12:10 <jeffreybenzos[m]> that's a deflective measure and nothing more. it doesn't invalidate my point whatsoever.
2021-07-28 22:12:16 <trylletre> jeffreybenzos[m]: i dont necessarily disagree with you but i find it quite funny your name is jeffrey benzodiazepines
2021-07-28 22:12:27 <jeffreybenzos[m]> trylletre: likewise
2021-07-28 22:12:42 <jeffreybenzos[m]> nice. matrix didn't send this line, c705: that's a deflective measure and nothing more. it doesn't invalidate my point whatsoever.
2021-07-28 22:12:51 <trylletre> it did send
2021-07-28 22:12:55 <c705> this is either someone who doesn't have a frontal cortext or the worlds worst troll. anyways, i really don't give a shit. bye
2021-07-28 22:13:29 <jeffreybenzos[m]> trylletre: ah, matrix just sucks, lol
2021-07-28 22:13:54 <atka> well matrix just got 30m so maybe it'll get better!
2021-07-28 22:13:56 <trylletre> never used it myself
2021-07-28 22:14:04 <onelegend> matrix is cool
2021-07-28 22:14:07 <onelegend> I like it
2021-07-28 22:15:13 <jeffreybenzos[m]> onelegend: it works fine as a chat program, but element and the irc bridges are a bit bogged
2021-07-28 22:16:46 <onelegend> there's this dope client called cinny
2021-07-28 22:16:53 <onelegend> it's still a work in progress but it's really ncie
2021-07-28 22:17:01 <onelegend> https://cinny.in/
2021-07-28 22:22:26 <jeffreybenzos[m]> oh hell yeah, onelegend
2021-07-28 23:06:24 <fluix> this cam out 10 hours ago?
2021-07-28 23:06:27 <fluix> s/cam/came
2021-07-28 23:06:27 <alpine-meetbot> fluix meant to say: this came out 10 hours ago?
2021-07-28 23:13:36 <jeffreybenzos[m]> I'm in.
2021-07-28 23:13:41 <jeffreybenzos[m]> I wonder how much better it is. I like the interface.
2021-07-29 09:28:10 <LACampbell> I got another human being using alpine. soon we'll be as big as ubuntu
2021-07-29 09:31:31 <jeffreybenzos[m]> ikke: did you make a change to the dt package? i see there's an -r1 now?
2021-07-29 09:31:47 <ikke> yes, but it was merged before I verified it actually worked
2021-07-29 09:34:27 <tankf33der> Hi all, i spent many dayz to setup wireguard between alpine and iphone. 
2021-07-29 09:35:01 <jeffreybenzos[m]> lol; it seems to be crashing for me.
2021-07-29 09:35:35 <ikke> Yes, I just set CXXFLAGS, but I suppose I should be setting other relevant flags as well
2021-07-29 09:36:05 <jeffreybenzos[m]> the darktable guys weren't as impressed as i was that -Os alone shrinks the binary by like a billion times
2021-07-29 09:36:20 <jeffreybenzos[m]> i was wondering why stuff on alpine downloaded and installed so insanely quickly; i think i have my answer
2021-07-29 09:36:32 <ikke> Well, what also helps is that we strip debug symbols
2021-07-29 09:36:57 <ikke> https://pkgs.alpinelinux.org/package/edge/community/x86_64/darktable-dbg
2021-07-29 09:37:11 <ikke> 76MB for the symbols
2021-07-29 09:37:40 <tankf33der> I failed again and again, i wonder if somebody have working wireguard configs somewhere. 
2021-07-29 09:43:39 <Skaface82> i'm loving this, since playing with alpine, i now have my desktop system on a USB stick that loads into ram which I can boot my home and work pc's from
2021-07-29 09:43:56 <Skaface82> no more setting up and maintaining separate systems
2021-07-29 09:44:03 <Skaface82> so good
2021-07-29 09:45:06 <jeffreybenzos[m]> ikke: how are the symbols literally so big
2021-07-29 09:48:05 <jeffreybenzos[m]> skaface82: how big's your install? it's really nice that the system and all of the programs i actually care about are smaller than a windows install disk, before decompression
2021-07-29 09:49:20 <Skaface82> 1.2Gb
2021-07-29 09:53:08 <jeffreybenzos[m]> Nice; a good drive would take like 10-20 to load all of that in RAM anyway.
2021-07-29 09:54:34 <Skaface82> its given me a new way of thinking really, like if i need to use say gimp, i just apk add gimp, do what i need to do, and next reboot its gone. keeps the system so clean
2021-07-29 09:56:13 <Skaface82> and the packages are cached on the usb stick for next time i need to use it
2021-07-29 10:01:03 <jeffreybenzos[m]> It's why I wanted to do some sort of situation where I pull an image off of disk into RAM, only operate there, and then flush to disk if I ever invoke any changes I actually want to commit.
2021-07-29 10:02:22 <Skaface82> it makes so much sense
2021-07-29 10:04:36 <Skaface82> i was trying to get a network PXE boot thing going, but I gave up as I couldn't find a way to stop it from sourcing the packages from an internet mirror on each boot
2021-07-29 10:05:15 <Skaface82> but now i have a few usb sticks, and a script that backs up/restores the apkovl.tar.gz to my NAS so I can easily keep my sticks in sync
2021-07-29 10:08:42 <Skaface82> my next project will be my virtual machine server, I think ill make a USB alpine image for that so that the hypervisor system just runs from ram, keeping the SSD drive just for operating system images
2021-07-29 19:09:55 <meph> hey uhh is there an easy way to install alpine on an x86_64 system over a serial console?
