2024-04-01 00:24:42 Last kernel buildable with tcc is 2.4.x ... 2024-04-01 01:52:55 linux barely builds with clang 2024-04-01 01:52:57 let alone any other compiler 2024-04-01 09:09:41 hello, this morning I tried to upgrade my laptop under alpine edge, I have the following error 2024-04-01 09:09:50 ERROR: bubblewrap-0.9.0-r0: BAD signature 2024-04-01 09:10:51 how is that possible? 2024-04-01 09:40:34 should’ve put the packages in the fridge or upgraded sooner—the signature spoiled 2024-04-01 09:58:43 o/ 2024-04-01 09:59:04 I'm checking kernel-hooks to make a custom uki file after each upgrade of linux kernel 2024-04-01 09:59:16 but it looks like the hooks are called before mkinitfs is called 2024-04-01 10:03:52 anybody knows how to fix this? (1/1) Installing bubblewrap (0.9.0-r0) 2024-04-01 10:03:54 ERROR: bubblewrap-0.9.0-r0: BAD signature 2024-04-01 10:04:07 i saw on 'so' about clearing /var/cache/apk, but there is nothing there for me 2024-04-01 10:04:26 (i'm installing the package in a custom root directory, with apk add -p) 2024-04-01 10:05:52 welcome to the club 2024-04-01 10:09:52 maybe there's something wrong with the bubblewrap package itself? i manage to clean up a bit, and the other dependencies for luakit are installing correctly 2024-04-01 10:10:35 I remember having this when date was incorrect 2024-04-01 10:10:46 the bubblewrap package was built today also 2024-04-01 10:10:54 and yesterday i installed luakit on my main os for testing 2024-04-01 10:12:42 huh, I see that too from a couple of repositories 2024-04-01 10:15:13 usually "BAD signature" means apk with untrusted key? or something else 2024-04-01 10:18:53 It can also mean that the package with the same version was rebuilt, and our cdn cached the old version, so the signature in the index does not match the package in the cache 2024-04-01 10:20:32 That does not seem to be the case with bubblewrap 2024-04-01 10:21:37 source tarball checksum is ok. `apk verify ./bubblewrap*.apk` is ok too... not sure what it is tbh 2024-04-01 10:24:10 The sha256sum on the builder matches what's on the master mirror, and matches what I get from dl-cdn 2024-04-01 10:26:56 is it possible that the package was signed with a key not present in `alpine-keys`? i dont know how to check the key that an apk was signed with 2024-04-01 10:30:10 rdbo: tar tzf 2024-04-01 10:30:39 the file .SIGN.RSA.*.rsa.pub must match a key in /etc/apk/repositories 2024-04-01 10:32:46 rdbo: it seems like apk-tools also received an update, not sure if it's related to that 2024-04-01 10:33:33 f3b3fc2428f880a18257685fc89612487bdfb4cbcfa326a7d76c9b86f7c18546 .SIGN.RSA.alpine-devel@lists.alpinelinux.org-6165ee59.rsa.pub 2024-04-01 10:33:35 207e4696d3c05f7cb05966aee557307151f1f00217af4143c1bcaf33b8df733f /etc/apk/keys/alpine-devel@lists.alpinelinux.org-6165ee59.rsa.pub 2024-04-01 10:33:38 did i do this wrong? 2024-04-01 10:33:44 looks like an issue 2024-04-01 10:34:56 rdbo: the file itself contains a signature 2024-04-01 10:35:01 actually /etc/apk/keys/ is bigger 2024-04-01 10:35:12 it's just the filename that needs to match 2024-04-01 10:35:27 oh ok 2024-04-01 10:36:14 This is also being discussed in #alpine-devel, where the apk-tools maintainer is 2024-04-01 10:36:38 Seems to be an issue with how the index is generated 2024-04-01 10:37:59 https://irclogs.alpinelinux.org/%23alpine-devel-2024-04.log 2024-04-01 10:56:37 I have previously installed Alpine Linux on a USB stick with XEN using a mini PC and then booted it successfully from the USB stick. I booted the install ISO from a USB stick when i first set it up. Now i wanted to make another installation on the same hardware but when the installer boots it gives the following error: "ERROR: modloop failed to start" 2024-04-01 10:57:34 And when reaching the login prompt the keyboard doesnt work (which it does when GRUB is loading and also when booting from my XEN USB stick). If someone could point me in a direction on how to solve this I'd appreciate it, because so far none of the things I've tried has helped. 2024-04-01 10:59:28 EvTheFuture if i'm not mistaken 'modloop' is what mounts /.modloop if i'm not mistaken (which is where the kernel modules are located) 2024-04-01 10:59:38 and without kernel modules, i guess the keyboard won't work 2024-04-01 11:00:45 i would check what the modloop service does and try to go from there 2024-04-01 11:01:01 maybe you are missing whatever it mounts that has the kernel modules 2024-04-01 11:13:18 rdbo: yes that's what I'm thinking since no kernel modules can be loaded. However I tested the XEn installation ISO and that one works just fine though... Kind of interesting 2024-04-01 11:17:39 Did you do a "sys-install" (install to hard disk) at any point? If so, if the kernels are different you might be running into a situation where the modules on the disk are different. 2024-04-01 11:19:40 nangel: No, I don't get that far since it is the installation media that fails to load the modloop and thus all kernel modules... 2024-04-01 11:19:58 But interestingly when using the xen image instead of the extended image it works 2024-04-01 11:48:22 hello all. I wonder if Alpine's Clang defaults have to do with the visibility of this musl issue: https://www.openwall.com/lists/musl/2024/03/29/3 2024-04-01 11:49:19 My system wakes up from sleep when I move my mouse. how do I disable wakeups for just this usb device? 2024-04-01 11:49:54 WhyNotHugo: have you checked /proc/acpi/wakeup ? 2024-04-01 11:50:03 I disabled evertyhing on a zzz hook 2024-04-01 11:50:40 WhyNotHugo, check in your BIOS first 2024-04-01 11:51:07 s/first/also/ 2024-04-01 11:52:57 oh yes, mine doesn't have any option related :( 2024-04-01 11:53:03 I only see PCI devices in /proc/acpi/wakeup. Most of them don't specify a sysfs node 2024-04-01 11:54:14 you can try disabling everything but maybe you end on my same stupid scenario where I could only wake up with power button and the power button triggered a shutdown :) 2024-04-01 11:54:59 jaja 2024-04-01 11:56:09 WhyNotHugo, note that the device type/id matters, not so much the bus it's connected with 2024-04-01 11:56:17 Could you paste the contents? 2024-04-01 11:57:08 (for example *HCI is USB) 2024-04-01 11:57:23 https://paste.sr.ht/blob/8d5472a619b071e45f405764ec0b99cf30801d6f 2024-04-01 11:57:32 I found it in /sys/class/wakeup too 2024-04-01 11:58:26 That's a lot of devices :D 2024-04-01 11:58:42 Maybe disable XHCI 2024-04-01 12:08:04 WhyNotHugo: get a PCI/PCIe USB card and disable the wake-up for that card. then plug all USB devices that need to to ignored into that card. 2024-04-01 12:54:36 we are working on fixing the APKINDEX 2024-04-01 12:54:45 need some manual steps 2024-04-01 12:54:50 to fix the builders 2024-04-01 13:37:05 damn, the timing 2024-04-01 13:37:16 just when i wanted to build some images for templating 2024-04-01 13:38:06 The APKINDEX file is missing: https://dl-cdn.alpinelinux.org/alpine/edge/community/aarch64/ 2024-04-01 13:38:29 I also got this when I tried to upgrade: ERROR: apk-tools-2.14.3-r0: BAD signature 2024-04-01 13:38:42 Yeah, it's being worked on ikkoma 2024-04-01 15:14:10 all repositories should be fixed again 2024-04-01 15:14:11 sorry 2024-04-01 15:14:29 all repos should be working 2024-04-01 15:14:37 \o/ 2024-04-01 15:22:32 \o/ 2024-04-01 15:24:12 what a day... 2024-04-01 15:24:20 yup 2024-04-01 15:34:40 hey everybody! Thank you for your patience! you have no idea how much it means 2024-04-01 16:10:19 well done! 2024-04-01 16:25:57 thanks @ncopa ; btw, I'm still getting ERROR: apk-tools-2.14.3-r0: BAD signature, on my raspberry pi 2024-04-01 16:27:34 which repo/mirror/version do you use? 2024-04-01 16:50:45 mirrors.lax.silicloud.com/alpine/v3.19 is still giving me bad signature for apk-tools ... rest of my machines (on different urls) are fine. 2024-04-01 16:51:45 last updated 3 hours ago 2024-04-01 16:55:24 yeah it should be fixed with next sync 2024-04-01 19:35:56 thank you, it is resolved now. thank you 2024-04-01 22:02:16 how would I control what routes are created when connected to a network? if I connect to wlan0 + eth0, both get added as destination 0.0.0.0, gateway but only one works (which is expected, only the wlan0 one has access to the outer network, eth0 is limited to 10.42.0.1/24) 2024-04-01 22:02:37 right now I've been manually and periodically `route del default gw 10.42.0.1 eth0` 2024-04-01 22:04:54 How do you connect to a network? 2024-04-01 22:06:07 wpa_supplicant does it automatically (though I'm not sure how it gets started, it's not in rc-status so will take a look) 2024-04-01 22:06:56 wpa_supplicant doesn't do anything with IPs or routes afaik 2024-04-01 22:07:28 And surely you're not using it on eth0 2024-04-01 22:07:30 ah, is it udhcpc -i wlan0 then 2024-04-01 22:07:33 I am not 2024-04-01 22:07:48 eth0 is only configured in /etc/networking/interfaces 2024-04-01 22:07:50 Then IPs and routes are configured by your dhcp client 2024-04-01 22:08:50 ah that makes sense. will take a look there then, thanks! 2024-04-01 22:09:41 IIRC udhcpc triggers some scripts, passing information given by the dhcp server. So you can modify these to do what you want. 2024-04-01 22:11:38 didn't know udhcpc was part of busybox, good to know 2024-04-01 23:48:22 hm, well I think I can do what I want, but I'm having trouble with something. when a connection is established everything is fine, but very quickly all requests stop working. running `udhcpc -i wlan0` fixes that for a very brief window of time 2024-04-01 23:48:47 no changes in `route -n` during that time 2024-04-01 23:49:15 `ip a` output is also identical when it works and doesn't 2024-04-02 00:09:09 something different: when it works, the wlan0 ip shows up on my router's dashboard, but when it doesn't, it doesn't 2024-04-02 00:09:29 connecting to devices on the LAN always works though 2024-04-02 00:13:34 nvm, connecting on the LAN only works because I was also connected through eth0 2024-04-02 03:29:25 hello 2024-04-02 03:30:02 does anyone know how to configure stable private ipv6? (rfc7217/private-stable) 2024-04-02 03:30:43 I've tried setting sysctl but it doesn't get used on boot, requiring an ifup/ifdown dance for it to get applied 2024-04-02 03:31:03 (net.ipv6.conf.default.stable_secret / net.ipv6.conf.default.addr_gen_mode) 2024-04-02 03:35:47 "privext 2" in interfaces didn't work either 2024-04-02 03:42:24 attempting to set sysctl in /etc/network/interfaces pre-up results in 2024-04-02 03:42:41 "cannot stat /proc/sys/net/...: No such file or directory" 2024-04-02 04:37:21 May I ask which filesystems are supported on Raspberry Pi 4? 2024-04-02 04:38:23 I use setup-disk and it prompts that only vfat is supported. 2024-04-02 05:02:01 might be related to that the rpi default format is fat32 to be able to boot 2024-04-02 05:02:13 not sure if newer models supports other partitiontypes for booting 2024-04-02 05:03:49 I believe they're all limited to vfat for /boot 2024-04-02 05:03:50 That's the boot partition, I'm talking about the root partition. 2024-04-02 05:05:42 The example used in the wiki is ext4 2024-04-02 05:05:46 https://wiki.alpinelinux.org/wiki/Raspberry_Pi 2024-04-02 05:07:01 I haven't tried ext4 yet, I just want to know which filesystems Alpine supports. 2024-04-02 05:07:23 generally speaking, the same as any other linux distro 2024-04-02 05:09:19 Other distributions support f2fs, but I'm not sure what to do with Alpine. 2024-04-02 05:10:27 grep CONFIG_.*FS= /boot/config-lts 2024-04-02 05:10:28 So I am trying to use filesystems supported by setup-disk. 2024-04-02 05:15:10 grawity: http://fxzx.nl/jpwj2j 2024-04-02 05:15:51 The kernel supports f2fs. 2024-04-02 05:16:06 damn, you've got a loading screen for a pastebin 2024-04-02 05:28:46 In some other places, it is mentioned that if I want to use a specific filesystem, I need to remake initramfs. I'm not sure if this is the problem here. 2024-04-02 05:29:40 the initramfs needs to contain drivers needed for mounting the '/' fs 2024-04-02 05:30:03 unless the drivers are compiled-in ("=y") 2024-04-02 05:30:31 so if you look at your kernel config, "=y" means it will always work and "=m" means it will work if the initramfs is rebuilt with it included 2024-04-02 05:30:43 well, or if it's already included there, ofc 2024-04-02 05:34:12 OK. That could be a mistake appearing elsewhere. 2024-04-02 05:37:38 Another guess: Is it that the root parameter in the config.txt file only supports device path like/dev/mmcblk0p2 and not UUID? 2024-04-02 05:56:22 with an initramfs, the parameter is interpreted by the initramfs, and in the case of alpine supports anything `mount` will accept, I think 2024-04-02 05:56:38 without an initramfs, it's interpreted by the kernel which only supports PARTUUID= and PARRTLABEL= 2024-04-02 06:13:33 answering myself but apparently it's by using dhcpcd - the config is on different location and ignores sysctl but at least this one seems to work 2024-04-02 08:47:25 hello u guys. I need help 2024-04-02 08:48:09 i apk add gdm gdm-openrc and rc-service gdm start 2024-04-02 08:48:20 the screen is black 2024-04-02 08:49:35 from console? 2024-04-02 08:50:45 yes 2024-04-02 08:51:28 is there nothing on one of the other vt's? 2024-04-02 08:51:43 just a _ 2024-04-02 08:52:09 no prompt event on vt1? 2024-04-02 08:52:11 even 2024-04-02 08:52:17 yes 2024-04-02 08:52:30 I run that command in tty1 2024-04-02 08:52:37 check the logs 2024-04-02 08:53:24 I dont know where's the log😂️ 2024-04-02 08:53:44 /var/log/messages maybe, check them all in that dir. you can order by time 2024-04-02 08:53:59 thank you man 2024-04-02 08:54:09 c u latter 2024-04-02 08:57:19 I checked it 2024-04-02 08:58:17 so gdm wont run if i dont install Xorg in alpine 2024-04-02 08:58:27 makes sense 2024-04-02 08:59:29 Im using opensuse without Xorg and gdm also works well🤣️ 2024-04-02 08:59:39 wayland? 2024-04-02 08:59:43 yes 2024-04-02 09:00:16 I already enable the settings in /etc/gdm/custom.conf 2024-04-02 09:01:12 do you have wayland installed? 2024-04-02 09:24:04 saddly. I installed wayland and xorg-server xf86-input-libinput xf86-video-amdgpu 2024-04-02 09:24:49 gdm seems working but black screen 2024-04-02 09:25:15 keyboard not work 2024-04-02 09:27:47 did you set this var? https://wiki.alpinelinux.org/wiki/Wayland 2024-04-02 09:30:23 oh. c u latter 2024-04-02 09:31:28 I have installed elogind and started it before starting gdm 2024-04-02 09:34:25 gdm log has changed. it looks like gdm will use Xorg as backend anyway 2024-04-02 09:38:09 i checked /var/lib/gdm/.local/share/xorg/Xorg.0.log. it looks like the gdm user has no "tty" permisson 2024-04-02 09:38:20 lmao 2024-04-02 09:45:21 that's a stupid path 2024-04-02 09:45:33 i add gdm to video group and it still do not work 2024-04-02 09:45:58 Xorg.log said 'xf86OpenConsole: Cannot open virtual console 1 (Permission denied)' 2024-04-02 11:03:26 you guys know any hex viewer and editor available in the repos? i found imhex but it doesnt work for me 2024-04-02 11:03:33 something like hexedit 2024-04-02 11:05:32 vim can do it in conjunction with xxd 2024-04-02 11:10:37 grawity: I still cannot install it on f2fs. I can't see the error, but I guess it's still related to the ROOT partition. 2024-04-02 11:47:02 It seems that I may need to know what the setup-disk did. 2024-04-02 11:55:38 Hi I am trying to install Alpine for the first time on bare metal but I get in trouble right away with my 2 ethernet controllers (RTL8111/8168/8411 PCI Express Gigabit Ethernet and RTL8125 2.5GbE Controller) not being recognized it seems (setup-alpine doesn't work and i got nothing in config -a) 2024-04-02 11:57:01 I do not see anything in dmesg and modprobe r8125 tells me "module r8125 not found" 2024-04-02 11:57:29 do you see them in lspci? 2024-04-02 11:59:14 but there is no lspci command available when I boot ... 2024-04-02 12:00:09 I used the alpine-extended iso 2024-04-02 12:13:15 https://pkgs.alpinelinux.org/contents?branch=edge&name=raspberrypi%2dbootloader&arch=aarch64&repo=main 2024-04-02 12:14:44 Actually, these files are usually located at /media/mmcblk0p1. 2024-04-02 12:16:09 https://pkgs.alpinelinux.org/contents?branch=edge&name=linux%2drpi&arch=aarch64&repo=main 2024-04-02 12:16:13 Moreover, the location of these files does not match the actual situation. 2024-04-02 19:23:25 anyone knows about my kernel-hooks question? (aka it run hooks before calling mkinitfs) 2024-04-02 21:37:02 does anybody know how to setup gtk themes? i thought i knew, but apparently i'm missing something 2024-04-02 21:37:10 i installed the theme at /usr/share/themes/ 2024-04-02 21:37:21 and setup .gtkrc-2.0 and .config/gtk-3.0 2024-04-02 21:37:39 yet, the theme is not being chosen by the apps (even though in lxappearance it shows as selected) 2024-04-02 21:38:15 also, i ran `gsettings get org.gnome.desktop.interface gtk-theme`, and it shows 'Adwaita' instead of my theme (even thou there is no Adwaita in /usr/share/themes) 2024-04-02 21:38:48 running `gsettings set ...` works, but i am setting up /etc/skel, not configuring for myself, so i can't really run that afaik 2024-04-02 21:40:48 my last hope is reverse engineering the source code of gsettings and figure out what it does, but i'm all ears for suggestions 2024-04-02 23:25:55 Is there an easy way to see what app might be requesting an exclusion rule? For instance, I can't seem to apk add dosbox-staging, as it breaks !dosbox-staging, which is not in my /etc/apk/world 2024-04-02 23:26:30 Sorry, I mean this: 2024-04-02 23:26:30 ``` 2024-04-02 23:26:30 ERROR: unable to select packages: 2024-04-02 23:26:30 dosbox-staging-0.81.0-r0: 2024-04-02 23:26:30 breaks: dosbox-staging-0.81.0-r0[!dosbox] 2024-04-02 23:26:32 satisfies: world[dosbox-staging 2024-04-02 23:26:34 ``` 2024-04-02 23:26:59 dosbox is not installed 2024-04-03 00:35:57 Why does https://wiki.alpinelinux.org/wiki/Nginx creates a newly 'www' user, when 'nginx' is already created? E.g. /var/lib/nginx already belongs to nginx. Why to change it to a newly created user? 2024-04-03 00:45:02 might not be up to date 2024-04-03 03:46:25 not sure if this is the right channel for this, but does anyone know what "0" in the secfixes section of an APKBUILD file means? for example https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/zlib/APKBUILD#L18 2024-04-03 03:50:02 I think I answered my own question by digging through the git history - it means "does not affect any version" (https://gitlab.alpinelinux.org/alpine/aports/-/commit/02ab5a25b2f11a81d1d698e5d8d585d89e46fe97) 2024-04-03 04:23:32 aaronwebber: correct 2024-04-03 05:56:04 is it good to ask support questions here? 2024-04-03 06:42:58 pijul: yes, that's mainly the goal of this channel 2024-04-03 06:51:22 @ikkes ok, thanks. Is it possible to follow this guide here: https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop and thereafter install gnome and distrobox? Are there any guides for that? 2024-04-03 08:40:18 Hello! 