2021-07-29 19:10:17 <meph> it doesn't look like grub is configured to use ttyS0 in the default image
2021-07-29 19:10:21 <ikke> https://wiki.alpinelinux.org/wiki/Enable_Serial_Console_on_Boot
2021-07-29 19:10:39 <meph> this system is entirely headless
2021-07-29 19:10:45 <meph> no vga output
2021-07-29 19:10:56 <meph> it just has a serial port
2021-07-29 19:11:09 <ikke> I've installed a system over serial recently
2021-07-29 19:11:20 <ikke> You do need to change the boot medium, though
2021-07-29 19:11:32 <meph> do I need to rebuild an iso with the correct grub config?
2021-07-29 19:11:42 <meph> i mean, i obviously need to get those params in there somehow
2021-07-29 19:12:35 <meph> ikke: any suggestions re: boot medium?
2021-07-29 19:13:57 <ikke> How do you use the ISO?
2021-07-29 19:14:57 <ikke> As an USB you can just modify the contents (but not if it's isofs)
2021-07-29 19:15:58 <meph> hmmm
2021-07-29 19:16:33 <meph> i dd'd the extended iso to a usb drive
2021-07-29 19:16:52 <meph> though I just popped open the back of my t440p and it looks like I have a free minipcie slot
2021-07-29 19:16:56 <ikke> if you have an alpine system, you can use setup-bootable
2021-07-29 19:17:13 <meph> I wonder if I can just shove the msata drive in there and dump a rootfs on there
2021-07-29 19:17:22 <ikke> then you get a bootable and writable usb disk
2021-07-29 19:17:45 <meph> ikke: the files i need to modify are on a iso9660 filesystem
2021-07-29 19:17:52 <meph> i thought that was inherantly RO
2021-07-29 19:18:01 <ikke> yes, that's why I suggested using setup-bootable
2021-07-29 19:18:06 <ikke> which copies the files from the iso
2021-07-29 19:18:11 <meph> ah
2021-07-29 19:18:23 <meph> im on adelie right now
2021-07-29 19:18:36 <meph> i'll just set up an alpine chroot for this
2021-07-29 19:18:49 <ikke> It's part of alpine-conf
2021-07-29 19:18:51 <meph> (this is for a pcengines apu3 btw)
2021-07-29 19:19:00 <meph> ty ikke
2021-07-29 19:19:09 <meph> I appreciate the advice
2021-07-29 19:19:58 <ikke> then you can change the boot config to enable the serial device
2021-07-29 19:20:19 <ikke> But I think what I did is to manually change the kernel cmdline on boot
2021-07-29 19:20:21 <ikke> through serial
2021-07-29 19:22:03 <ikke> So I did not even modify the boot medium
2021-07-29 19:23:51 <meph> i do actually get an isolinux prompt
2021-07-29 19:24:04 <ikke> right
2021-07-29 19:24:06 <meph> oh hmm
2021-07-29 19:24:15 <ikke> If you press tab, you can edit the cmdline
2021-07-29 19:24:20 <meph> why's there a grub config on the boot medium?
2021-07-29 19:24:29 <meph> er, the iso rathet
2021-07-29 19:24:32 <meph> rather*
2021-07-29 19:34:13 <meph> %n on %B%F{white}%m%f%b %~ %B%F{green}%#%f%b setup-bootable alpine-extended-3.14.0-x86_64.iso /media/sdb1
2021-07-29 19:34:13 <meph> cat: can't open '/media/sdb1/.new/.alpine-release': No such file or directory
2021-07-29 19:34:18 <meph> ok lol
2021-07-29 19:35:06 <meph> doesnt look like it's actually extracting the iso
2021-07-29 19:39:30 <meph> oh it does, it's just that the current release iso's don't contain a .alpine-release file
2021-07-29 19:39:36 <meph> that seems like a possible regression
2021-07-29 19:42:07 <ikke> oh, interesting
2021-07-29 19:46:38 <meph> yeah this script is entirely broken without that file
2021-07-29 19:46:58 <meph> if i try to add it beforehand to .new the script will remove .new when it starts up
2021-07-29 19:47:22 <meph> it | die's out on failure when trying to copy the file
2021-07-29 19:48:16 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimage.sh#L169
2021-07-29 19:49:45 <ikke> jeffreybenzos[m]: pushed a new version that should actually work
2021-07-29 19:49:52 <ikke> Tested it in my VM
2021-07-29 19:54:13 <meph> ikke: plot twist: .alpine-release is in the actual iso, and I can see it if I loopmount the iso, although it doesn't look like uniso is extracting it
2021-07-29 19:55:18 <meph> uniso is what the setup-bootable script is using
2021-07-29 19:56:04 <ikke> hmm, that's also part of alpine-conf
2021-07-29 19:56:25 <meph> how do i check what package provides a file again?