2024-04-03 13:54:31 regarding merged-usr, I keep thinking the same thing: bind mount /bin from the initfs onto the final / mount 2024-04-03 13:58:28 you don't have the plan 9 bind(1) 2024-04-03 14:01:10 mount --bind 2024-04-03 14:02:09 umm... I thought that work differently? 2024-04-03 14:02:20 Ermine: mount --bind is opaque, plan9 bind was transparent by default (like an overlayfs) 2024-04-03 14:02:40 though I missed why that kind of bind mounting /bin would be necessary 2024-04-03 14:03:39 I think mount --bind is what WhyNotHugo meant 2024-04-03 14:04:05 If that mount --bind is enough merge-usr would not exist 2024-04-03 14:05:09 union directory 2024-04-03 14:22:17 grawity: what does transparent and opaque mean in this context? 2024-04-03 16:00:54 Is there another way to ask for support for Alpine other than IRC and mailing list (registrations closed)? Somehow IRC doesn't make sense to me.. I closed and reopened it and I've lost all what has been written.. 2024-04-03 16:02:04 pijul: you can subscribe to the mailing list without registring 2024-04-03 16:03:19 But, the mailing list is not that active 2024-04-03 16:05:26 pijul: if you don't have a bouncer you can view irc logs here: https://irclogs.alpinelinux.org/ 2024-04-03 16:07:53 kinda offtopic, but i slightly wish Alpine had a Discourse instance, or just some search-engine-indexable place for support questions that isn't aports issues 2024-04-03 16:08:30 there was a forum.a.o, what happened to it? 2024-04-03 16:08:44 archived after disus 2024-04-03 16:08:47 disuse 2024-04-03 16:09:06 https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum/installation/guts-casino.html 2024-04-03 16:09:08 i see.. 2024-04-03 16:09:10 ( lol ) 2024-04-03 16:10:17 fluix: yes, I know, but somehow I wasn't able to find anything of what I wrote in the log which should have been of today.. 2024-04-03 16:12:03 It's there 2024-04-03 16:12:11 2024-04-03 05:56:04 is it good to ask support questions here? 2024-04-03 16:15:45 ikke: Yes, now I found it. But I see that there has been no answer to my last question? 2024-04-03 16:19:59 yep 2024-04-03 16:20:02 correct 2024-04-03 16:23:32 discourse supports mailing lists so in theory that could be 2-for-1, but discourse needs to own the mailing list for them to writable from the forum side 2024-04-03 16:27:02 How should I handle the modloop service? 2024-04-03 16:27:46 I installed Alpine on Raspberry Pi 2024-04-03 16:28:47 Fxzxmic: what about it? 2024-04-03 16:29:18 However, the boot partition where modloop is located may fail to start due to not being mounted. 2024-04-03 16:30:31 I wrote a boot partition mount in fstab, but it seems too late. 2024-04-03 16:34:18 ikke: Do you think it is possible? Any place I should read myself into? 2024-04-03 16:43:43 pijul: have you seen this page: https://wiki.alpinelinux.org/wiki/GNOME ? 2024-04-03 16:47:15 Just a moment ago.. what about distrobox? Here maybe: https://wiki.alpinelinux.org/wiki/Running_glibc_programs 2024-04-03 16:50:10 there is a distrobox package 2024-04-03 16:55:29 OH, I see. 2024-04-03 16:56:16 It said "we should not try start modloop on sys install" 2024-04-03 16:56:28 in setup-disk 2024-04-03 16:56:42 Why do you want modloop on a sys install? 2024-04-03 16:57:09 It's not what I want. 2024-04-03 16:57:31 It's just not explicitly stated in any documents. 2024-04-03 16:58:33 modloop is something that is used on run-from-ram installations to provide kernel modules and firmwhere while preventing these files from taking up a lot of memory 2024-04-03 17:00:20 o/ 2024-04-03 17:00:30 I have a strange issue regarding mount-ro at shutdown 2024-04-03 17:00:51 I see it being called if I turn openrc verbose but I don't see any messages from it trying to unmount my filesystems 2024-04-03 17:01:01 Can I directly delete the modloop-rpi file now? 2024-04-03 17:01:02 as such at boot I always get a fsck.fat warning of being dirty 2024-04-03 17:18:47 dok: Is the link right for distrobox? 2024-04-03 17:20:37 pijul: i don't know, i never tried it. but looks ok to me 2024-04-03 20:04:53 Sorry for bad question I have x220 I want alpine on but no wifi? I don't know what package has the iwlwifi... 2024-04-03 20:10:43 Sorry it solve 2024-04-03 20:11:02 a lesson for us all 2024-04-03 20:54:05 umount doesn't work with fuse mounts. why not? 2024-04-03 20:54:13 i know i can use fusermount, but things like nemo use umount 2024-04-03 21:26:50 WhyNotHugo: why doesn't it? 2024-04-03 21:27:05 command output would be helpful 2024-04-04 04:07:58 Whats the busybox equivalent to fc? 2024-04-04 07:23:28 WhyNotHugo: transparent = you can see original files through the mount, i.e. like an automatic overlayfs 2024-04-04 11:06:03 dwfreed: umount: can't unmount /path/to/fuse/mount: Operation not permitted 2024-04-04 11:06:38 are you running umount as the mounting user? 2024-04-04 11:06:52 yes 2024-04-04 11:07:26 i tried both for encfs and sshf 2024-04-04 11:07:27 s 2024-04-04 11:10:33 util-linux mentions support for unmounting fuse as of 2.34. Maybe busybox doesn't support this? 2024-04-04 11:38:38 Hi there, I'm trying to use GNU Octave with latex but it seems that the required "dvisvgm" binary is missing - is there any chance it can be included/added? 2024-04-04 11:39:27 bananicorn: Best way to get that to happen is to either make an MR, or open up an issue (https://gitlab.alpinelinux.org/alpine/aports) 2024-04-04 11:39:27 Other distros either have a texlive-dvisvgm or a dvisvgm package available that includes this binary, to give an example 2024-04-04 11:40:57 ikke: Thanks, will do :) Would it make sense to contact the maintainer of texlive in a case like this? 2024-04-04 11:41:14 bananicorn: that will happen when you do either of those 2 things :) 2024-04-04 11:41:51 Alright, I will create an issue then, since I am unfortunately lacking the time for packaging it myself at the moment :) 2024-04-04 11:47:58 iwd offloads a lot of work to the wireless chipset, and in the past wpa_supplicant implemented some functionality in software instead. is this still the case with wpa_supplicant? 2024-04-04 11:55:40 hi 2024-04-04 11:55:55 WhyNotHugo: wpa_supplicant only implements that which it needs to; everything else is done by the chipset or the kernel 2024-04-04 12:35:37 Hi all, I am trying to install openjdk on alphine 3.18.6 but it seems to fail, I can install it on 3.19.1. I can't see any restrictions or something? 2024-04-04 12:46:12 what is the exact error you get? 2024-04-04 12:46:17 sounds like a bug 2024-04-04 12:49:36 openjdk21 (no such package) 2024-04-04 12:49:48 tried an alpine 3.19.1 image and it worked 2024-04-04 13:02:46 3.18's highest openjdk version is 17 -> openjdk17 2024-04-04 13:09:10 autojoin add 2024-04-04 15:46:55 aron, where can I see that ? 2024-04-04 15:53:49 is the tcp wrappers (libwrap0 on debian) package available on alpine? 2024-04-04 16:11:00 HegelianFascist[m]: I don't think we have a package for that 2024-04-04 16:13:26 HegelianFascist[m]: besides being compatible with the TCP wrapper config from other distros/ hosts, why not use nf/ ip-tables? 2024-04-04 16:13:50 otherwise, did you have a in depth look into the packages provided https://pkgs.alpinelinux.org/ ? 2024-04-04 16:16:00 theJoker8814: i need it for compiling https://github.com/chaos/diod 2024-04-04 16:17:51 I see some packages using --without-tcp-wrappers 2024-04-04 16:18:03 or --disable-tcp-wrappers 2024-04-04 16:19:03 The question is if DIOD uses tcp_wrappers or any function from the library. 2024-04-04 16:19:51 HegelianFascist[m]: without specifics, why not just compile it using debian, in case tcp_wrappers are required and used by diod. 2024-04-04 16:21:16 or are you planning to build a package for use with alpine after compiling? 2024-04-04 16:29:28 ikke: diod seems to require tcp_wrappers for the build unfortunately 2024-04-04 16:30:08 theJoker8814: I need a static binary for another machine and glibc doesn't really play well with static binaries 2024-04-04 16:31:40 I'm a bit rusty on my c. For sake of argument, compile DIOD on Debian (or any other distro with a packaged version of TCP wrapper); can't you configure your compile tool chain to link libraries statically, if DIOD really requires TCP wrappers lib. Then package all files required by TCP wrappers (i.e. host.allow, host.deny) and run it on alpine 2024-04-04 16:31:40 (same arch)? 2024-04-04 16:32:06 alpine uses musl as far as I'm aware. 2024-04-04 16:32:53 yes, musl is the best option when building static binaries nowadays 2024-04-04 16:33:04 glibc errors out half the time 2024-04-04 16:33:09 so build it on debian using musl 2024-04-04 16:33:23 and then copy the binary to alpine + tcp required files. 2024-04-04 16:33:51 if you want to make it clean for the future build a tcp_wrappers package for alpine :) 2024-04-04 16:41:24 HegelianFascist[m]: just a naive quick search in the DOID repository, revealed the wrapper is included libdiod/diod_sock.c and there is a knob to turn it off "HAVE_TCP_WRAPPERS", because it's only used to echo an error if TCP wrappers deny access. Which cannot happen on systems that don't have TCP wrappers. 2024-04-04 16:49:06 HegelianFascist[m]: maybe it's too obvious, but have you tried to compile it on alpine without TCP wrappers? 2024-04-04 16:51:03 oh it turns out I'm just dumb 2024-04-04 16:51:52 I tried compiling without another dependency and it errored out but it compiles now 2024-04-04 17:04:08 Dex2x: https://pkgs.alpinelinux.org/ 2024-04-04 19:11:42 Ello all 2024-04-04 19:12:41 Where does openldap look for the CA certifcates in the default file location? 2024-04-04 19:23:12 NM 2024-04-04 19:50:03 Hey all, I'm trying to build a go based pkg (newrelic infra agent). I can manually compile it but when I try with my APKBUILD on the same host I get "flag provided but not defined: -Wl,--as-needed,-O1,--sort-common" then it spits out the help/usage for /usr/lib/go/pkg/tool/linux_amd64/link. I dont know much about go, what could be different / effecting that in the abuild env? 2024-04-04 19:58:51 mintyhippoxyz: those are flags comming from CFLAGS 2024-04-04 21:49:31 where do openrc's service logging go? 2024-04-04 21:56:03 if you're using rc_logger, wherever rc_log_path is set to, defaults to /var/log/rc.log 2024-04-04 21:58:45 is rc_logger not being used out of teh box 2024-04-04 22:01:19 no 2024-04-04 22:02:08 also it's worth noting that rc_logger only captures the initscript output, not the actual daemon output 2024-04-04 22:05:17 awh what 2024-04-04 22:05:22 thats boring 2024-04-04 22:05:59 daemons should be doing their own logging 2024-04-04 22:06:05 there's a reason syslog was invented 2024-04-04 22:06:20 what is syslog solving? 2024-04-04 22:07:26 logging, like it has for last several decades 2024-04-04 22:12:09 for daemons that don't do their own logging 2024-04-04 22:12:48 hand them a cluebat 2024-04-04 22:12:54 hell, shell out to 'logger' 2024-04-04 22:13:27 shouldn't there be a solution that gathers daemons logging? i don't feel like each daemon doing its own logging is a great idea when it comes to administration 2024-04-04 22:13:42 that's why they log to syslog... 2024-04-04 22:14:05 why not make it syslog implementation independent and just use stderr 2024-04-04 22:14:32 syslog is implementation dependent already 2024-04-04 22:14:41 i see 2024-04-04 22:14:42 there's like 3 or 4 different syslog implementations 2024-04-04 22:14:53 s/dependent/independent/ 2024-04-04 22:15:09 but why do programs need to shell out to a syslog if stderr is easier 2024-04-04 22:20:09 they don't, there's literally a library call that handles all of this already, and basically every useful programming language has support for syslog 2024-04-04 22:20:33 but shelling out to logger is a great fallback, especially if you're writing in shellin the first place 2024-04-04 23:04:23 Hello All. How is everyone? 2024-04-04 23:06:41 i'm good. how are you? 2024-04-04 23:08:53 I'm still alive. And I got food, and maybe even a beer hiding in the back of the fridge, so I guess I am doing ok. 2024-04-04 23:22:48 I have been worse. 2024-04-04 23:24:39 Can anyone tell me if Alpine will continue to support 32-bit? 2024-04-04 23:26:11 I know it's basically paperweight-class hardware, but I do have a couple machines that are old x86 boxes that I have a sentimental attachment to. 2024-04-04 23:27:49 I ask because it seems a lot of distros seem to be considering dropping support for 32-bit, and not without good cause, really. 2024-04-04 23:55:36 m1rr0r5h4d35: there are no plans to drop 32bit support 2024-04-04 23:59:13 aron, I don't see a way to see what version of alpine that package supports? 2024-04-05 00:28:49 Nice! 2024-04-05 00:32:27 I can keep my netbook and HP tablet thing a little longer. 2024-04-05 00:35:37 Not that I can squeez a great deal of meaningful work out of them. 2024-04-05 00:38:17 Anyone here into retrogaming handhelds? 2024-04-05 05:01:14 Dex2x: you can select your desired release, 3.18 in this case, your arch (x86_64 I presume, but 99.99% it does not matter), and search for pattern "openjdk*" 2024-04-05 05:01:21 like this: 2024-04-05 05:01:28 https://pkgs.alpinelinux.org/packages?name=openjdk*&branch=v3.18&repo=&arch=x86_64&maintainer= 2024-04-05 05:01:57 if you browse that list, you'll see different versions (2 only pages, so not that long) 2024-04-05 05:02:16 -> openjdk17 is the highest 2024-04-05 05:04:21 (sorry for the typos, it's early morning here) 2024-04-05 08:38:07 hello, i'm having a minor issue with alpine stable 2024-04-05 08:38:40 when i install both inetutils-ftp and inetutils-telnet apk starts throwing an error, and it reports it every time 2024-04-05 08:40:01 ERROR: inetutils-telnet-doc-2.4-r0: trying to overwrite usr/share/info/inetutils.info owned by inetutils-ftp-doc-2.4-r0. 2024-04-05 08:40:47 to reproduce it you just # apk add inetutils-ftp inetutils-telnet 2024-04-05 08:41:05 al_: would it be possible to report this on https://gitlab.alpinelinux.org/alpine/aports/-/issues 2024-04-05 08:41:53 sure, just let me make a gitlab account. 2024-04-05 08:46:45 oh yeah i forgot 2024-04-05 08:46:58 this only happens if you have the docs package installed 2024-04-05 08:47:15 because it pulls both inetutils-ftp-doc and inetutils-telnet-doc 2024-04-05 08:47:27 if you don't have one of those it does't happen 2024-04-05 09:57:31 done, issue created. 2024-04-05 09:57:33 https://gitlab.alpinelinux.org/alpine/aports/-/issues/15957 2024-04-05 10:08:59 Fixed and issue closed. Sorry for bothering. I should have read the documentation. 2024-04-05 10:29:36 ikke, don't forget to reopen that ticket 2024-04-05 10:30:04 done 2024-04-05 10:30:13 (y) 2024-04-05 16:34:40 hello all 2024-04-05 16:36:21 I have a question about a package that seems to have missing deps. The package is Freecad, which is currently "flagged out of date" and can not install because the package "py3-pyside2". Is there anyone here who could give some hints to workaround? 2024-04-05 16:37:33 I was also looking for a way to install librecad, but no package seems to be available. How hard is it for a non-coder to get that working in Alpine? 2024-04-05 16:39:52 you don't necessarily be a coder, you do need to know how build systems work 2024-04-05 16:41:56 I built (compiled) some programs like xst (an st variant) and a few other c programs, but I got confused as hell with the aports documentation so was able to make installs on some suckless type apps. Never gone beyond that though. 2024-04-05 16:43:23 Seems to be 2 packges that depend on pyside2 2024-04-05 16:43:24 is there an easy way from the "make" to easily port over to a build package before "make install"? Last I recall I had to set up a whole chroot system and I didn't understand the context of the instructions 2024-04-05 16:43:38 6b057ac70e027350666b0be5f0c06c15e4e2fee9 2024-04-05 16:44:26 oh oh ... that means no freecad? I wonder what they'll do as a replacement? 2024-04-05 16:46:13 I imagine one could create a pyside2 package fairly easily if it's just python script? 2024-04-05 16:47:02 Not only pyside2, but also shiboken2 was removed 2024-04-05 16:47:28 right, two packages 2024-04-05 16:50:57 is there an easy way to take the apkbuild file and just run an install? These packages seem to be just python scripts 2024-04-05 16:52:33 talking about freecad?> 2024-04-05 16:52:34 Also, what about "appimage" on alpine? Is this possible? 2024-04-05 16:53:06 ikke: talking about the two missing packages. Freecad is in the system, but can't install because of the reported pyside2 package 2024-04-05 16:53:12 Don't know appimage 2024-04-05 16:53:37 SagestMage: THe reason they got removed is probably because they have build issues, either now, or with python3.12 that will soon be pushed 2024-04-05 16:59:49 Seems the two packages are very interrelated https://doc.qt.io/qtforpython-5/ 2024-04-05 17:00:01 yes 2024-04-05 17:04:15 And there is a third package testing/py3-pivy that is affected as well 2024-04-05 17:05:32 it seems that pyside2 version is way out of date, there is one from last month. Perhaps there is a move from qt5 to qt6 with these apps. 2024-04-05 17:06:04 I wish I had time to fiddle with getting a build system up. 2024-04-05 17:06:48 The latest is pyside6 2024-04-05 17:10:39 SagestMage: there are docker images that help with it 2024-04-05 17:15:57 ikke: what docker images? 2024-04-05 17:16:17 alpinelinux/dabuild 2024-04-05 17:16:44 I must admit, I don't really understand docker (nor appimage, snap, etc) 2024-04-05 17:17:03 docker is a chroot on steroids 2024-04-05 17:22:04 So, you're saying there's a way to use docker images as build setups? 2024-04-05 17:25:20 yes 2024-04-05 17:25:26 our CI builds packages in docker 2024-04-05 18:22:34 hey everyone! is anybody familiar with nlplug-findfs? 2024-04-05 18:23:39 trying to discover disks in init with the following command: "find /sys/ -name modalias | xargs sort -u | xargs -n 1 modprobe 2> /dev/null" 2024-04-05 18:24:10 and it does not detect some disks (works fine for virtio, but not for a regular usb stick on a physical host) 2024-04-05 18:24:41 while nlplug-findfs works everywhere, but also does stuff i do not want (auto mounts drives etc) 2024-04-05 18:24:56 i use it like this: "nlplug-findfs -p /sbin/mdev -a /dev/null;" 2024-04-05 18:25:12 this will detect all drives all the time, but cant stop it from mounting them 2024-04-06 00:36:43 you can read the code https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/nlplug-findfs/nlplug-findfs.c 2024-04-06 00:36:58 atm0s: probably you should just call mdev directly 2024-04-06 02:25:09 Hello71: thanks, will try the direct mdev call! 2024-04-06 07:33:26 why on earth is partprobe not refreshing disks 2024-04-06 07:33:43 kernel /sys and dmesg say partitions are detected but in /dev its missing 2024-04-06 07:36:17 that's [um]dev that should create those entries 2024-04-06 07:48:37 ikke: its occurring in the live iso 2024-04-06 07:48:49 it is very frustrating to have to reboot but this has to have some solution 2024-04-06 08:23:28 it was mdev -s 2024-04-06 14:31:17 does udhcpc only run once at default service level? 2024-04-06 14:31:40 or will it use udhcpd 2024-04-06 18:38:22 hey, i want boot a diskless alpine system but instead of loading an apkovl (and installing packages into a ramdisk) i want to mount a squashfs (+ tmpfs write overlay). has anyone tried that before? i read through the init script from mkinitfs (very lean early boot process, i like it), and it looks like it'd just work if i had a parition with a squashfs on it mounted. could 2024-04-06 18:38:24 nlplug-findfs be used to find and mount a filesystem with a squashfs file on it? 2024-04-06 18:58:49 minus: i'm doing something like that in https://gitlab.alpinelinux.org/ptrcnull/alpine-live, but that uses a modified init 2024-04-06 18:59:23 that said, if you don't mind hand-editing /usr/share/mkinitfs/initramfs-init, it would be very much possible 2024-04-06 19:02:37 oh cool, that's very nice as a reference 2024-04-06 19:03:19 i think i don't quite know what i'm aiming for exactly yet 2024-04-06 19:04:01 well, i do know that the part of installing apks to ram makes my raspberry pi run out of ram 2024-04-06 19:04:46 would be cool if lbu built a squashfs that just gets mounted instead of a tar that gets extracted then 2024-04-06 19:05:21 so if i'd want to achieve that there's no way around changing a bunch of stuff in alpine 2024-04-06 19:06:32 yeah, that's kinda what i was aiming for 2024-04-06 19:06:52 as in, running some images with graphical environments and the packages not taking up most of the ram 2024-04-06 19:07:38 but saving "snapshots" to squashfs images instead of prebuilding them.. that would require some mildly cursed scriptery 2024-04-06 19:08:02 tar2sqfs, interesting 2024-04-07 01:33:01 Alpine Linux fucks. Favorite distro. 2024-04-07 01:33:33 Thanks devs for your continued hard work :) 2024-04-07 05:24:59 Is there any way to install alpine linux so that the wifi works while using a wifi connection? Because even by using the extended iso for the installation, it cannot find the packages iwlist and wpa_supplicant anywhere.. 2024-04-07 05:25:19 also trying to mount the USB drive where the extended iso is, it complains that it can't find the device in /etc/fstab 2024-04-07 05:25:50 When I mount drives in Alpine I always have to use the 2024-04-07 05:26:04 -t flag, to indicate the filesystem type 2024-04-07 05:27:31 > mount -t ext4 /dev/sda1 /mnt/usb 2024-04-07 05:37:12 gamingirc: busybox mount does not have a way to detect the filesystem to automatically load the correct module 2024-04-07 05:37:30 mount should be util-linux, though? 2024-04-07 05:37:47 according to https://pkgs.alpinelinux.org/contents?file=mount&path=&name=&branch=edge 2024-04-07 05:38:02 dwfreed: busybox applets are not listed there 2024-04-07 05:38:09 ah, dumb 2024-04-07 05:38:16 they're symlinks 2024-04-07 05:38:37 they're still "files" :P 2024-04-07 05:38:46 But not part of the package 2024-04-07 06:02:05 Hi, does anyone if alpine has something like void's xlocate, for finding out what package provides a file? 2024-04-07 06:02:52 No, the only thing that has an index of files to packages is pkgs.alpinelinux.org 2024-04-07 06:03:08 (There are some unofficial tools that scrape it though) 2024-04-07 06:05:39 Oh, https://pkgs.alpinelinux.org/contents I suppose could work. Thanks 2024-04-07 06:14:34 I have the same problem by using the -t option for mount.. it still complains that it cannot find the device in fstab even though blkid finds all the devices flawlessy. 2024-04-07 06:18:49 pijul: that sounds like you are trying to mount as non-root 2024-04-07 06:21:16 or trying to mount by target directory 2024-04-07 06:21:33 (missing the source device, so it will try to look it up in fstab) 2024-04-07 06:23:46 Anyone know if there's a package that provides (erlang) mix? erlang-dev doesn't and I'm out of ideas 2024-04-07 06:28:42 https://pkgs.alpinelinux.org/contents?file=mix&path=&name=&branch=edge&arch=x86_64 2024-04-07 06:36:13 Oh that's how you use it. Thanks :D 2024-04-07 07:05:09 pijul: I had that issue yesterday but i fixed it by restarting 2024-04-07 07:05:24 kept telling me "No device found" when trying to mount a usb, event with the -t flag 2024-04-07 07:08:13 Yeah, I did restart and the problem persisted for me.. 2024-04-07 07:24:29 pijul: what command do you run? 2024-04-07 07:27:12 ikke: For example, mount -t filesystemtype /dev/sdg1 /mnt/usb 2024-04-07 07:28:22 And what is the exact error you get? 2024-04-07 07:38:03 some USB drives needs delay option, and some simply don't mount during boot 2024-04-07 07:39:26 i have sony-32-gb(made in india) that simply won't mount during boot 2024-04-07 07:39:59 and transdent 8gb - that may mount sometimes 2024-04-07 07:40:35 transcend^ 2024-04-07 07:47:43 kingston, pny, samsung do nice 2024-04-07 08:18:16 ikke: The error I'm getting is: mount: can't find /dev/sdbX in /etc/stab 2024-04-07 08:19:10 does /dev/sdbX exist? 2024-04-07 08:19:54 (that message can only appear if you do not specify the source device for the mount) 2024-04-07 08:26:16 ikke: Yes, blkid finds it. 2024-04-07 08:27:16 satmd: I did specify it and just for the sake of it, I tried with the other devices too and the error message was all the times, the same.. 2024-04-07 08:31:28 Hello. I tried to install GUI as described here: https://gitlab.alpinelinux.org/alpine/docs/user-handbook/-/blob/master/modules/Working/pages/post-install.adoc#user-content-getting-a-graphical-environment 2024-04-07 08:32:07 But OS complains about missing alpine-desktop package. Please say what correct command is and update documentation 2024-04-07 08:34:18 the setup-desktop script should be used instead 2024-04-07 08:34:28 59ed4dcbaef9dec859eb014a003cb7a4c89780b0 2024-04-07 08:37:06 ikke: thank you, I will try it. please correct documentation if you can 2024-04-07 08:39:14 Vort: would you mind opening an issue here: https://gitlab.alpinelinux.org/alpine/docs/user-handbook? 2024-04-07 08:41:25 yes, I will, but slightly later 2024-04-07 08:42:33 ikke: oh, it is already done: https://gitlab.alpinelinux.org/alpine/docs/user-handbook/-/issues/2 2024-04-07 08:46:30 Changed my setup so I no longer have doas (or sudo), removed my user from wheel, and enabled autologin on tty1 for my user, and autologin on root for tty2 2024-04-07 08:47:24 Once i enter encryption password I just get shoved to a desktop, anytime I need root I just switch to tty2, no password needed. Convenient yet secure. 2024-04-07 08:48:58 Could even lock the root account if I wanted, autologin still works 2024-04-07 08:52:50 only secure if you never let your device alone :P 2024-04-07 08:56:41 There are two states that matter. Encrypted, or not. If you leave it encrypted and someone gets to it. You are likely safe from them collecting data, but they might leave keyloggers so best to ditch it. Or its unencrypted, in which case having the root terminal doesnt matter as the data can easily be read. 2024-04-07 08:57:29 In my threat model, physical possesion is game over unless encrypted. So root terminal has zero impact on security. 2024-04-07 08:57:56 while running, you should treat it as unencrypted 2024-04-07 08:58:14 Duh. 2024-04-07 09:16:38 why "mount: can't find /dev/sdbX" ... and and not sdaX .. if using more than one USB plugout all and only try with bootable on 2024-04-07 09:18:01 always switch to base mode when trying to find issuue, 1. use a reference bootable usb , 2. plugout all and try only one 2024-04-07 09:19:22 reference bootable usb - is one that you have created and worked on most devices or got madef from a tech guy 2024-04-07 09:20:18 and have couple of them, a "2gb sandisk" that i created 10+yrs back 2024-04-07 09:20:32 i have couple of them^ 2024-04-07 15:31:28 Hey, I have opened a PR to update forgejo (!63556). Version 1.21.10-0 includes a security fix for CVE-2023-45288 and there is a "secfixes" section in the APKBUILD. How is that field supposed to be handled? 2024-04-07 15:32:41 famfo: you add an entry for the version that fixes the CVE 2024-04-07 15:32:51 New entries are added on top 2024-04-07 15:33:26 okay, ty 2024-04-07 15:34:14 As an FYI, the secfixes are collected here: https://secdb.alpinelinux.org/edge/community.json 2024-04-07 15:34:30 Oh, but not for testing 2024-04-07 17:01:11 So, the only thing I could try based on what I'm reading from the log, is unplugging all other usb devices and see if that way mount can see the other devices? 2024-04-07 17:01:47 But then why can it not see the other non-usb internal devices? 2024-04-07 17:29:32 pijul: did you try mount from util-linux 2024-04-07 17:41:19 ikke: I don't think so. I would that be? 2024-04-07 17:49:12 *how 2024-04-07 18:02:36 apk add util-linux 2024-04-08 09:37:28 Pfff 2024-04-08 09:37:46 Rebooting after an update: start-stop-daemon: /sbin/dhcpcd does not exist 2024-04-08 09:38:01 yeah, because it's not installed 2024-04-08 09:38:52 Why would rc try to start it 2024-04-08 09:39:14 rc-status doesn't list it 2024-04-08 09:42:14 Any idea? 2024-04-08 09:43:17 If I do a rc-service networking restart, it works without issue, no mention of dhcpcd 2024-04-08 09:45:33 Of course, that doesn't restart unbound though 2024-04-08 10:42:04 Is there an update guide on how to actually make the wifi work? Because the one here: https://wiki.alpinelinux.org/wiki/Wi-Fi doesn't go very far.. the linux-firmware package installs with MANY errors and the way that guide makes you configure wpa_supplicant.conf doesn't work, the one here: https://venenux.github.io/alpine-wiki/tutorials/alpine-tutorial-wifi-routering.html works better, but it's apparently still not enough 2024-04-08 10:42:56 pijul: did you try setup-interfaces? 2024-04-08 11:23:57 ikke: No. Where is written about it? Isn't that setup included in setup-alpine? 2024-04-08 11:38:35 pijul: yes, it is 2024-04-08 11:39:17 setup-interfaces tries apparently to install iw and gets: "iw (no such package) required by: world[iw]" 2024-04-08 14:47:25 "setup-interfaces tries apparentl..." <- I think there's something wrong with your mirror or you need to refresh them 2024-04-08 17:17:12 moonzero[m]: How? I couldn't connect to any online mirror while setup, because I couldn't connect to the internet, since I never managed to establish a wifi connection following a few different guides, including the official one.. 2024-04-08 17:19:02 pijul: did you use the standard or extended image? 2024-04-08 17:36:53 ikke: extended 2024-04-08 17:38:16 both standard and extended should have iw 2024-04-08 17:39:45 It seems like they don't. I also verfied the sha512 before writing the image to an usb disk.. and tried executing the iw command separately.. 2024-04-08 17:39:53 What do you suggest? 2024-04-08 19:52:08 would it be ok to send a log file as plain text in a message? I have an issue with xorg, most likely its my fault and I haven't installed something or haven't set up something but I don't know what and would be very thankful for any advice. 2024-04-08 19:52:28 use a pastebin 2024-04-08 19:52:31 ^ 2024-04-08 19:53:39 wouldn't the matrix bridge send a link with the full message anyway? 2024-04-08 19:54:38 postgresql-common-openrc doesn't have the executabe itself as a dependency, so it fails to start 2024-04-08 19:54:50 is this intentional? 2024-04-08 19:55:09 triallax: yes, not in a very helpful way: "Foo sent a long message: " 2024-04-08 19:55:24 yes but that's not much different from sending a normal pastebin link 2024-04-08 19:55:40 If you just paste a link, no, but usually you give some context 2024-04-08 19:56:43 same caveats apply to either, but i digress 2024-04-08 19:57:01 https://pastebin.com/qP406Mva 2024-04-08 19:57:46 what's the issue exactly? 2024-04-08 19:58:06 it crushes and won't start 2024-04-08 19:58:52 or is it because I haven't told it what wm or de to use? it should default to the ugly one anyway right? 2024-04-08 19:59:01 ugly one? 2024-04-08 19:59:23 what's the default wm of x11? 2024-04-08 19:59:27 what are you running? startx? 2024-04-08 19:59:29 was it TWM? 2024-04-08 19:59:31 there isn't a "default wm" 2024-04-08 19:59:53 triallax: yes 2024-04-08 20:00:13 i haven't used xorg in a long time but do you have something in .xinitrc? 2024-04-08 20:00:26 ~/.xinitrc to be exact 2024-04-08 20:00:51 triallax: nope 2024-04-08 20:00:52 Damn, feels good to finally having networkmanager working on alpine 2024-04-08 20:01:18 shigetsune[m]: you need to have a wm installed and start it in .xinitrc 2024-04-08 20:01:25 a wm or de or whatever 2024-04-08 20:02:37 is there really no default wm coming with x11 in alpine? well considering how ugly it was I'm not loosing much, ok I'll try it out 2024-04-08 20:02:42 thank you 2024-04-08 20:02:51 i really have no idea what you're talking about 2024-04-08 20:02:57 maybe alpine does something different than usual, idk 2024-04-08 20:03:05 why would there be one, just install a package with a wm and have fun using it 2024-04-08 20:03:40 I assume you haven't used one of that do-it-yourself distros. Arch has a similar experience imo 2024-04-08 20:06:16 lassebq: actually I've used arch and freebsd, dunno about arch since I've straight off installed wm and never cared but at least on freebsd there were twm coming with x11, maybe just bsd thing tho, wouldn't be surprised 2024-04-08 20:06:37 oh no, wait I believe I've used it on void too 2024-04-08 20:06:43 pretty sure void doesn't have a default 2024-04-08 20:06:47 just bsd thing for sure 2024-04-08 20:07:36 there really shouldn't be a default wm since xorg doesn't have to depend on any wm. If you installed a wm package though, xorg would likely be fetched as a dep 2024-04-08 20:11:06 btw, lacking a wm was the issue, it launched fine when I specified it, I just wanted to make sure x11 is working before I configure wm and stuff 2024-04-08 20:11:33 sorry for causing trouble for nothing I guess 2024-04-08 20:11:35 you can pass it to to startx or do exec in ~/.xinitrc so you only need to run startx 2024-04-08 20:11:47 the latter is a better idea probably 2024-04-08 20:12:07 it's not a problem 2024-04-08 20:15:09 haha, I always found it so hard to set up Xorg on Alpine... so easy to "crash" the system by getting X to start but no keyboard or mouse... 2024-04-08 20:15:18 so I kinda gave up and just use Wayland, much easier... ;) 2024-04-08 20:15:33 at least Sway won't start up anyway if it can't find any input devices 2024-04-08 20:17:30 having said that - in above pastebin X server found input devices and video seems fine too... so all that's left is finding SOMETHING to start inside the session 2024-04-08 20:18:19 ah, you figured that out too by now as well 2024-04-08 20:19:03 personally I just really hate how startx loves bringing up an Xterm - in which you can't type or select anything because "oops you forgot one of the 5000 weird services that are needed for X to detect input devices" 2024-04-08 20:19:21 at which point all I can do is reboot using the power button (or ssh in) 2024-04-08 21:24:07 hez 2024-04-08 21:24:09 hey 2024-04-08 21:24:58 french speaker spotted 2024-04-08 21:26:46 one question, which app can I install to get auto adaption of app-windows, such that when I go with a window in the upper right corner that a window automatically adapts to this spot in xfce4, I have already xfwm4 but that seemed to not do the trick. 2024-04-08 21:29:20 triallax: more like german 2024-04-08 21:29:28 french don't use qwertz 2024-04-08 21:29:45 germans and czechs do 2024-04-08 21:30:06 french are weirder they have azerty 2024-04-08 21:30:15 oh right 2024-04-08 21:35:42 hey I wrecked my Internet with uncomplicated firewall xD 2024-04-08 21:38:52 omg why cant I connect to https and http when I connect the firewall 2024-04-08 21:39:15 did you add the rules to let the connections pass 2024-04-08 21:39:25 assuming they're denied by default 2024-04-08 21:40:02 https://pastebin.com/C0606Rxi 2024-04-08 21:40:07 this is how it looks like 2024-04-08 21:40:37 Hello 2024-04-08 21:40:53 did you do something like doas ufw allow out http 2024-04-08 21:40:55 looking for welp 2024-04-08 21:40:55 and same for https 2024-04-08 21:41:26 badb: what's going on? 2024-04-08 21:41:48 https://pastebin.com/h9f607BW triallax the list of commands I executed 2024-04-08 21:42:21 zcrayfish: when activating firewall I have no connection to http and https 2024-04-08 21:42:35 Hey z, I'm trying to decide which vetrsion of alpine shall I get. Extended or standart. I'm planning to use alpine with my potato home server pc for the purpose of add blocking dns by using docker container installed dns apps 2024-04-08 21:42:43 you denied all outgoing and are surprised that internet does not work? 2024-04-08 21:42:43 version* 2024-04-08 21:42:47 simon_: just try the commands i suggested 2024-04-08 21:42:58 q66: where did I denied them ? 2024-04-08 21:43:02 alright triallax thanks 2024-04-08 21:43:03 doas ufw default deny outgoing 2024-04-08 21:43:09 this denies all outgoing connections 2024-04-08 21:43:11 by default 2024-04-08 21:43:18 it defaultly denies oncoming 2024-04-08 21:43:25 oh 2024-04-08 21:43:27 try ufw status command 2024-04-08 21:43:47 https://pastebin.com/C0606Rxi badb 2024-04-08 21:43:51 badb: the the main difference between the two is if you are going to use them offline. if you're doing ad blocking then it surely will be online, thus the standard should be fine. 2024-04-08 21:45:46 okay, I got the answer. Also curious about online offline thing. Does extended comes as offline? confused 2024-04-08 21:47:06 extended has more packages which makes it more useful for machines that are going to be offline 2024-04-08 21:48:43 Actually home server machine I own has no wifi or network connection. Using extended looks way more better for me but in the download page it say "it runs from RAM" that is the main thing concerning me. I don't know the pros and cons of it. 2024-04-08 21:50:00 Sorry if i'm asking a lot and bothering you zcrayfish 2024-04-08 21:50:16 it's no problem at all; you can set it to run either way during setup 2024-04-08 21:50:50 what about runs from ram thing. Would it be more or less stable for a dns server specific usage. 