2021-07-29 19:56:40 <ikke> https://pkgs.alpinelinux.org/contents?file=uniso&path=&name=&branch=edge
2021-07-29 19:57:00 <meph> ok yup
2021-07-29 19:57:07 <ikke> There is also a 3rd party tool that provides a command line tool for that
2021-07-29 19:58:05 <meph> im gonna do something sketchy
2021-07-29 19:58:38 <meph> ok nvm this isn't flexible at all
2021-07-29 19:59:03 <meph> in the script it does something like:          (cd "$destdir"/.new; exec ${UNISO:-uniso}) < "$src" 
2021-07-29 19:59:16 <meph> this should probably just loopmount the iso
2021-07-29 19:59:41 <meph> i can't really shim a loopmount into $UNISO based on how that's written
2021-07-29 20:00:14 <meph> tl;dr: this script appears to be hosed
2021-07-29 20:00:39 <meph> im just gonna try and see if I can get isolinux to give me a way to edit the kernel args
2021-07-29 20:00:42 <meph> brb
2021-07-29 20:01:09 <ikke> you can always just do it manually
2021-07-29 20:01:25 <meph> yeah that'll be last ditch effort
2021-07-29 20:01:31 <ikke> extract the iso, copy the files
2021-07-29 20:01:46 <ikke> install syslinux
2021-07-29 20:02:14 <meph> it's just a lot of work for a param which I think should probably be default in the installer
2021-07-29 20:03:05 <meph> I can't picture a scenario where sending stuff out a serial port on system install would be an issue
2021-07-29 20:03:50 <meph> I guess it could screw with industrial control system stuff if someone tries to do some maintanence in field
2021-07-29 20:04:10 <meph> or print out a few pages of bullshit on a serial printer if one happens to be attached
2021-07-29 20:05:05 <meph> unless of course syslinux or the kernel will complain and die if no serial port is found. I'm not actually sure about that.
2021-07-29 20:06:12 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues/9324
2021-07-29 20:06:46 <meph> lol
2021-07-29 20:07:04 <meph> so im suffering because some john doe couldn't handle his log being blown up
2021-07-29 20:07:41 <ikke> Read the message from ncopa
2021-07-29 20:08:23 <meph> yeah, ncopa's reply implies that it's intentional and possibly should still be that way
2021-07-29 20:08:39 <meph> is this ONLY on the virt image though?
2021-07-29 20:09:22 <ikke> On the virt image, it _is_ enabled
2021-07-29 20:09:29 <ikke> at least, according to that post
2021-07-29 20:10:22 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh#L80
2021-07-29 20:10:32 <ikke> And then there is always the question of what serial device to use
2021-07-29 20:10:43 <ikke> Is it always going to be ttyS0?
2021-07-29 20:20:39 <meph> ya on these devices
2021-07-29 20:21:04 <ikke> I mean, in general
2021-07-29 20:21:27 <meph> i mean, in my experience when I've wanted to do installs via serial it has been
2021-07-29 20:22:05 <meph> most BMC's will pass through ttyS0 to the bmc itself, so you can do everything from the bmc management console
2021-07-29 20:22:27 <meph> though they also usually have VGA support so it's not like necessarily needed in that case
2021-07-29 20:22:44 <meph> lots of hypervisors will also watch ttyS0 for kernel logs and stuff
2021-07-29 20:23:05 <meph> which is almost definitely why it's enabled in the virt image
2021-07-29 20:23:15 <meph> anyways, thanks for the help, I'll give the virt image a shot
2021-07-29 20:28:59 <captainchris> Hi everybody 
2021-07-29 20:29:53 <captainchris> I have a little problem with app like qjackctl tuxguitar vimb qutebrowser. I have a big font ugly and I don't know how to change it 
2021-07-29 20:30:31 <captainchris> I have install font update my fc-cache but always the same font 
2021-07-29 20:32:09 <ikke> Just installing a font does not cause it to be used
2021-07-29 20:32:16 <ikke> You need to set it somewhere
2021-07-29 20:32:23 <ikke> where that is depends on your environment
2021-07-29 20:32:30 <meph> ikke: lmao so it's booted and im in
2021-07-29 20:32:58 <meph> this needs intel igb drivers
2021-07-29 20:33:41 <meph> the virt image basically only has drivers for virtio, hyperv, amazon ena, intel e1000 and some intel virtual functions
2021-07-29 20:33:56 <meph> sooooo... it booted but I have no network to install from :D
2021-07-29 20:34:06 <ikke> yes, that's the ideaeh
2021-07-29 20:34:14 <ikke> idea behind the virt image
2021-07-29 20:34:23 <ikke> it's not meant to boot real hardware
2021-07-29 20:35:29 <meph> I am of the opinion that all the installers should support ttyS0, but that's just me
2021-07-29 20:35:47 <Hello71> unless you're actually booting from a cd-rom, it's not that hard to create a fat partition and run syslinux
2021-07-29 20:36:03 <captainchris> but where, when i can change font in application that does'nt take effect
2021-07-29 20:37:57 <meph> that's not particularly end-user friendly Hello71
2021-07-29 20:38:42 <Hello71> neither is serial
2021-07-29 20:38:49 <meph> lol.
2021-07-29 20:40:45 <meph> I'm not even suggesting that a getty for ttyS0 should be installed by default on the actual installed system for anything other than virt
2021-07-29 20:45:02 <meph> users that have systems like the pcengines apu's (hardware with a serial port and no vga port) should check their gettys before rebooting out of the installer if they know they need it
2021-07-29 20:45:45 <meph> also, alpine being a lightweight operating system, I would suspect there'd be an interest in supporting smaller embedded systems
2021-07-29 20:45:50 <meph> such as router hardware
2021-07-29 20:46:25 <ikke> That's kind of alpine linux' origins
2021-07-29 20:46:26 <Hello71> the iso needs to support data mode
2021-07-29 20:46:37 <onelegend> Hello71: what do you mean by data mode?
2021-07-29 20:46:37 <meph> your mom need to support data mode
2021-07-29 20:46:44 <Hello71> *plonk*
2021-07-29 20:51:33 <thexzoron> since grsecurity is no longer and PIE and stack smach protection is pretty common amongst distributions, what security features still set apart this distro?
2021-07-29 20:52:08 <Hello71> not that much. still have some like -Wl,-z,now etc
2021-07-29 20:53:02 <meph> Hello71: did you mean: The ISO 13490 standard is an extension to the ISO 9660 format that adds support for multiple sessions on a disc. Since ISO 9660 is by design a read-only, pre-mastered file system, all the data has to be written in one go or "session" to the medium. Once written, there is no provision for altering the stored content. ISO 13490 was created to allow adding more files to a writeable disc such as CD-R in multiple sessions. 
2021-07-29 20:53:28 <meph> I wonder if I could do something like that to the alpine iso lol
2021-07-29 21:03:34 <meph> lol ikke I just used the extended image and figured out all I needed to do was "lts console=ttyS0,115200"
2021-07-29 21:03:43 <meph> maybe there should just be a wiki entry on this
2021-07-29 21:04:09 <thexzoron> Hello71: can I get a list or more info on this?
2021-07-29 21:05:25 <meph> thexzoron: alpine uses musl-libc which is a much simpler codebase than glibc, and also focuses on more explicit posix compliance
2021-07-29 21:06:05 <ikke> meph: yes, that's what I referred to earlier
2021-07-29 21:06:27 <meph> ikke: i know, I can see now that I made this much more difficult than I needed to
2021-07-29 21:06:54 <meph> but hey, at least I discovered the fact that the setup-bootable script seems broken
2021-07-29 21:08:08 <ikke> Can you open an issue for that in the alpine-conf project?
2021-07-29 21:08:17 <meph> yup
2021-07-29 21:11:44 <meph> err sorry for the errant highlight
2021-07-29 21:19:15 <jeffreybenzos[m]> ikke: IT'S ALIVE
2021-07-29 21:19:26 <jeffreybenzos[m]> i'll thank you with some random dumb timelapse i'm gonna make from some clouds in a little
2021-07-29 21:21:32 <doggone> Is there a reason why nf_conntrack_bridge.ko is only included for mips64?
2021-07-29 21:23:00 <doggone> My poor VMs need to bridged for IPv6 and they're sad because they can't make outgoing connections :(
2021-07-29 21:34:15 <jeffreybenzos[m]> What font is everyone using?
2021-07-29 21:34:54 <doggone> Comic Sans MS
2021-07-29 21:45:37 <trylletre> iosevka 
2021-07-29 21:45:48 <invoked> jetbrains mostly
2021-07-29 22:57:04 <Naib> jetbrains is surprisingly good
2021-07-29 23:09:47 <invoked> to my eyes it's nice.
2021-07-29 23:11:41 <invoked> reasonable 'g' glyphs.  i really cannot abide the squiggly 'g' in other fonts. yes i have weird hangups.
2021-07-29 23:53:28 <danieli> i stick to droid sans mono
2021-07-29 23:53:40 <danieli> at times fira code
2021-07-29 23:54:04 <danieli> used to use hasklig for a while but i didn't really need it just for the ligatures so i quit using it
2021-07-30 01:50:48 <aftershock> hey so I have an install that I need to run in BIOS mode instead of UEFI because my gpu doesn't support GOP
2021-07-30 01:51:04 <aftershock> can I do this without needing to reinstall
2021-07-30 01:51:28 <aftershock> I have GPT partitioned disk 
2021-07-30 02:06:26 <jeffreybenzos[m]> i don't have direct experience with this, but i think you would just reconfigure grub
2021-07-30 02:06:42 <rcf> aftershock: see https://wiki.alpinelinux.org/wiki/Bootloaders#Installing_Syslinux
2021-07-30 02:06:44 <jeffreybenzos[m]> but i don't know how that interfaces with your particular scripts, so definitely wait until someone better joins in
2021-07-30 02:07:11 <rcf> aftershock: assuming you want to use syslinux, of course. 
2021-07-30 02:07:22 <aftershock> I don't mind
2021-07-30 02:07:28 <aftershock> if its easy I'm happy
2021-07-30 02:07:32 <rcf> Which you probably do, because GRUB requires repartitioning to boot BIOS from GPT 
2021-07-30 02:07:39 <aftershock> I'll tske a look
2021-07-30 02:07:52 <rcf> (it's the default for BIOS installs) 
2021-07-30 02:08:30 <aftershock> I'm trying to use btrfs which is why I can't use the default install
2021-07-30 02:10:13 <invoked> what does this have to do with your gpu?
2021-07-30 02:11:40 <aftershock> apparently if your gpu is old enough it won't implement the graphics output protocol required to boot the bios in uefi mode
2021-07-30 02:11:47 <Hello71> also "default install"?