2024-04-08 21:50:54 ? 2024-04-08 21:51:04 Also thanks for your kindness 2024-04-08 21:51:58 In my experience the stability is roughly the same. It really depends on you hardware though. 2024-04-08 21:52:27 it's a Dell wyse 7010 z90d7 pretty much a potato from 2013 2024-04-08 21:53:44 the DOM's on those are *slow* so the more RAM you have, the better. \o/ 2024-04-08 21:55:24 8 gigs of ram seem like going to a trouble for me so I'll be using standart version 2024-04-08 21:55:32 Thanks broski love ya 2024-04-08 21:55:42 np! I gotta run, laters!! 2024-04-08 23:22:35 hi again 2024-04-09 01:55:21 Hi, I am messing around with unshare, I suspect this is not particularly an Alpine issue, but after unpacking a minirootfs into /tmp/zram and running unshare -r -f /sbin/chroot /tmp/zram ./lib/ld-musl-x86_64.so.1 ./bin/busybox sh, I am "root" in the new chroot, but any call to apk add gets permission denied. 2024-04-09 02:01:29 OOI where does apk.static from https://gitlab.alpinelinux.org/api/v4/projects/5/packages/generic/v2.12.10/x86_64/apk.static come from? Can I pull this directly from a mirror instead of hitting gitlab? 2024-04-09 02:06:22 ah right, https://gitlab.alpinelinux.org/alpine/apk-tools 2024-04-09 05:00:57 I didn't see any answer to my question in the log - so, what should I try next? 2024-04-09 05:04:55 pijul: ask it again as some of us have no logs and no context? 2024-04-09 05:05:04 but how long did you wait? 2024-04-09 05:07:44 Basically trying to install iw throws the error "iw (no such package) required by: world[iw]".. I tried using the command setup-interfaces from the RAM and a sys installation and it' 2024-04-09 05:08:46 s always the same problem.. the wifi is needed to configure the mirrors and setup a desktopn environment, but it seems not to work. 2024-04-09 05:20:15 pijul: so, more context needed: what version of Alpine? How did you start the installer? What wifi card? What happens with setup-interfaces? I don't have a machine I can replicate this issue on, but would something like tethering a phone temporarily work? 2024-04-09 05:20:32 you may not have gotten an answer as there wasn't enough info to go on. 2024-04-09 05:27:35 pijul: just in case you come back to this, the alpine-extended ISO may also have the packages you need already. also the package is probably wireless-tools, not iw 2024-04-09 05:56:02 the package is `iw` but if you don't have a connection installing it isn't going to work. using the alpine-extended ISO is the right call 2024-04-09 06:28:31 thanks pukko, I couldn't find the name when I was looking 2024-04-09 06:32:53 Since Linux upgrade from 6.6.22, my laptop battery don't charge anymore 2024-04-09 06:33:15 It stays at whatever charge it was when AC is plugged in 2024-04-09 06:33:27 Note that I'm not sure if it's actually a kernel problem 2024-04-09 06:33:47 Did something change in battery charging handling? 2024-04-09 06:33:57 That's on a thinkpad X230 2024-04-09 06:34:59 does it charge when powered off? 2024-04-09 06:35:39 It does 2024-04-09 06:35:47 I can also see the ACPI event correctly: ac_adapter ACPI0003:00 00000080 00000001 2024-04-09 06:39:52 And it's discharging allright though :D 2024-04-09 06:40:37 And I'm not running things like tlp either 2024-04-09 06:41:21 (what I mean is that I haven't setup any third party battery charging model) 2024-04-09 06:43:58 fluix: I use the extended ISO already.. but I get that error thrown at me nonetheless 2024-04-09 06:45:32 [Tue Apr 9 08:44:02 2024] thinkpad_acpi: battery 1 registered (start 0, stop 100, behaviours: 0x7) 2024-04-09 06:46:39 pijul: do you have a network connection? what does /etc/apk/repositories look like? 2024-04-09 07:19:52 fluix: I have a wireless connection, not a wired connection. How should I see what is inside /etc/apk/repositories what is the command? Grep or something? 2024-04-09 07:22:00 err, `cat /etc/apk/repositories` 2024-04-09 07:25:06 fluix: Should I tried that for the live-usb before installing, or after install, or both? 2024-04-09 07:25:50 sorry, can you clarify the problem? did you install fine and now don't have internet *after* rebooting or what? 2024-04-09 07:44:12 fluix: No, I tried installing without completing the internet connection and mirror part, to see if by issue could be better solved from a sys installation. 2024-04-09 07:47:23 ah I've read the full backlog. I'm not sure why iw fails to install when it should be included in the images. probably good to report an issue on gitlab if it's definitely not in the image (this shouldn't need any kind of existing network connection or mirrors) 2024-04-09 07:47:48 sorry you've had to repeat this multiple times 2024-04-09 09:01:27 hey how can I grant my firewall premission to access oftc network chats 2024-04-09 09:02:12 do I need to open the oftc port in my firewall? 2024-04-09 09:03:38 for example ufw enable out 6693 2024-04-09 09:05:46 I type from phone because my browser can't access the webchat.oftc.net 2024-04-09 09:06:22 I mean it can access https but not oftc* now I got it 2024-04-09 09:06:25 :) 2024-04-09 09:06:55 fluix, pijul: can't repro on alpine-extended-3.19.1-x86_64.iso, do you mean you've installed from the ISO to a fixed disk and now can't install iw? 2024-04-09 09:11:05 is alpine not intended to be shutdowns ? I am stuck in booting process at "waiting for events to be processed..." 2024-04-09 09:11:30 voltagex: I did both. After installing on a disk and on the live-usb. Both had the same issue.. 2024-04-09 09:11:45 sigh 2024-04-09 09:18:18 I tried ufw enable out 6660:7000/tcp it adapted the rule but I have still no connection to oftc irc 2024-04-09 09:19:27 could you help by any chance 2024-04-09 09:19:58 simon_: what port are you trying to connect to oftc on? 2024-04-09 09:20:27 I just try to connect to the webchat.oftc 2024-04-09 09:20:37 if you disable the firewall temporarily does it work? 2024-04-09 09:20:44 yes 2024-04-09 09:20:58 simon_: webchat probably uses websockets 2024-04-09 09:21:10 webchat would be port 443, I would question the utility of restrictive outbound rules though. 2024-04-09 09:21:15 It does not directly connect to irc 2024-04-09 09:22:09 port 8443 is used by webchat 2024-04-09 09:23:10 yes some of them worked 2024-04-09 09:23:37 heyy it worked 2024-04-09 09:24:34 https://pastebin.com/VF3MUNBq can you spot any security risks in my ufw status? 2024-04-09 09:24:37 FWIW restrictive outbound rules will make your life difficult and I would go back to thinking about what threat/s you are trying to mitigate 2024-04-09 09:24:50 security risks from what? from whom? 2024-04-09 09:25:23 idk just a basic knowledge question I guess, maybe I missconfigured something what could be a vulnerability? 2024-04-09 09:26:40 if I was doing a *risk analysis* I would say that the main risk here is something not working (say, VoIP, video calls) due to overly restrictive firewall policies. This (except for port 123 and 6000+) is more what I would expect from a kiosk or locked down corporate box. 2024-04-09 09:27:33 from an "infosec hat" point of view, any malware is just going to go out on port 443 anyway so this wouldn't stop it. 2024-04-09 09:29:12 can I close port 443? 2024-04-09 09:30:33 so you suggest closing 443 and 123/udp anad 666:700/tcp? 2024-04-09 09:30:59 or where did you see 6000+ 2024-04-09 09:32:48 ohhh, and another question does gcompat allow glibc apps? like appimages and such? 2024-04-09 09:42:45 simon__: what do you think will happen when you close port 443? 2024-04-09 10:14:30 hey how can I adapt pipewire in alpine linux 2024-04-09 10:25:52 simon_: what do you mean by 'adapt'? 2024-04-09 10:26:19 like installing and using 2024-04-09 10:36:56 simon_: please don't DM people without permission, also it is very hard to keep track when you keep quitting and rejoining as different usernames. A lot of the questions you are asking suggest you need to do some more research. Think about what your end goal is and work backwards. 2024-04-09 17:38:12 If /sys exposes a superset of the information in /dev/, why do we keep using /dev for everything? 2024-04-09 17:38:37 Like 'mount /sys/class/block/nvme0n1p1/dev /mnt' makes sense to me 2024-04-09 17:46:47 WhyNotHugo: it's magic. it's what attributes are exportable to userspace, which is not everything. the model for the fs is basically sysctl 2024-04-09 17:53:10 invoked: what can't be exported via the filesystem? 2024-04-09 17:55:27 there's lots, at least with respect to being writable, that can't be exported from drivers 2024-04-09 18:00:43 the question is why we have both magic file systems—sysfs and devtmpfs 2024-04-09 18:01:51 the answer is probably “it’s always been like that” (/dev, not devtmpfs) 2024-04-09 18:03:55 heyaa 2024-04-09 18:16:59 yeah, i have the impression that historical reasons is the answer to most of hese question 2024-04-09 18:21:39 it's complicated. say you have a storage device. it has structures it registers with the system bus, but those belong to the system bus, not the driver. the driver can't (or shouldn't) export stuff it doesn't own 2024-04-09 18:22:34 sysfs is just a view of what can be readable and writable from userspace 2024-04-09 18:27:49 does sysfs give you some equivalent of /dev/serial/by-id ? 2024-04-09 18:29:25 guys I have a question is it better to compile a pkg my self or to use an appimage in alpine? If you say I should compile it I also would need to compile https://github.com/moonrepo/proto heh 2024-04-09 18:29:25 :D 2024-04-09 18:30:02 Habbie: probably under /sys/class somewhere. 2024-04-09 18:30:35 simon_: appimages don't work in alpine unfortunately 2024-04-09 18:30:58 does a proto pkg exist by any chance 2024-04-09 18:30:58 ? 2024-04-09 18:31:29 i don't see one 2024-04-09 18:35:42 proto? 2024-04-09 18:35:57 oh, i see the link 2024-04-09 18:35:58 don't think so 2024-04-09 18:36:53 but then, if it downloads random executables from the internet, that's not gonna work on Alpine 2024-04-09 18:37:16 well, they might, if they were built on alpine 2024-04-09 18:37:40 ptrc: does it do this? 2024-04-09 18:38:25 i'm looking through the sources of moonrepo/proto, because obviously it would be too much effort to document how it works 2024-04-09 18:39:03 nope, they match on architecture and OS (linux/darwin/windows), but not on libc 2024-04-09 18:39:15 so you're just gonna get broken glibc executables 2024-04-09 18:39:42 so proto will not work? 2024-04-09 18:41:47 you can try gcompat but probably not. 2024-04-09 18:43:25 proto itself works, but it's useless 2024-04-09 18:43:26 https://ptrc.gay/FycCFBGA 2024-04-09 18:43:59 sad 2024-04-09 18:44:00 D; 2024-04-09 18:44:08 what's this proto thing 2024-04-09 18:44:16 packager, looks like 2024-04-09 18:44:19 found it 2024-04-09 18:44:53 a wrapper over wget + tar xf 2024-04-09 18:44:55 gives me corporate vibes 2024-04-09 18:44:59 /j 2024-04-09 18:45:17 excellent way to herd people into being owned, that 2024-04-09 18:48:38 so I should just use another computer for appimages than alpine 2024-04-09 18:48:39 ? 2024-04-09 18:48:52 https://github.com/LedgerHQ/ledger-live I tried compiling this 2024-04-09 18:49:15 if you want to use appimages you'll have to. that's a glibc thing 2024-04-09 18:49:31 would they work in a chroot 2024-04-09 18:49:39 never actually tried that 2024-04-09 18:49:41 in some cases you can use flatpaks instead of appimages, Flatpak works on Alpine just fine 2024-04-09 18:49:42 and that doesn't mean you have to use appimages 2024-04-09 18:50:10 triallax: iirc appimages do some fuse mount shenanigans 2024-04-09 18:50:14 so i heard 2024-04-09 18:51:16 but huh, apparently they work..? 2024-04-09 18:52:25 in a chroot you also get fun perks as "chromium sandbox is broken af" 2024-04-09 18:52:42 i don't use a real chroot 2024-04-09 18:52:43 bwrap ftw 2024-04-09 18:53:55 https://ptrc.gay/sOZXZmXl 2024-04-09 18:54:56 even if you could get this proto thing to download binaries that run under basic usage i wouldn 2024-04-09 18:55:03 't rely on them 2024-04-09 18:55:24 yeah this is (at worst) a security vulnerability wrapped in a fancy tui presentation 2024-04-09 18:55:41 and at best just a wrapper over "download a tarball from the website and extract it into /usr/local" 2024-04-09 18:56:28 people are so keen to add layers that don't add any value and just make stuff more complicated. 2024-04-09 18:56:50 the one tool to end all tools 2024-04-09 19:19:09 can I not just install glibc on alpine 2024-04-09 19:19:19 to run appimages 2024-04-09 19:22:16 no. 2024-04-09 19:23:04 invoked: alpine is pretty unuseable for a desktop user isnt it 2024-04-09 19:23:07 Combining different libc libraries is asking for trouble 2024-04-09 19:23:24 simon_: i use alpine as my primary desktop 2024-04-09 19:23:38 simon_: flatpak is an option 2024-04-09 19:24:05 simon_: i use a musl system as my desktop and it's alright 2024-04-09 19:24:08 ye but flatpak doesnt pass ledger 2024-04-09 19:24:14 for a couple of glibc pograms i just use a chroot 2024-04-09 19:24:24 and also flatpak helps 2024-04-09 19:24:43 what do you mean by chroot 2024-04-09 19:24:43 ? 2024-04-09 19:24:53 everything the majority people want to do with a desktop can be done with alpine. it comes down to how alpine does those things differently. 2024-04-09 19:25:22 it goes back to what reasons are you using alpine for 2024-04-09 19:25:23 simon_: just look it up 2024-04-09 19:25:35 to be specific i use bwrap though, not chroot 2024-04-09 19:28:25 triallax: and inside bubblewrap you put an operating system? 2024-04-09 19:33:34 very roughly speaking, yes 2024-04-09 19:33:52 bruh bwrap seems so hard 2024-04-09 19:34:18 it did take some setup but now it works without issues 2024-04-09 19:34:36 is there a guide on internet maybe which is not arch wiki 2024-04-09 19:34:42 what's wrong with arch wiki 2024-04-09 19:34:51 too much text 2024-04-09 19:35:02 you don't need to read all of it 2024-04-09 19:35:10 surely there's a youtube tutorial 2024-04-09 19:35:13 /s 2024-04-09 19:35:19 for me it was mostly 2024-04-09 19:35:22 fuck around and find out 2024-04-09 19:35:34 i can share my script if that helps 2024-04-09 19:35:44 yes please 2024-04-09 19:36:03 https://0x0.st/s/hxIQWSu8uGTzLGsEsP1Wuw/Xina.sh 2024-04-09 19:59:43 trillax the script doesnt work 2024-04-09 19:59:45 :( 2024-04-09 19:59:58 of course it wouldn't work as-is for you 2024-04-09 20:00:06 you probably need to tweak some things 2024-04-09 20:00:35 what's the problem 2024-04-09 20:01:18 ~ $ bash ./lol.sh usage: bwrap [OPTIONS...] [--] COMMAND [ARGS...] --help Print this help 2024-04-09 20:01:45 you need to pass it a command to run 2024-04-09 20:01:52 e.g. bash ./lol.sh echo hi 2024-04-09 20:02:16 as a suggestion, you shouldn't run scripts off the internet without reading them first 2024-04-09 20:02:49 i should've mentioned this, but my intention wasn't for you to run the script as-is 2024-04-09 20:02:58 but as a basis for your own and to save you the work 2024-04-09 20:03:03 bwrap: Can't find source path /var/lib/dbus/machine-id: No such file or directory 2024-04-09 20:03:26 /etc/machine-id 2024-04-09 20:04:47 bwrap: execvp echo: No such file or directory 2024-04-09 20:05:08 i really suggest you read the arch wiki 2024-04-09 20:05:09 and now I could put echo into this thing? 2024-04-09 20:05:23 there's no point to doing it blind like this 2024-04-09 20:06:03 but isnt there maybe a already setup script? 2024-04-09 20:06:13 bwrap is rather low-level, and if you want to use it directly you should know exactly what the command is doing 2024-04-09 20:07:13 i'm just going to say it: maybe alpine isn't for you, if you want magic helpers for everything 2024-04-09 20:07:30 if you're going to download random scripts from the internet and run them without reading 2024-04-09 20:07:32 (etc) 2024-04-09 20:09:50 is there no such thing as wsl just ll? 2024-04-09 20:10:15 and I mean no virtual machine 2024-04-09 20:10:22 docker 2024-04-09 20:10:28 can docker do gui? 2024-04-09 20:10:32 It can 2024-04-09 20:10:33 i've heard of distrobox 2024-04-09 20:10:36 perfect 2024-04-09 20:10:36 but i never used it 2024-04-09 20:11:08 simon_: it does mean you need to forward some things and have appropriate drivers installed for hw acceleration 2024-04-09 20:14:17 I will be right back 2024-04-09 20:34:20 Hi, is there any way to get Python 3.12 on alpine? 2024-04-09 20:36:42 skinkie: Not yet, we're working hard dealing with all the fallout of python3.12, pytest7, cython 2024-04-09 20:36:52 Lots of python packages need to be fixed 2024-04-09 20:41:33 what's the problem with py3.12? 2024-04-09 20:43:02 oGEosFMtUw8: removing things that have been deprecated for ages but some packages still using 2024-04-09 20:47:06 ohshit 2024-04-09 20:47:11 thx. ikke. 2024-04-09 21:12:13 I cant destroy alpine can I ? 2024-04-09 21:12:26 destroy alpine? 2024-04-09 21:12:49 like typing shit code in console and getting completly wrecked and bugged system 2024-04-09 21:13:01 it's certainly possible 2024-04-09 21:13:10 which code? 2024-04-09 21:13:24 why 2024-04-09 21:13:26 ? 2024-04-09 21:13:45 I executed many weird stuff from internet 2024-04-09 21:13:46 :D 2024-04-09 21:13:54 want a bomb to explode in my computer 2024-04-09 21:14:01 what stuff 2024-04-09 21:14:13 basically blackbox.ai stuff 2024-04-09 21:14:19 and he wrote much gibberish 2024-04-09 21:14:26 oO 2024-04-09 21:14:29 is that something like chatgpt 2024-04-09 21:14:34 yes 2024-04-09 21:15:03 any funny system destroyer command except rm -rf / 2024-04-09 21:17:52 doas chmod -R 777 / 2024-04-09 21:20:59 are my messages still arriving? 2024-04-09 21:22:49 ikke: the problem I am facing is homeassistant not being able to upgrade unless 3.12 is present, and sure I can compile stuff on my own, I was just hoping something was in the pipeline. 2024-04-09 22:18:05 them dropping 3.11 support feels like a move to force people to move to HAOS or docker deployments... 2024-04-10 09:29:32 `apk info --rdepends tesseract-ocr` doesn't report any reverse dependencies, this this must be a manual package (listed in world) yet it isn't, so why doesn't `apk fix` remove it? 2024-04-10 09:29:32 i struggle to perform an upgrade, as there is a dependency conflict. it involves icu / icu-libs . here's the conflict erro https://dpaste.com/5F7M2BN6U.txt . how to fix my way out of this? 2024-04-10 10:37:48 Piraty: what version of Alpine you running? 