2021-07-30 02:12:05 <aftershock> through setup-disk
2021-07-30 02:12:12 <Hello71> eh? normally you can still boot in uefi mode with csm emulation for option rom
2021-07-30 02:12:28 <aftershock> mine won't for some reason
2021-07-30 02:12:31 <Hello71> if your mobo can't do that it's real crap (i mean they all are, but yours is more crap than usual)
2021-07-30 02:12:33 <rcf> I take it this is also an early UEFI implementation. 
2021-07-30 02:12:39 <rcf> Which means yes, it is probably really crap 
2021-07-30 02:12:43 <aftershock> msi pr vdh
2021-07-30 02:13:00 <Hello71> i think if it can't do that then it should still allow booting and have the os initialize it
2021-07-30 02:13:01 <rcf> (unless you just like using ancient GPUs in more modern systems, I won't judge) 
2021-07-30 02:13:03 <Hello71> but again, crap firmwares
2021-07-30 02:13:06 <aftershock> and I flashed it
2021-07-30 02:13:41 <Hello71> rcf: my 6950 works fine in asrock b450 pro4
2021-07-30 02:13:59 <aftershock> well ancient gpu doesn't matter if its just a server
2021-07-30 02:14:06 <Hello71> efi mode and all, but it won't let you turn off legacy option roms, and if you do it temporarily bricks it
2021-07-30 02:14:13 <aftershock> yeah my gtx 560 says no
2021-07-30 02:14:17 <Hello71> i think it needs a manual bios reset
2021-07-30 02:14:35 <rcf> Hello71: I have some experiences with early Insyde UEFI that have taught me that expecting things to work -- no matter how reasonable -- can be a lost cause. 
2021-07-30 02:14:52 <aftershock> I'm just gonna try the install in bios mode
2021-07-30 02:14:59 <Hello71> heh, i also have an insyde efi that won't boot from internal efi or some shit
2021-07-30 02:15:13 <Hello71> or it needs to be mbr formatted or something ridiculous, i don't recall exactly
2021-07-30 02:15:56 <Hello71> i tried flashing it externally but wound up accidentally breaking the bios chip, and i can't be assed to buy a new one
2021-07-30 02:16:10 <rcf> aftershock: if you're willing to reinstall I'd try installing syslinux and configuring it, I think it works with btrfs. 
2021-07-30 02:16:18 <rcf> aftershock: if it fails, oh well, you can just reinstall. 
2021-07-30 02:16:23 <rcf> If it works... you've saved yourself some trouble. 
2021-07-30 02:16:43 <Hello71> rcf: iirc it works as well as syslinux does in general, which is "mostly, as long as you're using a typical configuration and don't poke it"
2021-07-30 02:16:50 <rcf> Indeed. 
2021-07-30 02:17:10 <Hello71> unfortunately syslinux is pretty unmaintained, even more so than grub
2021-07-30 02:17:29 <Hello71> which is also basically unmaintained
2021-07-30 02:17:34 <aftershock> <rcf> ok thanks yeah I'll try syslinux
2021-07-30 02:17:59 <Hello71> nobody wants to write reliable 16-bit real mode x86 code anymore, whodathunkit
2021-07-30 02:18:15 <aftershock> I use refind on my main machine
2021-07-30 02:18:16 <Hello71> aftershock: don't do that on irc, it's quoting people
2021-07-30 02:18:19 <aftershock> works ok
2021-07-30 02:18:36 <Hello71> refind works better than syslinux or grub because it intentionally doesn't do much
2021-07-30 02:18:52 <Hello71> it's a boot manager, not a boot loader
2021-07-30 02:19:04 <Hello71> best way to write software is to punt the work to someone else
2021-07-30 02:19:11 <aftershock> true
2021-07-30 02:19:25 <Hello71> i only wish the big software companies did that more
2021-07-30 02:21:05 <aftershock> how do you reply to someone on irc?
2021-07-30 02:21:19 <aftershock> < > doesn't work?
2021-07-30 02:21:49 <rcf> <nick>: implies that nick said something. nick: is how you address someone. 
2021-07-30 02:22:00 <aftershock> oh ok
2021-07-30 02:22:02 <aftershock> thanks
2021-07-30 04:45:56 <bolangi> quit
2021-07-30 04:46:57 <ikke> dramatic exit :)
2021-07-30 08:48:09 <Skaface82> hey im trying to run libvirt, I have changed libvirtd.conf to set the group to libvirtd, restart the service, and the permissions on libvirt-sock are still root:root
2021-07-30 08:48:35 <Skaface82> it seems like its not reading libvirtd.conf on startup?
2021-07-30 08:49:36 <lopid> is the socket ownership supposed to change?
2021-07-30 08:53:13 <Skaface82> as far as im aware.. i set   unix_sock_group = libvirtd
2021-07-30 08:53:30 <lopid> oh
2021-07-30 08:56:50 <Skaface82> sorry i meant libvirt
2021-07-30 09:09:52 <ncopa> Skaface82: I use libvirt here
2021-07-30 09:10:10 <ncopa> $ ls -la  /var/run/libvirt | tpaste
2021-07-30 09:10:10 <ncopa> https://tpaste.us/aZZZ
2021-07-30 09:10:29 <ncopa> srwxrwx---  1 root libvirt   0 Jul 30 08:38 libvirt-sock
2021-07-30 09:10:45 <Skaface82_> weird
2021-07-30 09:11:10 <Skaface82_> is your config file at /etc/libvirt/libvirt.conf ?