2024-04-10 13:17:51 cyclisme24[m]: this is alpine3.19 with postmarketos23.12 repos 2024-04-10 13:18:33 the issue appear first when migrating from alpine3.18+pmos to alpine3.19+pmos 2024-04-10 13:19:41 so i guess at least one of manual packages in apk/world is not provided by the binary package repos and depends on icu , hence i want to know the full revdep tree of them to see which one it is 2024-04-10 17:03:16 I am not sure if it's already known but the writefreely package doesn't have a maintainer, and it has been outdated since November 2022. 2024-04-10 17:16:44 nep_nep: are you interested in maintaining it? 2024-04-10 17:18:34 I am fairly certain I don't have the experience required to do it confidently. 2024-04-10 17:41:46 nep_nep: everyone has to start somewhere :) 2024-04-10 17:46:37 I appreciate it but I don't think starting on a widely used distribution is a very good idea. 2024-04-10 17:46:50 I don't trust myself to not break anything 2024-04-10 18:34:06 nep_nep: that's why other people are around to review and help 2024-04-10 18:34:26 you get experience by doing it 2024-04-10 18:36:56 Starting small is the standard for gaining experience when the lack of it would have a high risk, for a good reason. 2024-04-10 18:37:17 writefreely isn't a critical package probably 2024-04-10 18:37:20 so i'd say go for it 2024-04-10 18:38:52 Well, if that many people are supportive I suppose I'm probably overestimating how risky it is, how does the process of becoming a package maintainer work? 2024-04-10 18:39:07 you basically slap your name on it 2024-04-10 18:39:12 actually i don't know how it works in alpien 2024-04-10 18:39:51 Surely there's some kind of access control. 2024-04-10 18:40:23 i don't think package maintainers usually can merge changes directly (they need to send MRs), but i can't say how it works for alpine 2024-04-10 18:40:27 i'll let the experts speak 2024-04-10 18:43:27 https://wiki.alpinelinux.org/wiki/Developer_Documentation is a good place to start 2024-04-10 18:45:02 or more specifically: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Bumping_a_package_version 2024-04-10 20:57:54 anybody who can assists me with a question to unbound? 2024-04-10 20:58:23 > slow channel, ask your question and wait for a response. 2024-04-10 20:59:05 (you don't need to ask to ask) 2024-04-10 21:14:30 how many querys it must done to server fail unbound with max querys alert and can i ignore somehow this client 2024-04-10 21:21:34 this querys are imcp packets, but crashes unbound btw. 2024-04-10 21:41:55 is it possible to remap mouse buttons without installing a binary? 2024-04-10 22:05:04 ikke, ptrc, hey what about !63556? Do you want to wait for an upstream fix in forgejo/node (probably a node issue, hard to debug for me locally, the stuff builds for me in the riscv64 docker)? 2024-04-10 23:53:36 ` apk list --upgradable` shows a package update, yet `apk upgrade` doesn't upgrade the mentioned package. why may that be, and how to find out? 2024-04-11 01:05:01 does ``apk upgrade -a'' fix it? 2024-04-11 01:09:27 nep_nep: if you're still around, hope !63907 will help you out a bit in your quest 2024-04-11 01:10:43 haven't actually tested the artifact to see if it actually runs, but may get you closer, maybe? 2024-04-11 01:11:18 Sounds good 2024-04-11 01:11:23 if you want to take maintainership of it, for example ^^" 2024-04-11 01:12:02 cheers 2024-04-11 06:58:01 famfo: i'm trying to reproduce the same issue for now, to make sure it doesn't hit the builders; if i won't be able to, then i think it's ready to merge 2024-04-11 07:20:48 Okay good, ty 2024-04-11 12:01:33 zcrayfish: no 2024-04-11 12:02:55 Piraty: try: apk add -s 'package=~' 2024-04-11 12:06:27 thanks ikke, that does the job i guess 2024-04-11 12:06:38 reports a conflict with an installed pkg 2024-04-11 12:06:45 yup 2024-04-11 12:06:49 that's the ideas 2024-04-11 12:07:53 how to grab a revdep tree though? in previous case (icu-libs upgrade gives conflict) the broken packages are not in world but transitive (i assume, yet didn't get why tesseract behaved the way it did) 2024-04-11 12:08:27 I haven't found a good way yet except step by step narrowing it down 2024-04-11 12:08:39 one shortcut may be: apk list --orphan 2024-04-11 12:08:47 but I recall you use pmos 2024-04-11 12:09:00 so if it's a conflict between the 2, that would not help 2024-04-11 12:09:22 apk should be indifferent regarding package file repo ...? 2024-04-11 12:09:37 But those repos should be consistent 2024-04-11 12:09:46 i'll dig more into it 2024-04-11 12:10:16 If one repo holds back a package that the other repo needs the latest version of, it's a conflict 2024-04-11 12:13:23 sure 2024-04-11 12:22:09 may it be that a revbump was missed on nheko ? the conflict is for gst-plugins-good-qt which breaks/satisfies nheko 2024-04-11 12:23:58 what does apk info -P nheko return? 2024-04-11 12:24:38 nheko-0.11.3-r2 provides:\ncmd:nheko=0.11.3-r2 2024-04-11 12:24:46 (inlined for your convenience) 2024-04-11 12:24:55 sorry, I meant -R 2024-04-11 12:25:00 which cannot be easily inlined 2024-04-11 12:26:10 the conflicing lib is : nheko-0.11.3-r2[so:libgstqmlgl.so 2024-04-11 12:26:15 ] 2024-04-11 12:26:56 for me, nheko does not depend on that 2024-04-11 12:27:30 https://paste.rs/XrtjI 2024-04-11 12:27:41 that is on 3.18 2024-04-11 12:27:52 oh, 3.18 2024-04-11 12:28:15 comparing commit log of 3.18 to 3.19 now ... 2024-04-11 12:28:19 but the soname is not versioned 2024-04-11 12:28:25 yeah duh 2024-04-11 12:28:40 .. 2024-04-11 12:28:50 upstream doing good work 2024-04-11 12:29:06 but then, apk wouldn't care in this case and i'd get runtime errors 2024-04-11 12:29:13 yup 2024-04-11 12:29:30 apk info -P gst-plugins-good-qt for good measure 2024-04-11 12:29:36 (this time I meant -P) 2024-04-11 12:30:01 thanks for holding my hand with this 2024-04-11 12:30:21 https://termbin.com/anye 2024-04-11 12:31:20 ok, so the new version only provides qml6 2024-04-11 12:31:23 i see nheko/APKBUILD specifies all the lib deps manually, wth 2024-04-11 12:31:45 c9549a356d6db87c523345750fe9511199bf5260 may have been useful on 3.18 as well 2024-04-11 12:32:31 that's it 2024-04-11 12:32:46 are patches to 3.18 still accepted? 2024-04-11 12:33:33 lol nheko doesn't even link this lib 2024-04-11 12:33:52 I think something like this to fix dependencies should not be an issue 2024-04-11 12:34:02 `git mv community wild-wild-west` 2024-04-11 12:34:12 lol 2024-04-11 12:34:19 what's the difference? 2024-04-11 12:34:20 ;_ 2024-04-11 13:00:03 I like kvm because it's simple. But I like xen more, because it has powerfull features and the VMs running smoother (e.g. smooth desktop experience). Does a third hypervisor exist you recommend which I don't know so far? 2024-04-11 13:00:19 I've heard of proxmox 2024-04-11 13:01:20 Ain't proxmox a VM-manager based on KVM? 2024-04-11 13:01:29 possibly 2024-04-11 13:17:34 ProxMox is both KVM and LXC/LXD 2024-04-11 15:11:20 I use virt-manager for KVM, Xen or LCX_LXD to setup local VMs on my laptop. On productive machines I use Xen and LXC_LXD manually. 2024-04-11 15:28:34 same here. 2024-04-11 16:49:59 yeah proxmox uses qemu server which uses kvm for virtualization 2024-04-11 16:50:13 an opensource competitor would be xcp-ng who uses xen instead 2024-04-11 16:50:31 seems to be the two which former vmware customers switch to 2024-04-12 01:25:15 hi 2024-04-12 15:43:56 how come openssh always pulls in openssh-server? 2024-04-12 16:20:27 elly: appears to be primarily a meta package 2024-04-12 16:22:01 if you want just the client, install just openssh-client-default or openssh-client-krb5 if you need kerberos support 2024-04-12 16:40:20 aha 2024-04-12 16:40:40 cheers, thanks :) 2024-04-12 19:35:42 after installing minimal xfce based on the wiki, I have a pecular issue with Firefox's interaction with thunar: I can download files as long as "ask where to save" is not set but the files do not appear in thunar; although they do appear in Firefoxes download dialogue 2024-04-12 19:38:46 not sure if this has something to do with my home directory only having the Desktop folder after initial install 2024-04-12 19:49:09 click the folder icon in the firefox downloads popup, it should open the location in thunar 2024-04-12 19:56:05 yep, tried that, nothing happens 2024-04-12 19:57:19 when I click on the file itself, it asks me if I want to open the link in the system handler 2024-04-12 20:00:57 I am using the flatpak version of Firefox 124.0.2 2024-04-12 20:24:23 chmod +x on that flatpak as with appimage ? 2024-04-12 22:02:13 @Apachez you may be on to something, I can try it 2024-04-12 22:26:05 So how unsafe would be for me to pull incus@edge while on 3.19 🤔 2024-04-12 22:40:56 it may want to pull in some other dependencies and then... 2024-04-12 22:41:18 unsupported 2024-04-12 22:54:59 So the other option is to go to edge until 6.0 gets into some non edge release of alpine 2024-04-12 22:57:17 it will be in the upcoming 3.20 release, maybe as early as sometime in may 2024-04-12 23:24:52 on edge, installing sdl2-dev appears to be impossible: https://bpa.st/APUQ - sdl2-dev depends indirectly on xcb-proto which seems to depend on a different version of python3 than a bunch of other programs I have installed, and I'm unsure why 2024-04-12 23:25:09 (`apk upgrade` makes no changes) 2024-04-12 23:27:42 yes, we're in the middle of rebuilding every python related aport against python 3.12 2024-04-12 23:27:42 possibly my system is screwed up somehow but I'm unsure quite how 2024-04-12 23:27:45 oh okay 2024-04-12 23:27:47 that'd do it 2024-04-13 04:41:13 Just a heads up. fsck from util-linux 2.40 is segfaulting on me. Worked around by turning off fsck in fstab. In a bit I'll try to get a decent backtrace. 2024-04-13 07:24:59 Hi, guys. I have a question 2024-04-13 07:25:23 When I suspend my computer, slightly moving the mouse will wake it up 2024-04-13 07:25:32 Is there a way to stop my USB mouse from waking up the computer? 2024-04-13 07:27:01 https://wiki.archlinux.org/title/Power_management/Wakeup_triggers this also applies on alpine 2024-04-13 07:28:11 i assume a mouse event would be part of XHCI events but i'm not sure 2024-04-13 07:57:35 Thanks caskd! 2024-04-13 08:47:55 is it jsut me or does the packages page not load? 2024-04-13 08:48:51 maybe just a hiccup 2024-04-13 08:48:57 loads fine here 2024-04-13 08:50:00 idk, every time ive tried it recently, takes several minutes to load 2024-04-13 08:50:27 but it's fine, went to the GH mirror instead 2024-04-13 15:10:59 can i use ap (from lua-aports) to find recursive dependencies across repositories? e.g.: to find packages in testing that depend on a given package in testing? 2024-04-13 15:12:11 e.g.: using 'ap recursive-revdeps python3' on testing yields no results 2024-04-13 20:18:00 I just had a weird breakage - on bootup, fsck crashed with a segfault. 2024-04-13 20:18:00 I already fixed it - running another apk upgrade did the trick - but it was tricky to recover enough to that point. 2024-04-13 20:18:14 It seems one can't skip fsck in interactive boot, as virtually everything else depends on fsck. 2024-04-13 20:18:54 So what I ended up doing was touch /run/openrc/started/fsck, which then let me rc-service iwd start and from there on I could run an upgrade that fixed it. 2024-04-13 20:19:04 Would be nice if interactive boot let me "fake a service running" like I did there 2024-04-13 20:20:44 Basically, not complaining about the fact that fsck had a segfault - it's already fixed, and breakages are kinda expected on edge. Just wondering if we should allow for better recovery methods in openrc's interactive mode. 2024-04-13 20:22:09 hm... rc-service -D iwd start might also have done the trick to get wifi running, that might be a more documented way to do this 2024-04-13 20:52:59 divVerent: 'touch /fastboot' to skip fsck while having it report success 2024-04-13 21:06:55 Good afternoon. Greetings from Brazil..... (full message at ) 2024-04-13 21:42:35 "divVerent: 'touch /fastboot..." <- Good one too, not sure if it would have worked in my case, but good to know 2024-04-13 21:43:02 you would have had to manually remount root readwrite first to do that, yes 2024-04-13 21:43:02 mount -a had segfaulted too, but mounting individual fs was fine 2024-04-13 21:43:30 probably an issue in a util-linux lib that's only used by the -a flags in mount and fsck 2024-04-13 21:43:31 no, I mean, not sure if fsck segfaulted before looking at /fastboot 2024-04-13 21:43:42 or is that file read by the init script only? 2024-04-13 21:43:45 the /fastboot check is in the initscript 2024-04-13 21:44:00 ah, I see, then it too would have worked 2024-04-13 21:44:23 just felt a bit weird that the interactive boot feature of openrc was so close to helping, but did not 2024-04-13 21:44:35 but of course the login prompt did help 2024-04-13 21:45:47 basically, "skip this service but continue assuming it is running" would be neat to have for cases like this 2024-04-13 21:46:10 yeah, specifically for services that are important in the depgraph normally 2024-04-13 21:46:10 as openrc is rather strong at enforcing dependencies 2024-04-13 21:47:09 or maybe alternatively an option to launch a service ignoring deps (like rc-service -D) 2024-04-13 21:47:17 so I could kinda have shoehorned iwd in 2024-04-13 21:47:27 and from there I had enough to run apk 2024-04-13 21:49:41 as I knew I had run apk upgrade yesterday and the issue of crashing fsck breaking boot is so serious, it was kinda obvious that a) this kind of stuff happens in edge, and b) if it happens, it likely also is fixed very quickly 2024-04-13 21:51:03 still amazing I could fix this at all without booting from usb media 2024-04-13 21:51:32 when I had such an issue regarding early mounting on Ubuntu a few years ago, I only ended up in initrd shell with no tools to fix anything 2024-04-13 21:52:20 on alpine I got a hostnameless root login prompt and could play around with openrc stuff from there to get it to boot sufficiently 2024-04-13 21:54:10 i don't know about alpine but here i can just pass pass init=/bin/sh kernel param in grub, remount root as read-write, and boom 2024-04-13 23:35:57 yeah, I think what Ubuntu did different is doing more mounting related stuff from the initrd - so that won't help and the initrd shell had no tools to fix it. That was during upstart times though, pre systemd. 2024-04-13 23:37:59 whereas Alpine does the more traditional thing of mounting rootfs ro early and if something breaks I have at least a rootfs to work with 2024-04-14 02:04:39 Hello. I'm having some trouble getting chromium to play back youtube videos (it just spins). Audio and video otherwise seem to be working. Any ideas? 2024-04-14 02:06:26 I should mention that I am running a minimal install with seatd and sway 2024-04-14 02:06:43 firefox works fine 2024-04-14 04:25:04 anyone seen local mount segfault recently? 2024-04-14 04:25:40 i updated edge yesterday and booted up first time today and saw it has segfaulted 2024-04-14 04:28:08 many reports of segfaults with fsck and mount -a; it is fixed if you update after mounting manually, purportedly 2024-04-14 04:29:49 fsck is getting segfaulted 2024-04-14 04:30:21 ah sorry following what you are saying now 2024-04-14 04:45:17 hmm cant even get networking up 2024-04-14 04:46:05 touch /fastboot after manually mounting / read-write, and then /etc/init.d/fsck will "start" successfully 2024-04-14 04:46:39 or you can trick openrc into thinking fsck is already started by touching /run/openrc/started/fsck 2024-04-14 04:47:10 the latter does not persist across reboots 2024-04-14 04:50:56 ok rebooting, fingers crossed 2024-04-14 04:52:18 wooo! working thanks dwfreed 2024-04-14 13:08:03 should i assume that all other stuff that depends on python3~3.11 will eventually be rebuilt, like libreoffice-common? 2024-04-14 13:09:49 er, it already is. i don't understand what's going on here then: http://0x0.st/X-tO.txt 2024-04-14 14:12:14 Not sure if others are affected, too, but alsa/sof seems to be broken in recent -lts kernels, latest tested on 6.6.25 2024-04-14 14:12:34 sof errors out with sof-audio-pci-intel-tgl 0000:00:1f.3: error: firmware intel/sof/sof-tgl.ri contains unsupported or invalid extended manifest: -22 2024-04-14 14:23:24 nvm, found the bug, seems sof-firmware is now separate or had been uninstalled for some reason 2024-04-14 15:26:29 Hi, i'm trying to install ffmpeg but i'm getting this :apk add ffmpeg 2024-04-14 15:26:29 ERROR: unable to select packages: 2024-04-14 15:26:29 so:libmbedcrypto.so.7 (no such package): 2024-04-14 15:26:34 required by: librist-0.2.10-r0[so:libmbedcrypto.so.7] 2024-04-14 15:28:38 i must have done something wrong since this happens both on my laptop and on my raspberry pi homelab. Btw they're both on edge 2024-04-14 15:46:48 Hello. I am not sure if anyone else is having a problem with installing boost-build or boost-dev. It seems there is some kind of python version conflict. 2024-04-14 15:50:28 There's a python conflict, i am able to install both boost-build and boost-dev 2024-04-14 15:50:34 q 2024-04-14 15:52:38 Anyone to help i got hacked and got alpine installed by a rootkit so im stuck in lkvm please!! 2024-04-14 15:53:05 Thank you repack. Do you know if there is any way to reconcile my system? 2024-04-14 15:53:16 All my device af 2024-04-14 15:53:30 All my device are co.pr 2024-04-14 15:53:36 Compromised 2024-04-14 15:56:04 I cant do anything since im stuck in vm and i know very little about alpine but im so freaking desperate honestly 2024-04-14 15:56:24 do you know openrc ? 