2021-07-30 09:11:12 <ncopa> let me see how it is configured
2021-07-30 09:11:28 <Skaface82_> ohhh its libvirtd.conf
2021-07-30 09:11:51 <Skaface82_> never mind Ive got it
2021-07-30 09:11:56 <ncopa> yeah, there you go :)
2021-07-30 09:11:56 <Skaface82_> but thanks
2021-07-30 09:12:04 <ncopa> np
2021-07-30 13:52:17 <greatwielder[m]> when doing rc-update add srv, is default added by default to the end?
2021-07-30 13:53:27 <ikke> greatwielder[m]: It just creates a symlink, there is no ordered list
2021-07-30 14:17:19 <greatwielder[m]> functionally, there is no difference other than the association with a list? what about ordering?
2021-07-30 15:38:11 <captainchris> Hi everybody
2021-07-30 15:40:09 <wener> hi
2021-07-30 15:41:07 <ajin> hi captain chris
2021-07-30 16:01:57 <captainchris> I have a problem with suckless slstatus it returns me a segfault due to musl
2021-07-30 16:02:30 <captainchris> file containing a list of commands to execute on startup
2021-07-30 18:02:35 <captainchris> Re everybody
2021-07-30 22:06:56 <nbastin> is there a trick in /etc/network/interfaces to configuring ovs bridge ports?
2021-07-30 22:07:11 <nbastin> I have an interface v42-mgmt, and a block in /etc/network/interfaces for it with a static ip, but ifup does not do anything
2021-07-30 22:07:48 <nbastin> I have a functionally similar block for eth2, and that works fine
2021-07-30 22:11:12 <nbastin> if I use --force it works, there's something about the lock file already existing, even though it's not actually configured
2021-07-30 22:12:19 <Hello71> 3.14?
2021-07-30 22:12:25 <nbastin> yeah
2021-07-30 22:13:45 <nbastin> all the interfaces defined in /etc/network/interfaces have /run/ifstate.*.lock files, whether they're actually configured or not
2021-07-30 22:13:52 <nbastin> which seems to block any subsequent ifup from working
2021-07-30 22:17:28 <Hello71> Ariadne: ^
2021-07-30 22:18:45 <Ariadne> those are deleted on exit of ifupdown
2021-07-30 22:18:57 <Ariadne> you can just delete them
2021-07-30 22:19:44 <nbastin> sure, although I'm wondering why it doesn't work on boot
2021-07-30 22:19:48 <nbastin> when presumably they are not there.. :-)
2021-07-30 22:20:00 <nbastin> something seems to write them there, but not succeed at actually bringing the interfaces up
2021-07-30 22:24:00 <nbastin> anyhow I gotta run, I'll poke at this some more later
2021-07-31 01:31:48 <Ariadne> okay
2021-07-31 01:32:11 <Ariadne> try ifup -v and also post your /etc/network/interfaces
2021-07-31 01:32:28 <Ariadne> (to a pastebin not here)
2021-07-31 01:34:24 <trylletre> Ariadne: unrelated, but cute hair!
2021-07-31 01:43:38 <onelegend> lol
2021-07-31 01:44:37 <Ariadne> which one
2021-07-31 01:49:37 <Ariadne> trylletre: which one
2021-07-31 01:50:00 <trylletre> the new hair colors
2021-07-31 01:50:09 <onelegend> the cretan princess?
2021-07-31 01:50:33 <trylletre> but specifically hair #43,756 on the left side of the head :)
2021-07-31 02:19:44 <riverdc> how is it possible that I have edge/testing in my apk repositories (and have just updated), but `apk search --no-cache ocaml-camlp5` gives no results?
2021-07-31 02:20:29 <riverdc> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/testing/ocaml-camlp5
2021-07-31 02:20:33 <riverdc> hasn't been updated recently or anything
2021-07-31 02:21:49 <riverdc> ah, got it, sorry! looks like no architectures are enabled in the APKBUILD
2021-07-31 02:22:55 <trylletre> yep, the last commit lists that its incompatible with some ocaml version
2021-07-31 02:23:01 <trylletre> p4 exists instead
2021-07-31 02:31:25 <TranceKat> Hi all.  I'm experiencing some difficulty in getting iSCSI to work properly in alpine.
2021-07-31 02:31:25 <TranceKat> 3.12.7
2021-07-31 02:31:25 <TranceKat> nexus [~]# cat /etc/alpine-release
2021-07-31 02:31:25 <TranceKat> nexus [~]# iscsiadm --version
2021-07-31 02:31:25 <TranceKat> iscsiadm version 2.1.2
2021-07-31 02:31:27 <TranceKat> Logging in to [iface: default, target: iqn.2021-07.lan.abyss:apps-iscsi, portal: abyss,3260]
2021-07-31 02:31:27 <TranceKat> nexus [~]# iscsiadm --mode node --targetname iqn.2021-07.lan.abyss:apps-iscsi --portal abyss --login
2021-07-31 02:31:29 <TranceKat> It all looks correct, but I don't get a new block device to mount.
2021-07-31 02:31:29 <TranceKat> Login to [iface: default, target: iqn.2021-07.lan.abyss:apps-iscsi, portal: abyss,3260] successful.