2024-04-14 15:57:14 Me? 2024-04-14 15:57:19 yes 2024-04-14 15:57:33 Not really 2024-04-14 15:57:35 alpine do not use systemd but openrc 2024-04-14 15:58:27 brief intro to maybe help you: rc-service *service* start/stop... rc-update add/del *service* default (to add a service at boot) 2024-04-14 15:58:57 Its openrc then. Sorry having bother you guys. Do you know if if still possible t 2024-04-14 15:59:07 if you want root priveleges use doas, not sudo, it works the same way 2024-04-14 15:59:13 To get back accesover my system 2024-04-14 15:59:26 maybe shutting down the alpine vm will help 2024-04-14 15:59:27 I didny 2024-04-14 15:59:44 Idndt know that. Thanks man! 2024-04-14 16:00:40 So then i would clean user and password but i would still have to clean my ram since it used DMA 2024-04-14 16:00:57 It also flashed my bios 2024-04-14 16:01:02 damn 2024-04-14 16:01:13 Pretty much fuckedmy wholesetup 2024-04-14 16:01:24 Pcs cellphone and tv 2024-04-14 16:01:39 how 2024-04-14 16:01:41 All that shit t 2024-04-14 16:02:00 I gave someone remote acces over helpcez 2024-04-14 16:02:15 Helpdesk and he flashed my driver 2024-04-14 16:02:27 Wich started the whole process 2024-04-14 16:02:48 Everything running linux was fucked 2024-04-14 16:02:55 try changing firewall settings in your router 2024-04-14 16:03:46 Its already been reset and setup but it seems to be operating by http wich i cant block 2024-04-14 16:04:51 i dont know if you have the setup/router for but try making vlans for each devices 2024-04-14 16:04:53 From my research its a twin of the blacklotus viruz 2024-04-14 16:05:23 But using vm to stop owner from accessing bootloader 2024-04-14 16:05:48 what happened to your phone ? 2024-04-14 16:06:22 All because of a fivem gta server i was building andgot too much attention with it fuck sake im sick anx tired man😔 2024-04-14 16:06:56 Good luck man 2024-04-14 16:06:58 From what i know it spread over wifi and bluetooth 2024-04-14 16:07:32 Phone are working but in a workgroup vm 2024-04-14 16:07:44 Cant change any root files 2024-04-14 16:08:18 i think youre pretty fucked tbh 2024-04-14 16:08:46 It cant be man there .ust 2024-04-14 16:08:55 Must be so.ething i can do 2024-04-14 16:09:31 This is so horrifying idk how to help anymore 2024-04-14 16:09:40 If i can get ssh root acces then i could reinstall alpine and trash those setting 2024-04-14 16:10:22 Its ok brother ill find a way i c 2024-04-14 16:10:31 good luck 2024-04-14 16:10:35 Simply cant affordtolose 2024-04-14 16:11:03 All that money. Sorry i have hard time typing on my phone 2024-04-14 16:11:14 its ok 2024-04-14 16:12:40 if you any questions about alpine i can help 2024-04-14 16:16:07 Not specific to alpine as you told me its openrc 2024-04-14 16:18:32 It seems my system has both python3-3.11.8-r0 and python3-3.12.3-r0 installed at the same time. APK reports they "break" each other and list every python program I have installed as "satisfies". So, I'm in some kind of dependancy loop and can't install a bunch of programs. This now makes sense with another program I was trying to install last week that is in the repos. 2024-04-14 16:18:51 How can I start to fix this problem? 2024-04-14 17:07:25 can I force uninstall then reinstall python3 package? How do I remove one and not the other, or reconcile that there are two versions installed? 2024-04-14 17:08:31 SagestMage: how did you verify both versions are installed at the same time 2024-04-14 17:08:34 ? 2024-04-14 17:21:06 ikke: it says from "apk info python3" 2024-04-14 17:21:39 SagestMage: that tells what versions are available, not installed 2024-04-14 17:21:44 oh 2024-04-14 17:21:46 woops 2024-04-14 17:21:48 sorry mate 2024-04-14 17:21:59 ok, I'm getting something though with the conflict 2024-04-14 17:22:34 I get two different "breaks" statements. One for each version 2024-04-14 17:22:36 apk does not allow you to installed the same package twice 2024-04-14 17:22:46 s/installed/install 2024-04-14 17:22:47 how do I know what is installed now? 2024-04-14 17:22:53 apk version python3 2024-04-14 17:23:27 Installed: Available: python3-3.11.8-r0 < 3.12.3-r0 2024-04-14 17:23:32 you hav 3.11 2024-04-14 17:23:37 while the repo now contains 3.12 2024-04-14 17:23:40 yeah, it seems stumped on the upgrade 2024-04-14 17:23:47 apk list --orphan 2024-04-14 17:23:53 I had another thing last week connected to python that was out of wake 2024-04-14 17:23:55 something is holding back python3 2024-04-14 17:23:59 (3.12) 2024-04-14 17:24:10 fyi, python3.12 has just been pushed 2024-04-14 17:24:15 like a couple of days ago 2024-04-14 17:24:36 I see, maybe some packages are stll bound to the 3.11? 2024-04-14 17:24:47 nothing pops out from list --orphan 2024-04-14 17:24:49 if you have packages from testing installed, yes 2024-04-14 17:24:55 errr, orphaned 2024-04-14 17:24:56 testing is still being built 2024-04-14 17:25:07 I indeed have from testing installed 2024-04-14 17:25:15 no problem if I have to wait a little 2024-04-14 17:25:17 SagestMage: what does: apk add -s python3=~3.12 return 2024-04-14 17:25:29 mostly the 'breaks' part 2024-04-14 17:25:54 ERROR: unable to select packages: python3-3.12.3-r0: breaks: ueberzug-18.2.2-r0[python3~3.11] py3-bcrypt-4.1.2-r0[python3~3.11] py3-dateutil-2.9.0-r0[python3~3.11] py3-cairo-1.26.0-r0[python3~3.11] py3-pillow-10.3.0-r1[python3~3.11] hplip-3.23.12-r1[python3~3.11] syncthing-gtk-0.9.4.5-r0[python3~3.11] 2024-04-14 17:26:14 but also lists a bunch of stuff in "satisfies" that I didn't paste 2024-04-14 17:26:33 euberzug is in testing 2024-04-14 17:26:46 It makes sense if the system is being bui8lt in the repos that it'll just take a little while to get fixed. I could wait a little, it's not essential right now 2024-04-14 17:27:06 im curious about py3-bcrypt, which is in community 2024-04-14 17:27:15 yeah, I installed quite a few from testing 2024-04-14 17:27:26 SagestMage: did you run apk update recently? 2024-04-14 17:29:13 I did this morning, but this problem I started noticing was the day before yesterday 2024-04-14 17:29:35 ok, because for me, py3-bcrypt already depends on python3.12 2024-04-14 17:30:05 I'm guessing there's no way to "force" the update? 2024-04-14 17:30:30 no 2024-04-14 17:30:43 things would not work as expected anyway if you tried 2024-04-14 17:32:48 yeah, so I've got the bcrypt package installed but it also says it breaks 2024-04-14 17:33:14 perhaps because that package is held back as well 2024-04-14 17:33:18 I never really use ueburzug or syncthing ... I will try removing them 2024-04-14 17:33:35 I only installed to test for another application 2024-04-14 17:33:45 was going to get around to doing more but haven't yet. 2024-04-14 17:35:20 ok, it seems to work now, I removed euberzug, syncthing, and hplip 2024-04-14 17:35:30 now I have 147 updates 2024-04-14 17:35:45 thanks a lot ikke, you're a blessing 2024-04-14 17:35:57 SagestMage: glad it worked out 2024-04-14 17:36:25 holy, it was holding back a lot of good and important stuff 2024-04-14 18:03:49 ikke: I'm well on my way with boost and compiling away. Thank you again for the help, I wouldn't have figured it out without your intervention. 2024-04-14 22:36:02 I guess I should start getting ready to rebuild all the python stuff in my edge repo (really just ceph, qemu, and samba) 2024-04-15 10:54:02 how does 'abuild rootbld' determine which mirror to use? 2024-04-15 10:54:24 i seem to have configured a mirror when travelling abroad and can't figure out how to revert that 2024-04-15 10:57:31 rootbld uses .rootbld-repositories in the repo 2024-04-15 10:57:44 so it would be for example in main/.rootbld-repositories 2024-04-15 11:00:27 Ah, yes, and that uses the `mirror` variable which is set in my end 🤦 2024-04-15 11:00:28 thanks! 2024-04-15 12:04:29 is it expected that util-linux's dependency /lib/libeconf.so.0 is a symlink to /usr/lib? 2024-04-15 12:06:32 sorry, rather that it's just in /usr/lib. no symlink involved 2024-04-15 16:04:11 does the kernel have an api via wihch i can monitor the current wifi network for a given interface? 2024-04-15 16:04:35 yes, netlink 2024-04-15 16:04:44 usable via iw 2024-04-16 00:52:59 following up on my Firefox download to thunar issue - I resolved it by installing the xdg-desktop-portal-gtk package and reseting the downloads folder in Firefox settings 2024-04-16 07:05:24 hmm seems khard is broken https://clbin.com/fudoC 2024-04-16 07:05:41 does this happen because python upgraded? 2024-04-16 07:06:08 https://pkgs.alpinelinux.org/package/edge/community/x86_64/khard 2024-04-16 07:16:13 transitional issue 2024-04-16 07:16:27 you just need to wait until all python 3.12 issues are sorted out 2024-04-16 07:16:36 but also 2024-04-16 07:16:44 whatever you're doing is blatantly wrong 2024-04-16 07:17:34 ( https://wiki.alpinelinux.org/wiki/Repositories#Upgrading_to_edge ) 2024-04-16 07:28:58 it's actually being built in a container, and only that package is pinned 2024-04-16 07:29:29 unfortunately khard only exists in edge 2024-04-16 07:33:02 fortunately i have image still on the other machine so i'll save and load it on this one 2024-04-16 09:31:28 so what's the plan for testing? i see main and community have everything rebuilt for python3~3.12, but testing has plenty outstanding 2024-04-16 09:31:50 (also i think it's impossible to check this using https://pkgs.alpinelinux.org ?) 2024-04-16 11:20:40 is there known problem with aarch64 builds? I got a huge python3 upgrade conflict message 2024-04-16 11:20:51 looks like 3.11 -> 3.12 2024-04-16 11:23:04 staceee: testing is still being built 2024-04-16 11:23:10 so if you have packages from testing, they will cause conflicts 2024-04-16 11:23:44 okok 2024-04-16 12:02:08 the x86_64 builders are idle 2024-04-16 12:02:17 can the new index be shipped for x86_64 edge at least by now? 2024-04-16 12:14:25 should be, what are you missing? 2024-04-16 12:15:05 there are a few aports that are yet to be rebuilt against python 3.12, but I think they're all in testing/ 2024-04-16 12:19:45 looks this way, is there a better way to check for this than $ for f in /var/cache/apk/APKINDEX.*;do tar xOf $f APKINDEX|grep -q python3~3.11 && echo $f;done # particularly how do i map repo name <-> hashsum? 2024-04-16 12:45:05 libreoffice seems to be broken #16005 2024-04-16 12:48:15 anjan: river -> 0.3.0 (also: community?) 2024-04-16 13:31:23 is the Python 3.12 upgrade/build still on going or is it complete? 2024-04-16 13:32:44 testing is still being built on some arches 2024-04-16 13:35:01 reason why I ask is /usr/bin/ics2rem from remind-caldav has issues 2024-04-16 13:35:22 on x86 2024-04-16 13:35:54 worked fine on Python 3.11 2024-04-16 13:37:19 py3-remind is in testing, so still needs to be rebuilt / uploaded 2024-04-16 13:37:59 ok. that's what I figured but wanted to verify 2024-04-16 13:39:40 thanks 2024-04-16 13:56:07 Hi 2024-04-16 13:57:11 I'm considering arm64 hosting on hetzner using alpine. Anyone here have experience with it? 2024-04-16 13:57:55 One use case is vpn via wireguard and also remote backups etc. 2024-04-16 13:58:08 do you have any question in particular 2024-04-16 14:04:55 Any ideas how I can debug this SEGFAULT in the xdg-desktop-porta? https://gitlab.alpinelinux.org/alpine/aports/-/issues/15996 2024-04-16 14:05:48 try running it directly in gdb/lldb 2024-04-16 14:06:30 though actually not sure that would help much 2024-04-16 14:06:42 I did, the stacktrace is in the issue. 2024-04-16 14:06:52 i thought you got that from a corefile 2024-04-16 14:06:53 I'm not sure how I can obtain some of the missing symbols; there's no xdg-desktop-portal-dbg 2024-04-16 14:07:01 No, I got that from gdb / bt 2024-04-16 14:07:13 you can do that with a core dump too 2024-04-16 14:07:20 gdb -c core.* 2024-04-16 14:07:20 bt 2024-04-16 14:08:30 but you can try installing the -dbg package for the makedepends of xdg-desktop-portal 2024-04-16 14:08:32 packages* 2024-04-16 14:08:37 triallax: you mean me? Just in general how the arm64 vps are with Alpine vs amd64. You get 4arm64 cores 8g ram for roughly the same as 2 amd cores with 4gb ram 2024-04-16 14:09:09 i don't think that question is particularly alpine-specific 2024-04-16 14:09:55 Is anyone else using xdg-desktop-portal here? Does it run fine with the latest packages from edge? 2024-04-16 14:10:12 I want to understand if it's a "just me" issue or not. 2024-04-16 14:16:20 Triallax maybe not, except I want to use Alpine with it, wireguard etc. 2024-04-16 14:16:39 I'll take it to off topic channel 2024-04-16 14:16:54 no it's fine to ask here, i didn't imply it's not 2024-04-16 14:17:07 but i just mean that the answer isn't going to be much different depending on what distro you choose 2024-04-16 14:21:15 WhyNotHugo: works fine 2024-04-16 14:23:50 WhyNotHugo: sorry, i should include the context in which i use it: flatpak stuff, eg steam 2024-04-16 15:21:27 invoked: thanks 2024-04-16 15:21:27 it starts if i delete my configuration file 2024-04-16 15:26:55 someone else reported it upstream: https://github.com/flatpak/xdg-desktop-portal/issues/1330 2024-04-16 21:21:46 Forza: not a VPS, but my rock5b's have been working pretty smoothly on Alpine (as smooth as you can expect with edge anyway) 2024-04-16 21:58:46 iggy: thanks 2024-04-16 22:00:11 I do have an issue. Hetzner uses virtiogpu and that's not loaded in initramfs before it asks for luks key 2024-04-16 22:06:36 virtio-gpu or virtio-console? It seems weird a cloud provider would be using virtio-gpu 2024-04-16 22:15:37 Iggy virtio_console 2024-04-16 22:32:45 Hm not sure how to fix . The console is not active until th 2024-04-16 22:33:27 the os is booted. Grub works though, then a black screen. 2024-04-16 22:33:56 Hello everyone I'm a Grower and supplier of cannabis and psychedelics products such as;... (full message at ) 2024-04-16 22:34:14 if you have virtio in features in mkinitfs.conf (default I think), it should be included 2024-04-16 22:34:47 Forza: you might just have to tell inittab to add a serial console on hvc0 2024-04-16 22:35:39 Inittab isn't the issue, but I have no prompt to open luks 2024-04-16 22:36:38 or maybe on the kernel command line? 2024-04-16 22:38:02 I have both in my alpine VM build scripts (add `hvc0::respawn:/sbin/getty -L hvc0 115200 vt100` to inittab and `console=hvc0` to kernel command line) 2024-04-16 22:39:42 https://paste.tnonline.net/files/8lh186yweQmV_Screenshot_20240417_003838_RVNCViewer.png sure looks like virtio gpu 2024-04-16 22:40:00 That's booted from the alpine iso 2024-04-16 23:32:40 then you'll probably have to patch mkinitfs to include the virtio-gpu module with the virtio feature 2024-04-16 23:32:52 and maybe send that upstream 2024-04-16 23:35:15 edit /etc/mkinitfs/features.d/virtio.modules and add `kernel/drivers/gpu/drm/virtio/virtio-gpu*` (I think that should do it) 2024-04-16 23:35:37 then remake your initramfs 2024-04-17 01:23:51 Iggy in the end i needed "modules=simpledrm,sd-mod,usb-storage,ext4 console=tty1" to get a visible luks prompt in initrd 2024-04-17 01:29:53 got to sleep now. Iggy, minimal, thanks for the help 2024-04-17 07:41:37 Is there a better way to list redundant entries in /etc/apk/world than simulating deleting each in isolation, checking for dependents keeping them present? `cat /etc/apk/world | while read line; do apk del -s $line; done` 2024-04-17 08:09:29 Sofia, what do you mean by redundant entries? 2024-04-17 08:10:03 I assume "ones that would be in the dependency tree even without being explicitly listed" 2024-04-17 08:14:18 quinq: As grawity said. 2024-04-17 08:15:07 Speaking of the tree, any useful tools to visualize this in topological order? 2024-04-17 08:15:45 I think there's a graphviz tool 2024-04-17 08:16:01 Yeah, apk-dot 2024-04-17 08:16:36 Is that in edge or? 2024-04-17 08:17:05 apk dot 2024-04-17 08:17:12 Part of apk-tools 2024-04-17 08:18:40 Oh 2024-04-17 08:18:44 No dash 2024-04-17 08:18:45 ACTION nods 2024-04-17 08:19:23 Well… apk-dot(8) ;) 2024-04-17 08:19:33 Right 2024-04-17 08:19:37 ACTION doesn't even have the man pages heh 2024-04-17 08:20:12 I've been running "dot -Tpng apk.world.dot" for a couple minutes at full CPU already :/ 2024-04-17 08:20:26 What monster will come out of that 2024-04-17 08:20:28 Uh oh 2024-04-17 08:21:18 ACTION runs with time 2024-04-17 08:21:43 quinq: How many packages? 2024-04-17 08:21:47 I have 659. 2024-04-17 08:22:35 $ apk list -I | wc -l 2024-04-17 08:22:35 1529 2024-04-17 08:22:44 wow 2024-04-17 08:22:48 so many 2024-04-17 08:23:12 ACTION thought her 659 was many :p 2024-04-17 08:23:17 haha me too 2024-04-17 08:23:33 Well, it's also the separation of base/dev/doc 2024-04-17 08:23:44 Basically almost every package is triplicated 2024-04-17 08:23:50 I don't have a lot of dev, no doc. 2024-04-17 08:23:55 :/ 2024-04-17 08:24:26 New word: triplicated. 2024-04-17 08:24:35 :D 2024-04-17 08:24:55 oh, it seems that's actually a word 2024-04-17 08:25:01 Huh 2024-04-17 08:25:02 https://en.wiktionary.org/wiki/triplicate#English 2024-04-17 08:28:49 New word~ to me. 2024-04-17 08:30:27 Well, we have the dup function, guess we should have a trip function too. Not to be mistaken for tripping over exceptions. 2024-04-17 08:30:46 Huh. Poor dot... 2024-04-17 08:30:51 quinq: Any output yet? XD 2024-04-17 08:38:08 Ah whoops. I didn't do --installed, so the dot output was everything. 2024-04-17 08:39:48 dot: graph is too large for cairo-renderer bitmaps. Scaling by 0.620587 to fit 2024-04-17 08:39:51 real 0m 57.79s 2024-04-17 08:40:52 39.6M apk.png 2024-04-17 08:40:53 Hmm 2024-04-17 08:40:59 My image viewer doesn't like this image. 2024-04-17 08:42:47 Just get a white rectangle. 2024-04-17 08:43:46 Unless dot just failed? 2024-04-17 08:44:08 Sofia, I went back to work and forgot about it :D 2024-04-17 08:44:16 Of course it output the png to stdout, I have to do it again 2024-04-17 08:44:38 Same, cause I used your command :p 2024-04-17 08:44:43 hahaha, sorry 2024-04-17 08:44:55 I was questioning it after running. 2024-04-17 08:45:09 Suspected either stdout or adding .png to the input. 2024-04-17 08:45:35 Along with questioning if that was the output and it wanted dot as input, which would explain it "doing nothing" if it was waiting for me. :p 2024-04-17 08:45:53 6534x32767 image. Useless to me. 2024-04-17 08:46:00 Get a bigger screen 2024-04-17 08:46:10 It just doesn't have anything 2024-04-17 08:46:39 I'm using the `imv` image viewer, it took a moment to load but desn't seem unhappy to handle it. Just.... big white rectangle. No data. 2024-04-17 08:46:45 Maybe it's so small that you have to zoom in 10000% 2024-04-17 08:46:57 I legit did, 10k% was the max. 2024-04-17 08:47:02 haha 2024-04-17 08:49:25 (still running) 2024-04-17 08:49:35 Good luck 2024-04-17 08:53:34 apk dot sway | dot -Tpng | imv - 2024-04-17 08:53:43 This works and already pushes it enough heh 2024-04-17 08:55:28 :) 2024-04-17 08:58:51 dot: graph is too large for cairo-renderer bitmaps. Scaling by 0.212886 to fit 2024-04-17 08:58:54 real 772.34 2024-04-17 08:59:03 Sofia: apk add docs 2024-04-17 08:59:27 29M /tmp/apk.world.png 2024-04-17 08:59:39 That'd take me from 659 to 870. 2024-04-17 08:59:57 $ doas apk add docs 2024-04-17 08:59:57 After this operation, 1498 MiB of additional disk space will be used. 2024-04-17 09:00:00 hum 2024-04-17 09:00:54 Sofia, nsxiv doesn't have any trouble rendering the graph, if that can help 2024-04-17 09:01:15 Note that it doesn't make it more readable though :D 2024-04-17 09:01:24 Is a terrible representation anyway. 