2021-07-31 02:31:31 <TranceKat> nexus [~]# fdisk -l
2021-07-31 02:31:31 <TranceKat> Disk /dev/sda: 64 GB, 68719476736 bytes, 134217728 sectors
2021-07-31 02:31:33 <TranceKat> 65536 cylinders, 64 heads, 32 sectors/track
2021-07-31 02:31:33 <TranceKat> Units: sectors of 1 * 512 = 512 bytes
2021-07-31 02:31:35 <TranceKat> Device  Boot StartCHS    EndCHS        StartLBA     EndLBA    Sectors  Size Id Type
2021-07-31 02:31:39 <TranceKat> Any insights as to what I may be doing wrong, please?
2021-07-31 04:02:49 <riverdc> how should I package a library that doesn't ship anything other than development files? if I follow the normal subpackage setup, the main package will be empty
2021-07-31 04:53:40 <Thermi> riverdc, look at pybind11
2021-07-31 04:53:50 <Thermi> riverdc, that package is header-only, too
2021-07-31 04:55:41 <Thermi> Huh. Seems like it's not header only at all, although it's supposed to be.
2021-07-31 04:55:52 <Thermi> Ah. Tests.
2021-07-31 04:57:03 <riverdc> it's an ocaml library and it looks like the existing ocaml library packages don't even have a -dev subpackage
2021-07-31 04:57:21 <riverdc> i realized it wouldn't be completely header only as well, because there are some shared objects
2021-07-31 04:59:28 <Thermi> Okay, and I realized there are python files in the main package
2021-07-31 04:59:35 <Thermi> so pybind11 isn't header only at all.
2021-07-31 05:00:31 <Thermi> riverdc, if the shared objects are usable without the headers of the -dev package, they should be packaged seperately
2021-07-31 06:00:06 <jeffreybenzos[m]> guhhhh, firefox's audio is completely bugged
2021-07-31 06:00:16 <jeffreybenzos[m]> tried to install pipewire to fix it, but nope
2021-07-31 06:11:21 <jeffreybenzos[m]> no clue how applications even plug in to pipewire instead of alsa, actually
2021-07-31 07:05:42 <wej> jeffreybenzos[m]:you can either enable alsa emulation with pipewire or use the pipewire-pulse tool for pulseaudio emulation
2021-07-31 07:16:34 <jeffreybenzos[m]> wej: how do either of those work? the wiki seems to be incorrect
2021-07-31 07:16:44 <jeffreybenzos[m]> can't see anything about with-alsa on the pipewire wiki anymore
2021-07-31 07:18:21 <wej> as for alsa, it is basically copying of some config file to the alsa config dir /etc/conf.d/alsa
2021-07-31 07:18:35 <wej> it is explained here: https://github.com/PipeWire/pipewire/blob/master/INSTALL.md
2021-07-31 07:18:59 <wej> for pulseaudio you just need to run pipewire-pulse and it should work
2021-07-31 07:20:47 <jeffreybenzos[m]> oh, instead of running pipewire &?
2021-07-31 07:20:57 <jeffreybenzos[m]> i'm running that in bspwmrc; is there somewhere else it should be run?
2021-07-31 07:37:49 <wej> jeffreybenzos[m]: no, in addition to it. it is basically a proxy that translates between the pulseaudio protocol and pipewire
2021-07-31 07:38:13 <wej> running it with the start of your session should be okay
2021-07-31 07:39:03 <jeffreybenzos[m]> Just to be sure what you're saying, you want me to change bspwmrc to first do pipewire &, and then do pipewire-pulse &?
2021-07-31 07:39:17 <jeffreybenzos[m]> seems like my flatpak stuff is working fine just doing pipewire-pulse &
2021-07-31 07:42:51 <wej> yep
2021-07-31 07:43:21 <wej> that is weird. you could check if there is a pipewire process running elsewhere
2021-07-31 07:52:44 <captainchris> Hi everybody 
2021-07-31 07:53:53 <captainchris> I installed nyxt browser but when I don't have nyxt binary. How I can list files with apk ??
2021-07-31 07:54:10 <ikke> apk info -L nyxt
2021-07-31 07:55:26 <captainchris> yes man 8 apk-info 
2021-07-31 07:55:34 <captainchris> sorry 
2021-07-31 07:56:14 <captainchris> I read man 8 apk-index and I didn't found informations 
2021-07-31 07:56:46 <Skaface82> TranceKat: I actually tested out iscsi on alpine yesterday
2021-07-31 07:57:10 <Skaface82> i followed the same steps as you, and it created all of the /dev/sda1 2 3 etc.
2021-07-31 07:58:20 <Skaface82> TranceKat: it looks like a blank unpartitioned device?
2021-07-31 07:58:41 <Skaface82> do you need to create partitions and format them?
2021-07-31 08:04:52 <jeffreybenzos[m]> i don't think so, wej: ps aux was just giving me the basics
2021-07-31 08:05:08 <jeffreybenzos[m]> unless pipewire is the big part here
2021-07-31 08:05:30 <jeffreybenzos[m]> i see pipewire run on its own, and then also pipewire-media-session, and then pipewire -c pipewire-pulse.conf
2021-07-31 08:06:28 <wej> so you do have a pipewire process running?
2021-07-31 08:06:32 <jeffreybenzos[m]> aye.
2021-07-31 08:06:57 <wej> then everything should be fine, i thought you only had pipewire-pulse running
2021-07-31 08:08:06 <jeffreybenzos[m]> Hmmmm. Interesting
2021-07-31 08:08:16 <jeffreybenzos[m]> At least now we both don't know.
2021-07-31 10:30:05 <Skaface> hey, im trying to install rpcsvc-proto and also nfs-utils, but I get an error saying  nfs-utils trying to overwrite usr/bin/rpcgen owned by rpcsvc-proto
2021-07-31 11:30:05 <Skaface> hmm i resolved it, im not sure how
2021-07-31 12:57:19 <Hello71> riverdc: i have proposed moving this package to unmaintained: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/23041
2021-07-31 18:31:30 <nbastin> Ariadne: if you're around at any point, https://termbin.com/1rs3 is my /etc/network/interfaces, https://termbin.com/rj9v is ifup -v v200-ntn (I was doing v42-mgmt last time, but I already forced that up)
2021-07-31 18:32:31 <nbastin> fwiw https://termbin.com/rzz9 is ovs-vsctl show
2021-07-31 18:33:07 <Ariadne> nbastin: can you open a bug at https://github.com/ifupdown-ng/ifupdown-ng
2021-07-31 18:33:41 <Ariadne> since you’re doing vlan stuff i rather have the person who wrote the vlan code look at this
2021-07-31 18:33:54 <nbastin> sure, although I'm not entirely sure what to report - is there any way to get it to log the equivalent of -v at boot?
2021-07-31 18:34:11 <Ariadne> uhh
2021-07-31 18:34:14 <nbastin> (absent just modifying it)
2021-07-31 18:34:15 <Ariadne> no idea :/
2021-07-31 18:34:36 <nbastin> ok, I can just modify it to +x and write stderr to a file
2021-07-31 18:35:15 <nbastin> because it's on boot where something goes awry, I think...it fails now just because the lock files are already there, but they either shouldn't be there or the interfaces should be up
2021-07-31 19:08:34 <nbastin> is there a reason that "search" is not in apk --help, or should I file a ticket somewhere?
2021-07-31 19:12:52 <ikke> nbastin: it should've been recently fixed
2021-07-31 19:16:16 <nbastin> yeah I finally found the issue and merge request, thanks
2021-07-31 19:17:00 <nbastin> although I'm not sure that fixed it...
2021-07-31 19:17:10 <nbastin> since it only modified the man pages
2021-07-31 19:18:32 <nbastin> maybe I'm reading this wrong, but 10746 in apk-tools says apk --help does not mention the command, but the MR that closed the issue doesn't change that
2021-07-31 19:19:26 <nbastin> unless --help is somehow derived from apk.8.scd
2021-07-31 19:21:19 <ikke> Yes, they are linked
2021-07-31 19:21:31 <nbastin> ok
2021-07-31 19:22:02 <ikke> There is a lua script that embeds the man page
2021-07-31 19:22:41 <nbastin> yeah I finally worked my way through that part of the build....it was more complex than I figured.. :-)
2021-07-31 19:28:55 <jeffreybenzos[m]> what's the command to explode firefox into little pieces
2021-07-31 19:31:59 <ikke> rm -rf /
2021-07-31 19:38:07 <ajin> I'm trying to start postgres for the first time with rc-service postgresql start
2021-07-31 19:38:17 <ajin> unning bootstrap script ... 2021-07-31 19:35:50.962 GMT [12972] LOG:  invalid value for parameter "log_timezone": "US//Eastern"
2021-07-31 19:38:20 <ajin> 2021-07-31 19:35:50.963 GMT [12972] LOG:  invalid value for parameter "TimeZone": "US//Eastern"
2021-07-31 19:41:09 <ajin> maybe this is because I install alpine 3.10, but then updated to 3.14?
2021-07-31 19:42:57 <ajin> /etc/timezone doesn't exist, so I don't know how postgres even knows my timezone 
2021-07-31 19:47:45 <ajin> apparently postgres only accepts EST5EDT for my timezone
2021-07-31 19:49:33 <ajin> sorry for bothering everyone
2021-07-31 20:01:15 <Hello71> or more likely it doesn't like multiple slashes
2021-07-31 20:01:59 <jeffreybenzos[m]> ikke: we're almost there. not even installing pipewire has fixed the firefox audio
2021-07-31 20:02:20 <jeffreybenzos[m]> if i use apk del to nuke it, does that remove profiles, etc?
2021-07-31 20:02:29 <jeffreybenzos[m]> i don't think i have anything really in there yet
2021-07-31 20:02:43 <jeffreybenzos[m]> but i assume uninstalling and reinstalling doesn't always work anyway
2021-07-31 20:02:44 <trylletre> profiles are generally under home and not deleted
2021-07-31 20:04:07 <trylletre> under ~/.mozilla or something, and cache under ~/.cache/mozilla, but i doubt itll fix it anyway
2021-07-31 20:04:13 <jeffreybenzos[m]> same
2021-07-31 20:04:37 <jeffreybenzos[m]> yeah, no dice
2021-07-31 20:04:46 <jeffreybenzos[m]> and chromium has random tab crashes
2021-07-31 20:04:52 <jeffreybenzos[m]> but at least sound works
2021-07-31 23:37:02 <Flimmy> problem: authentication request to the driver failed -- when connecting to a wifi access point, following: wpa_supplicant step