2024-04-17 09:01:30 ^ 2024-04-17 09:01:40 I just want topologically sorted list. 2024-04-17 09:01:42 I think that apk-dot is just an XKCD meme 2024-04-17 09:02:13 Sorry for the loss of time 2024-04-17 09:02:21 Is fine 2024-04-17 09:02:31 ACTION is sleepy and not doing anything really right now so... 2024-04-17 09:04:11 Oh, not a GMT+2 user :) 2024-04-17 09:05:10 ACTION is from Australia, so... plus a few more. :p 2024-04-17 09:05:42 +10 currently, other times +11. 2024-04-17 09:15:42 So you should be relaxing right now! 2024-04-17 09:16:00 Hummm, is there something I need to do manually in order to have the tcl man-pages in the system MANPATH? 2024-04-17 09:16:37 apk add tcl-doc && man n open 2024-04-17 09:16:46 man: No entry for open in section n of the manual. 2024-04-17 09:17:12 apk info -L tcl-doc | grep -F open 2024-04-17 09:17:17 usr/share/man/mann/open.n.gz 2024-04-17 09:17:36 a program i want to compile complains that it cannot find -lunwind, but i can't install libunwind-dev since it conflicts with llvm-libunwind. i have llvm-libunwind-dev installed and it doesn't seem to be detected by my program 2024-04-17 09:18:30 nevermind, i had to install llvm-libunwind-static 2024-04-17 11:33:51 I know this isn't an Apline sepcific issue, but for some reason my doas.conf does not respect the "nopass" at all 2024-04-17 11:41:37 What version of doas do you have? How does your conf file look? 2024-04-17 11:45:11 doas-6.8.2-r6 - https://0x0.st/XoiI.txt 2024-04-17 12:04:47 ikke: anything wrong in my doas.conf? 2024-04-17 12:06:49 P 2024-04-17 12:07:09 pfr: what command do you execute? 2024-04-17 12:11:33 any.. reboot, poweroff, brightnessctl ... they all require password even witn nopass set 2024-04-17 12:12:40 Well, commands are parsed in order, aren't they? 2024-04-17 12:13:24 No sorry, the opposite “The last matching rule determines the action taken.” 2024-04-17 12:14:23 hey is there something like alpine just based on systemd? 2024-04-17 12:14:28 I mean not debian 2024-04-17 12:14:34 just as good as alpine 2024-04-17 12:15:30 I tried arch but hell no, my system was unbootable after installation with linux kernel 2024-04-17 12:15:31 xd 2024-04-17 12:15:52 what's wrong about debian? 2024-04-17 12:16:44 nothing is wrong with debian. 2024-04-17 12:16:50 What do you mean by “like Alpine” though? 2024-04-17 12:16:56 stable 2024-04-17 12:17:08 That's kind of quite subjective :/ 2024-04-17 12:17:35 Or do you mean as a release model? 2024-04-17 12:17:51 well, when I was using alpine I never had issues like on arch linux that randomly stuff broke. 2024-04-17 12:18:03 I think it is rather about the developer working in an distro 2024-04-17 12:18:18 I need rolling release bleeding edge but stable 2024-04-17 12:18:19 pfr, are you really typing "poweroff" 2024-04-17 12:18:21 Or the full path? 2024-04-17 12:19:12 quinq: was this question ment for me? 2024-04-17 12:19:24 No, it was meant for pfr 2024-04-17 12:19:43 (y) 2024-04-17 12:19:45 :) 2024-04-17 12:19:55 quinq: no, I have an alias for poweroff but it's not the full path. it doesn't seem to matter though 2024-04-17 12:20:08 It would 2024-04-17 12:20:46 That was my thinking, that you have to provide the full path 2024-04-17 12:20:49 with or without full paths, it always asks for my password 2024-04-17 12:21:44 If you specify a full path, you need to issue the full path 2024-04-17 12:22:08 I've removed full paths, still wants a password 2024-04-17 12:22:32 Is your username actually "dave"? 2024-04-17 12:22:40 yup 2024-04-17 12:22:57 doas-6.8.2-r6 here works fine 2024-04-17 12:23:28 As in requires a password if full-path is specified but single name is issued 2024-04-17 12:23:39 Doesn't if full-path is issued 2024-04-17 12:23:50 is ubuntu rolling like alpine ? 2024-04-17 12:23:54 Doesn't either if single name is specified 2024-04-17 12:24:04 no matter what my configuration, I cannot get doas to honor nopass 2024-04-17 12:24:48 simon_: no, but debian and opensuse both have good rolling releases 2024-04-17 12:25:11 and they're very stable 2024-04-17 12:33:15 +1, I have heard opensuse tumbleweed fits the "up to date but stable" bill. 2024-04-17 12:56:20 yabobay[m]: how high is the chance that arch linux zen kernel will work? 2024-04-17 12:56:38 i used to use arch and the zen kernel just worked 2024-04-17 12:56:44 things for me in arch just worked in general 2024-04-17 12:57:13 yabobay[m]: can you do pacman -S linux and use it for a while with system shutdowns and so on? 2024-04-17 12:57:17 :D 2024-04-17 12:57:57 sorry what do you mean? 2024-04-17 12:58:16 oh your not using anymore arch? 2024-04-17 12:58:24 yeah, i'm using debian sid now 2024-04-17 12:58:27 which is also pretty good 2024-04-17 12:58:47 how did you install debian sid 2024-04-17 12:58:47 ? 2024-04-17 12:58:54 like what is the fastest way? 2024-04-17 12:59:18 you install a regular release of debian and then migrate it to debian sid 2024-04-17 12:59:18 https://wiki.debian.org/DebianUnstable 2024-04-17 12:59:44 oh no 2024-04-17 12:59:49 that is taking ages 2024-04-17 12:59:54 is there a faster way? 2024-04-17 13:00:28 no, but its actually not that bad 2024-04-17 13:00:44 yabobay[m]: you install the system twice 2024-04-17 13:00:51 :D 2024-04-17 13:00:52 not really 2024-04-17 13:01:03 you install it once and then change the repos 2024-04-17 13:01:13 and then dist-upgrade 2024-04-17 13:01:21 yeah 2024-04-17 13:01:32 doesnt that mean installing twice because every pkg is being exchanged 2024-04-17 13:01:38 :D 2024-04-17 13:01:46 not every package, just the new ones 2024-04-17 13:01:53 it should take a while but it's not really a big deal 2024-04-17 13:02:08 what if I took a debian testing iso and upgrade then? 2024-04-17 13:02:18 would that be faster 2024-04-17 13:03:03 maybe, but the difference shouldn't be huge and the wiki says it's not recommended 2024-04-17 13:03:06 you can try it though 2024-04-17 13:03:56 but what to do if there is a pkg which I want to have which is not available in debian 2024-04-17 13:04:33 i would say you can use flatpak or makedeb which requires a bit of effort 2024-04-17 13:04:45 but a lot of stuff is available in debian 2024-04-17 13:04:50 I have had hell of dependency issue when trying with ubuntu pkgs in debian 2024-04-17 13:04:59 yeah don't do that 2024-04-17 13:06:25 ok reinstall now 2024-04-17 13:06:25 xd 2024-04-17 13:06:31 the 105 time this week 2024-04-17 13:06:58 yabobay[m]: for the record 2024-04-17 13:07:05 this person is kind of an elaborate troll 2024-04-17 13:07:15 huh? how 2024-04-17 13:07:48 like how do we know i mean 2024-04-17 13:08:18 not quite sure how to explain it but 2024-04-17 13:09:37 hmm yeah i don't know how to 2024-04-17 13:09:39 lol 2024-04-17 13:10:22 never mind me i guess 2024-04-17 13:10:23 and i was thinking "hmm this persons acting kinda weird but im gonna be nice" :p 2024-04-17 13:10:47 well i thought that when i first talked with them 2024-04-17 13:12:48 ikke: https://0x0.st/Xo-T.png 2024-04-17 13:14:36 #ahey the person I talked to about Debian are u still here 2024-04-17 13:14:44 which desktop have you got 2024-04-17 13:14:58 pfr: can you try doas -C /etc/doas.conf? 2024-04-17 13:16:54 I will take cinnamon 2024-04-17 13:17:06 ikke: doesn't prompt for password and no output... what does that do exactly? 2024-04-17 13:17:26 Check the configuration 2024-04-17 13:17:39 ok, so no output is good 2024-04-17 13:17:53 I noticed that you used two colons for the group name 2024-04-17 13:18:21 I thought I had solved it earlier when I noticed the file /etc/doas.conf.apk-new 2024-04-17 13:18:39 I deleted that and thought it worked, but nope 2024-04-17 13:20:07 ikke: yeah, I saw someone somewhere on the web suggest two colons 2024-04-17 13:20:13 didn't make much difference 2024-04-17 13:20:49 I'm just so confused as to why it's not working..... works perfectly on NetBSD 2024-04-17 13:21:43 sorry had a dc 2024-04-17 13:22:40 pfr: does it work if you provide your password? 2024-04-17 13:23:19 yes 2024-04-17 13:28:24 i know they've left but i decided to look at the code and while i dont totally get it but around the nopass stuff there was #if defined(USE_SHADOW) so maybe they could install shadow? thats not a dependency on the doas package but maybe 2024-04-17 13:29:24 i dont use alpine but i'd be surprised if shadow isn't already installed 2024-04-17 13:29:49 It works for me without shadow 2024-04-17 13:29:57 hm 2024-04-17 13:30:11 hey I am back now from debian stable desktop 2024-04-17 13:30:25 hello 2024-04-17 13:30:39 triallax: alpine does not come with shadow by default 2024-04-17 13:30:55 right okay 2024-04-17 13:31:11 did not know that 2024-04-17 13:31:15 interesitng 2024-04-17 13:31:26 what provides stuff like passwd then? 2024-04-17 13:31:32 busybox 2024-04-17 13:31:35 ah 2024-04-17 13:31:50 ok then it makes sense yeah 2024-04-17 13:34:02 could be a perms issue or something, hard to know without a trace 2024-04-17 13:35:36 i tried reproducing it here a bunch of different ways, including the ::wheel thing. nope. 2024-04-17 13:42:47 should I remove from Debian SID grub boot? 2024-04-17 13:43:06 this is not a debian channel 2024-04-17 13:43:31 no there is one person in here which I refer the message to he's smarter than Debian developer 2024-04-17 13:43:56 we do have an #alpine-offtopic channel 2024-04-17 13:45:03 I'm on phone 2024-04-17 14:58:28 alright I have systemd-boot now setup on debian 2024-04-17 14:58:55 is the person I had conversation with still in this channeL? 2024-04-17 15:01:06 is there a log of the previous shutdown? My install seems to shutdown every few days 2024-04-17 15:12:09 zombz: it powers itself down? 2024-04-17 15:45:48 # apk add --no-cache py3-ansi2html --repository=http://dl-cdn.alpinelinux.org/alpine/edge/testing/ 2024-04-17 15:45:48 python3-3.11.9-r0: 2024-04-17 15:45:48 ERROR: unable to select packages: 2024-04-17 15:45:48 hi, this command has recently started failing in my alpine container (in my CI): 2024-04-17 15:45:48 breaks: py3-ansi2html-1.9.1-r1[python3~3.12] 2024-04-17 15:45:50 satisfies: py3-ansi2html-1.9.1-r1[python3] python3-3.11.9-r0[python3~3.11] 2024-04-17 15:45:50 what should i do to fix the error? (cc ptrc) 2024-04-17 15:46:38 cc0: switch your image to edge 2024-04-17 15:48:44 hmm, i'm using the docker:cli image, basically apk add-ing yq, this python package, and then running a simple docker build. changing to edge wouldnt be too easy 2024-04-17 15:49:15 maybe i should instead just create a python venv and pip install ansi2html there instead of using the package/the edge repository entirely 2024-04-17 15:50:35 If the package can be moved to community, then it will be part of 3.20 2024-04-17 15:59:21 out of curiosity, how much time is left for stuff to make 3.20/community? (thinking about whether zig 0.12.0 will show up in time) 2024-04-17 16:02:21 if it does not require a lot of other packages to recompile, right before the branch-off point (at release time) 2024-04-17 16:14:05 thanks 2024-04-17 16:23:17 I would like to see zig in community if someone's willing to maintain it 2024-04-17 16:34:41 move 'zig' # to community 2024-04-17 16:34:48 sorry, could not resist 2024-04-17 16:35:47 the thing about zig is, it comes with its own C and C++ compilers... so it could be interesting to see how much of the entire system one can build with it 2024-04-17 16:38:01 but also, supposedly zig/musl isn't quite working well yet: https://github.com/ziglang/zig/issues/11909 2024-04-17 16:44:00 yeah. getting zig strapped everywhere is taking a long time, sadly 2024-04-17 16:44:41 in principle though, i'm enthused to work with it instead of golang, when it's more available 2024-04-17 16:45:49 i agress. zig looks really promising 2024-04-17 17:09:06 personally I prefer golang but it doesn't have to be either-or; zig seems to be a worthy replacement for C, while golang more replaces stuff like Python and Perl for me 2024-04-17 17:09:24 (and rust... will take C++'s spot, probably) 2024-04-17 17:10:03 like, since Golang, my only remaining use for Perl was processing MIDI files. Nothing beats MIDI::Opus yet. 2024-04-17 17:11:29 Python was nice thanks to many libraries for quick and easy game development, but Golang is kinda getting in there too. Like, not just bare SDL, but actually full graphics/sound abstractions (e.g. Ebitengine for anything 2D). Basically pygame has no real reason to still exist ;) 2024-04-17 17:13:23 go is nice but i resent the association with google 2024-04-17 17:35:35 understandable, very so 2024-04-17 17:36:20 one thing I dislike about Go is actually a typical Google thing - like, they're breaking legacy OS support so quickly 2024-04-17 17:36:34 current Go can't even build for Windows 8.1 anymore 2024-04-17 17:37:08 similarly "Go 1.23 will require macOS 11 Big Sur or later." 2024-04-17 17:37:28 meanwhile https://www.tomshardware.com/software/operating-systems/thousands-of-apps-ported-back-to-windows-95-twenty-eight-years-later-net-framework-port-enables-backward-compatibility-for-modern-software 2024-04-17 17:37:56 of course, it's kinda fun... I have a 2012 macbook... running Catalina 2024-04-17 17:38:02 the go team is all on macs now, from what i've been told 2024-04-17 17:38:04 Go 1.23 is gonna no longer support that 2024-04-17 17:38:06 BUT the thing runs Alpine 2024-04-17 17:38:12 Go 1.23 is gonna run just fine on that.... 2024-04-17 17:38:41 I have a late 2013 retina MBP that has not seen macOS in several years 2024-04-17 17:39:13 i like go for what it provides (out of the box), but my # of complaints is going up. zig would address a lot of it. 2024-04-17 17:39:21 yeah, not booted it for a while either 2024-04-17 17:39:27 it's so sluggish on macOS 2024-04-17 17:39:45 for a while it ran cloudready, later upgraded to chromeos flex after Google bought Neverware; about a year ago I switched it to Debian XFCE 2024-04-17 17:39:48 let me say - if Ebitengine had been written for zig, I'd have made my game on zig 2024-04-17 17:40:00 if I want to something in bare SDL, I'll probably do it in zig already 2024-04-17 17:40:22 Ebitengine looks neat! 2024-04-17 17:40:48 just for the game I did, I really didn't want to deal with low level graphics libraries 2024-04-17 17:41:10 like, don't want to write my game three times, once for OpenGL, once for Direct3D, once for MetaL (and soon once more for Vulkan) 2024-04-17 17:41:29 if all I need is pure 2D surfaces though, no shaders... well, I can memcpy sprites around with pure SDL just fine ;) 2024-04-17 17:42:26 similarly I don't want to implement audio 42 times, anew every time Lennart Poettering changes his pants 2024-04-17 17:43:17 but bare SDL takes care of that already 2024-04-17 17:43:27 * that already, just not for 3D APIs 2024-04-17 20:41:43 hi... there is a command-line version of obsidian? 2024-04-17 20:52:26 obsidian is just a text editor, so in that sense, you have a lot of flexibility 2024-04-17 21:04:30 https://paste.debian.net/hidden/2f5fc52c/ 2024-04-17 21:04:32 guys 2024-04-17 21:04:38 can you help with code 2024-04-17 21:04:52 the code is working but it looks like shit 2024-04-17 21:05:20 nevermind fuck this code 2024-04-17 21:05:31 I will come later back when I understand it better 2024-04-17 21:05:31 xd 2024-04-17 23:16:22 "obsidian is just a text editor..." <- obsidian follow links 2024-04-17 23:19:29 yes but fundamentally at its core obsidian is nothing more than a glorified markdown editor 2024-04-17 23:19:51 so you can edit the markdown files with any editor you prefer 2024-04-17 23:23:04 triallax: that is why i am looking for CL version of obsidion 2024-04-17 23:23:19 oh sorry i thought you were somebody else 2024-04-17 23:23:33 didn't notice you're the one who asked the question 2024-04-17 23:24:08 ok 2024-04-17 23:24:21 i assume by "command-line version" you actually mean a tui? 2024-04-17 23:24:26 these two things aren't the same 2024-04-17 23:25:57 yes, tui ... noob addicted to ui... 2024-04-17 23:26:27 obsidian itself doesn't have a tui as far as i know 2024-04-17 23:26:59 https://github.com/toolleeo/cli-apps?tab=readme-ov-file#note-taking 2024-04-17 23:27:04 guess you can take a look at this list 2024-04-17 23:27:29 apparently https://github.com/shashwatah/jot in particular "is a feature-stripped version of Obsidian focused on rapid note management through the terminal" 2024-04-17 23:28:59 but i don't know if it lets you follow links and such 2024-04-17 23:29:21 jot mostly seems to be a cli actually 2024-04-17 23:32:15 triallax: cool.. thanks the attention... i will take a look 2024-04-17 23:45:07 its in rust... now i need to learn how to compile one more crap language... 2024-04-17 23:46:01 building rust programs is pretty easy though 2024-04-17 23:46:45 https://github.com/shashwatah/jot?tab=readme-ov-file#install-with-cargo 2024-04-17 23:55:14 i will try later... i will need to create an alpine docker in distrobox first, to avoid infect my system more... 2024-04-18 00:01:21 infect your system? 2024-04-18 00:07:33 these new languages infest the system with all kind of tools. like python, JS, etc... 2024-04-18 00:07:33 sad they no follow the java philosophy, a VM directory and a $PATH 2024-04-18 00:10:01 funny you say that 2024-04-18 00:10:04 i abhor dealing with anything java 2024-04-18 00:10:20 also it's not required to "infest" your system with cargo 2024-04-18 00:10:38 e.g. if you use any rust program that's packaged in the repos you don't have to install cargo and the rust compiler 2024-04-18 00:20:59 "e.g. if you use any rust program..." <- until today, that was my relationship with rust.... 2024-04-18 00:20:59 that is why I am learning openbsd, to keep my system in the KISS philosophy 2024-04-18 00:21:22 ¯\_(ツ)_/¯ 2024-04-18 00:25:17 I don't have specific links, but I'm sure there's plugins for emacs and (neo)vim that support obsidian-like features 2024-04-18 00:25:30 there's org mode for emacs 2024-04-18 00:39:33 "I don't have specific links, but..." <- emacs and vim ... of course... I need to switch to one of them and stop suffering... 2024-04-18 00:54:28 "¯\_(ツ)_/¯" <- cargo demands me to upgrade the compiler, to build jot, and i will need to upgrade my system to compile a text based editor.... 2024-04-18 00:54:28 No, thanks.. 2024-04-18 00:58:07 it's not really a text editor because it delegates the actual text editing to your choice of editor i believe 2024-04-18 00:58:08 but okay 2024-04-18 02:15:26 jot does look very neat 2024-04-18 02:36:32 All interpreted/JIT languages are cancer when they are not use used for use cases where there is a real need for it. 2024-04-18 02:37:51 is that related to something specific, or? 2024-04-18 02:38:21 A lot of folks are like, "I'm getting good at that interpreted/JIT language so I'll program this awesome thing with it", except if it can be done with a compiled language it should be. 2024-04-18 02:40:36 It is understandable that you're not going to do a backend web app with a compiled language, because the server may not have the CPU to execute it, and because web programmers are second class programmers. 2024-04-18 02:46:13 perhaps you want #alpine-offtopic? 2024-04-18 02:48:31 perhaps 2024-04-18 05:46:05 Is there any reason why dri driver's path is /usr/lib/xorg/modules/dri? 2024-04-18 05:46:31 Why isn't it just /usr/lib/dri? 2024-04-18 05:47:48 They're used on wayland too, so I don't see any reason to package it in xorg related directory 2024-04-18 06:00:56 hmm I know Arch used the xorg/modules path until ~2017, maybe it was just copied at the time 2024-04-18 06:03:07 Oh, huh, that I wasn't aware of, not like I used arch at that time anyway, but that's pretty interesting 2024-04-18 08:13:53 Dug a little bit in arch's IRC and apparently upstream is using /usr/lib/dri, just like Arch does now 2024-04-18 08:14:23 Do you think it would be a good idea to drop this historical path and start using /usr/lib/dri like upstream? 2024-04-18 08:31:45 Not sure if anyone is aware, but the python upgrade in edge breaks the aws-cli: 2024-04-18 08:32:01 python3-3.12.3-r0: breaks: aws-cli-2.15.14-r0[python3~3.11] 2024-04-18 08:32:17 Probably not needed, as in, cli might just need to be updated as a package 2024-04-18 10:08:17 telmich: aws-cli was disabled iirc because it was broken 2024-04-18 12:26:07 yes, aws-cli doesn't work with python 3.12 2024-04-18 12:26:12 upstream issue: https://github.com/aws/aws-cli/issues/8342 2024-04-18 12:48:41 Hello! 2024-04-18 12:48:45 你好！ 2024-04-18 12:50:35 Hi 2024-04-18 12:50:48 Hahahahaha! 2024-04-18 12:51:22 How man? 2024-04-18 12:51:43 Let`s play on c language! 2024-04-18 12:52:47 Hello everybody 2024-04-18 12:53:01 I'm sorry for my english!! 2024-04-18 12:53:01 I want to install alpine on LVM partitions, I created 1 EFI System partition. 2024-04-18 12:53:01 Do I still need a boot partition or do I mount the EFI partition in /mnt/boot/efi? 2024-04-18 12:53:14 ? 2024-04-18 12:53:31 What are your chat? 2024-04-18 12:58:07 Hello! 2024-04-18 13:02:01 Hello!/ 2024-04-18 13:02:05 Hello!// 2024-04-18 13:02:11 Hello!// 2024-04-18 13:02:13 Hello!/////// 2024-04-18 13:02:21 qiaoran: pease calm down 2024-04-18 13:02:24 please* 2024-04-18 13:02:36 hahahahaha! 2024-04-18 13:03:29 Do not use dirty text 2024-04-18 13:04:22 hello 2024-04-18 13:05:15 test 2024-04-18 13:06:30 what 2024-04-18 13:08:40 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa 2024-04-18 13:08:47 ==================================================== 2024-04-18 13:09:39 gnd_ro: you should be fine with just EFI on /boot/efi 2024-04-18 13:10:34 ok,thanks 2024-04-18 13:45:26 hello 2024-04-18 13:45:38 good morning 2024-04-18 13:45:58 I why well be ban chat permission? 2024-04-18 13:46:24 I why will be closed chat mode? 2024-04-18 13:47:08 AndroUser: You said you will behave yourself 2024-04-18 13:58:54 I apologize deeply for that. 2024-04-18 14:53:56 hello 2024-04-18 14:53:58 guys 2024-04-18 14:54:16 is it easy to setup alpine on proprietary hardware 2024-04-18 14:54:16 ? 2024-04-18 14:54:30 like with iwlwifi and bluetooth proprietary blobs 2024-04-18 14:54:30 ? 2024-04-18 15:01:11 most hardware is proprietary 2024-04-18 15:04:07 lopid so it will be no issue 2024-04-18 15:04:08 ? 2024-04-18 15:04:18 it wasn't for me 2024-04-18 15:04:48 is alpine intendet do be used as desktop lopid 2024-04-18 15:04:50 ? 2024-04-18 15:04:57 or is there a secure alternative 2024-04-18 15:05:05 whether it's easy really depends on the specific hardware but in general alpine hardware support is pretty good 2024-04-18 15:05:15 it is a fine desktop linux distro, quite a few folks use it that way 2024-04-18 15:05:27 I just heard that alpine is the least secure os out there 2024-04-18 15:05:33 from some arch users 2024-04-18 15:05:43 okay? 2024-04-18 15:06:05 unless you see that backed up with a pretty convincing argument I would not put too much weight on it 2024-04-18 15:06:24 kinda sounds like someone just trying to start a fight to me :P 2024-04-18 15:06:31 well they said every random person can sumbit pkgs 2024-04-18 15:06:41 that's a load of nonsense 2024-04-18 15:06:42 They'd not be correct with that statement 2024-04-18 15:06:45 anyone can *propose* one, they're all reviewed 2024-04-18 15:06:50 that applies to most distros out there 2024-04-18 15:06:51 including arch 2024-04-18 15:07:02 hmm 2024-04-18 15:07:04 anyway, alpine is a fairly pleasant desktop system 2024-04-18 15:07:04 anybody can submit package templates but they are reviewed 2024-04-18 15:07:10 ok then I will install alpine on my mini pc 2024-04-18 15:07:12 unlike say, the AUR, where nothing is reviewed at all 2024-04-18 15:07:36 does alpine support bluethooth? 2024-04-18 15:07:43 because I really require this 2024-04-18 15:07:56 like every distro, it depends on your hardware 2024-04-18 15:08:46 your best bet is probably to boot alpine off a usb stick and see whether the hardware works properly or not 2024-04-18 15:08:54 and if it does, then do a real install 2024-04-18 15:09:00 is there a bleeding edge iso or img 2024-04-18 15:09:07 yes, it's called "edge" 2024-04-18 15:09:17 There is no iso though 2024-04-18 15:09:32 community/ 17-Apr-2024 19:15 - main/ 14-Apr-2024 07:07 - releases/ 30-Jul-2021 10:02 - testing/ 16-Apr-2024 01:42 which one do I want? 2024-04-18 15:09:36 oh, iso, yeah 2024-04-18 15:10:16 that looks like an aports tree, whereas you probably want either standard or extended from here: https://alpinelinux.org/downloads/ 2024-04-18 15:10:22 do I take the main folder of edge? 2024-04-18 15:10:27 no you do not 2024-04-18 15:10:52 you boot an alpine image (from the downloads page) then you configure it to use the edge repos if you want edge packages 2024-04-18 15:11:01 https://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/ 2024-04-18 15:11:04 this or not? 2024-04-18 15:11:07 elly: i thought there isn't an iso? 2024-04-18 15:11:17 sadas: those are just the packages, not the installation medium 2024-04-18 15:12:15 there is no iso for edge, yeah 2024-04-18 15:12:43 ah yeah, right 2024-04-18 15:13:07 https://dl-cdn.alpinelinux.org/alpine/edge/releases/x86_64/ 2024-04-18 15:13:40 oh it is even beta 2024-04-18 15:13:41 lol 2024-04-18 15:13:46 should work or? 2024-04-18 15:14:14 sudo dd if=alpine-edge-140416-x86_64.iso of=/dev/sdc bs=4M status=progress oflag=sync 2024-04-18 15:14:29 woa this is a small image 2024-04-18 15:14:30 xd 2024-04-18 15:15:22 that image is from 2014... 2024-04-18 15:15:31 oh 2024-04-18 15:15:33 xD 2024-04-18 15:16:03 go to the actual downloads page, which is where the current official images are linked, and grab either a standard or an extended image for your arch, boot that, then configure it to use the edge repos 2024-04-18 15:22:41 where can I find mirrors in /etc/apk? 2024-04-18 15:22:43 or something 2024-04-18 15:23:40 do I change main to edge 2024-04-18 15:23:40 ? 2024-04-18 15:23:55 https://mirrors.alpinelinux.org/ 2024-04-18 15:24:50 I got it 2024-04-18 15:24:56 so how to upgrade alpine 2024-04-18 15:25:17 edit etc/apk/repo and apk upgrade 2024-04-18 15:25:42 nice 2024-04-18 15:26:30 I tried setup-desktop script it says sddm and dbus not available 2024-04-18 15:26:31 :o 2024-04-18 15:26:56 should I activate community stuff? 2024-04-18 15:27:48 community stuff got activated by default 2024-04-18 15:28:38 gnome works 2024-04-18 15:28:44 plasma does not work 2024-04-18 15:31:35 seems like it supports bluetooth 2024-04-18 15:31:35 nice 2024-04-18 15:35:46 how to turn on bluetooth in alpine? 2024-04-18 15:39:42 bruh 2024-04-18 15:39:45 I love alpine 2024-04-18 15:39:45 xd 2024-04-18 15:43:05 I just wished alpine coule do appimages 2024-04-18 15:43:15 then I would completly wipe all other systems 2024-04-18 15:43:15 xd 2024-04-18 15:45:31 where is patricia? 2024-04-18 16:02:13 what 2024-04-18 16:02:18 what do you need from her 2024-04-18 16:02:37 lmao 2024-04-18 16:02:48 do you know them lol 2024-04-18 16:02:59 i recognize the writing style 2024-04-18 16:03:05 random person from the mckay circus 2024-04-18 16:04:33 didn't notice they disconnected lol 2024-04-18 16:06:26 hey I am back 2024-04-18 16:06:38 can you tell me what I can do against missing fonts? 2024-04-18 16:07:27 you can try installing some fonts 2024-04-18 16:07:35 can I just install all of them 2024-04-18 16:07:35 ? 2024-04-18 16:08:42 is community repo in alpine like aur in arch? 2024-04-18 16:08:52 no 2024-04-18 16:09:41 animations in alpine are really smooth compared to other os 2024-04-18 16:21:20 can I compile openjdk22 my self on alpine? 2024-04-18 16:21:47 yes 2024-04-18 16:22:01 I did it already on linux mint but dunno if it is harder on alpine linux 2024-04-18 16:22:19 or is it maybe even smarter to run both debian and alpine? 2024-04-18 16:22:35 alpine for bluetooth and videos 2024-04-18 16:22:40 and debian for programming 2024-04-18 16:22:41 xd 2024-04-18 16:23:26 becuz I dont really have an alternative for an appimage 2024-04-18 16:41:13 how to install fuse? 2024-04-18 16:41:53 apk install fuse like almost anything else? 2024-04-18 16:42:06 apk add even... 2024-04-18 16:42:32 https://pastebin.com/pKCNLMUd 2024-04-18 16:42:48 I insstalled many fuse pkgs 2024-04-18 16:42:49 xd 2024-04-18 16:43:12 overkill, but works. 2024-04-18 16:44:44 is there a difference between libfuse2 and fuse on alpine? 2024-04-18 16:46:22 https://pastebin.com/97jZzmCp so I have no idea why this does not work but it does not work xd 2024-04-18 16:47:37 simon_: try 'modprobe fuse' 2024-04-18 16:48:11 ~/Downloads $ ./ledger-live-desktop-2.79.0-linux-x86_64.AppImage Error relocating /tmp/.mount_ledgerniDIFN/ledger-live-desktop: posix_fallocate64: symbol not found 2024-04-18 18:08:27 sorry had a disconnect, did someone wrote something about Error relocating /tmp/.mount_ledgerniDIFN/ledger-live-desktop: posix_fallocate64: symbol not found 2024-04-18 18:15:28 simon_: try installing gcompat 2024-04-18 18:16:00 I have gcompat installed 2024-04-18 18:16:18 ~/Downloads $ apk list gcompat gcompat-1.1.0-r4 x86_64 {gcompat} (NCSA) [installed] 2024-04-18 18:16:43 ~/Downloads $ gcompat /bin/ash: gcompat: not found when executing 2024-04-18 18:17:37 maybe I have to reboot? 2024-04-18 18:21:04 I don't think it'll help 2024-04-18 18:21:27 ~/Downloads $ scanelf -i ledger-live-desktop-2.79.0-linux-x86_64.AppImage TYPE INTERP FILE ET_EXEC /lib64/ld-linux-x86-64.so.2 ledger-live-desktop-2.79.0-linux-x86_64.AppImage 2024-04-18 18:21:32 I have also this info 2024-04-18 18:22:43 https://git.adelielinux.org/adelie/gcompat can you check this section ```Running binaries linked against glibc``` 2024-04-18 18:22:56 there is something unclear for me 2024-04-18 18:25:04 actually I dont really understand what is written there at all 2024-04-18 18:38:27 oh 2024-04-18 18:38:36 alpine no new version of glibc 2024-04-18 18:38:37 :( 2024-04-18 18:39:31 alpine doesn't use glibc 2024-04-18 18:39:41 damn 2024-04-18 18:40:52 would it be possible to build glibc on top of alpine? 2024-04-18 18:41:41 you can build glibc, but then why do you use musl-based distro in the first place? 2024-04-18 18:42:09 Ermine: alpine is cool 2024-04-18 18:42:20 very to no bugs 2024-04-18 18:42:24 but it doesn't satisfy your needs 2024-04-18 18:42:45 but no bugs 2024-04-18 18:44:22 but you still need to overcome the need for glibc 2024-04-18 18:45:02 my program only available for glibc 2024-04-18 18:45:04 :( 2024-04-18 18:47:36 There's little we can do about propietary programs 2024-04-18 18:47:55 gcompat 2024-04-18 18:48:04 I tried with gcompat 2024-04-18 18:48:08 did not work 2024-04-18 18:48:13 Missing symbols? 2024-04-18 18:48:31 :21] ~/Downloads $ scanelf -i ledger-live-desktop-2.79.0-linux-x86_64.AppImage TYPE INTERP FILE ET_EXEC /lib64/ld-linux-x86-64.so.2 ledger-live-desktop-2.79.0-linux-x86_64.AppImage 2024-04-18 18:48:36 too old I guess 2024-04-18 18:49:08 Not sure what this output is supposed to mean 2024-04-18 18:49:24 Error relocating /tmp/.mount_ledgerjClfFk/ledger-live-desktop: posix_fallocate64: symbol not found this 2024-04-18 18:49:35 There we go, much better 2024-04-18 18:49:38 This means that function posix_fallocate64 isn't present in current release of gcompat 2024-04-18 18:49:47 Yep 2024-04-18 18:50:00 You can wait for the next release of gcompat 2024-04-18 18:50:18 or use an older version of alpine 2024-04-18 18:50:20 I actually tried building gcompat myself and adding stubs for missing implementations 2024-04-18 18:50:33 Or using glibc's implementations in gcompat 2024-04-18 18:50:42 I took it as far as running Portal on musl 2024-04-18 18:50:48 It worked, but had bugs 2024-04-18 18:50:55 ikke: why older? 2024-04-18 18:50:59 I thought newer xd 2024-04-18 18:51:25 older versions of musl had shims for lfs64 symbols 2024-04-18 18:51:33 newer versions removed them as they have no purpose of musl 2024-04-18 18:51:37 Yeah edge branch is the best you can get without building youself 2024-04-18 18:51:46 can I just inject an older pkg in new alpine? 2024-04-18 18:51:55 unlikely without breaking things 2024-04-18 18:52:06 You can but why would you do that? 2024-04-18 18:52:08 can I roll back or reinstall? 2024-04-18 18:52:14 you can roll bakc 2024-04-18 18:52:16 reinstall 2024-04-18 18:52:21 oh, ok 2024-04-18 18:52:36 rolling back by just adapting source list and update upgrade? 2024-04-18 18:52:47 simon_: make sure you provide --available to the upgrade command 2024-04-18 18:52:57 I used distrobox + podman to run some glibc applications 2024-04-18 18:52:58 https://wiki.alpinelinux.org/wiki/Running_glibc_programs#Containers 2024-04-18 18:54:27 ikke: >older versions of musl had shims for lfs64 symbols. Dunno if that would help. Imo it's best if it was added on gcompat side 2024-04-18 18:54:36 yes, it would 2024-04-18 18:54:54 Therefore I don't think you should be using older musl 2024-04-18 18:54:55 And I may be misremembering that they were only compile time shims 2024-04-18 18:55:17 so would not help for missing symbols at runtime 2024-04-18 18:56:22 It is actually https://git.adelielinux.org/adelie/gcompat/-/commit/8a7e183ed3cf372563dc263066035870c3ea6995 2024-04-18 18:57:54 ikke: if musl had those symbols, removing them would constitute abi break 2024-04-18 18:58:01 nod 2024-04-18 18:58:25 ~/Downloads $ ./ledger-live-desktop-2.79.0-linux-x86_64.AppImage Error relocating /tmp/.mount_ledgercDJDfe/ledger-live-desktop: posix_fallocate64: symbol not found rolled back but still 2024-04-18 18:58:51 simon_: yeah, was just discussed, they were only compile-time shims, not runtime provided symbols 2024-04-18 18:59:08 simon_: Ermine linked to https://git.adelielinux.org/adelie/gcompat/-/commit/8a7e183ed3cf372563dc263066035870c3ea6995 2024-04-18 18:59:42 that means it doesnt work? 2024-04-18 19:00:31 correct, you need gcompat with that commit 2024-04-18 19:00:39 no, it does. I just don't think this commit is in alpine branches yet 2024-04-18 19:01:04 can I just takt this commit myself? 2024-04-18 19:01:05 xd 2024-04-18 19:01:20 simon_: Sure, you can build gcompat with that applied as a patch 2024-04-18 19:01:37 I'd just clone repo and build it lmao 2024-04-18 19:02:10 but where can I see which I need to clone + commit 2024-04-18 19:02:36 I cant really see anything on these sides 2024-04-18 19:02:44 I cant navigate through it 2024-04-18 19:03:05 oh wait 2024-04-18 19:03:07 nevermind 2024-04-18 19:03:11 I think I know how 2024-04-18 19:04:10 https://git.adelielinux.org/adelie/gcompat/-/commit/8a7e183ed3cf372563dc263066035870c3ea6995.patch 2024-04-18 19:05:48 Any gdm users on alpine? Just wondering if you're using xorg or wayland session 2024-04-18 19:06:46 they have just added this single line or libgcompat/fcntl.c \? 2024-04-18 19:08:15 simon_: read the patch, it adds posix_fallocate64 2024-04-18 19:09:37 +++ b/libgcompat/fcntl.c lassebq what I just mentioned that in this file there was no difference xd 2024-04-18 19:10:19 do I just build gcompat by writing make? 2024-04-18 19:10:31 What? Are you adding the patch on top of a tree with it merged already? 2024-04-18 19:10:49 make all, not make 2024-04-18 19:11:32 Or nvm make probably does the job too 2024-04-18 19:13:46 so I have build it 2024-04-18 19:13:54 still same error 2024-04-18 19:14:22 Did you also install it? Just building it is not enough 2024-04-18 19:14:49 install -D -m755 libgcompat.so.0 //lib/libgcompat.so.0 install: can't create '//lib/libgcompat.so.0': File exists make: *** [Makefile:98: install] Error 1 2024-04-18 19:15:14 oh I still have some part of this installedd ? 2024-04-18 19:16:09 can I see which pkg is in relation to gcompat 2024-04-18 19:16:31 apk info -W /lib/libgcompat.so.0 2024-04-18 19:16:57 No package in alpine linux itself should depend on gcompat 2024-04-18 19:17:13 ~/Downloads/gcompat $ doas apk del libc6-compat 2024-04-18 19:18:24 ~/Downloads $ ./ledger-live-desktop-2.79.0-linux-x86_64.AppImage /bin/ash: ./ledger-live-desktop-2.79.0-linux-x86_64.AppImage: not found 2024-04-18 19:18:29 ok this is another error xd 2024-04-18 19:18:57 You're missing the dynamic loader 2024-04-18 19:18:59 you didn't install built gcompat 2024-04-18 19:19:11 ~/Downloads/gcompat $ doas make install install -D -m755 libgcompat.so.0 //lib/libgcompat.so.0 install -D -m755 ld-linux.so.2 //lib/ld-linux.so.2 2024-04-18 19:19:18 I did :D 2024-04-18 19:19:46 it expects '/lib64/ld-linux-x86-64.so.2' 2024-04-18 19:20:09 strace it then ig 2024-04-18 19:20:51 Or maybe use that /lib64 path ¯\_(ツ)_/¯ 2024-04-18 19:21:34 tbh, easiest / less messy to include the patch file in the apkbuild file and build that 2024-04-18 19:25:37 I've been building git repo and manually doing `cp ld-linux.so.2 /lib/ld-linux.so.2`. Everything works as it should. 2024-04-18 19:26:10 and `cp libgcompat.so /lib/libgcompat.so.0` 2024-04-18 19:26:52 ET_EXEC points to /lib64/ld-linux-x86-64.so.2, so that file should at least exist 2024-04-18 19:27:07 symlink /lib64 to /lib 2024-04-18 19:27:11 that's what I do 2024-04-18 19:27:20 The filename also has the arch in it 2024-04-18 19:27:41 Uh.. yeah makes sense 2024-04-18 19:27:51 ld-linux.so.2 is for 32 bit 2024-04-18 19:28:13 I have both in /lib 2024-04-18 19:28:42 Btw, I (sorta) ran alpine with 32 libs 2024-04-18 19:31:58 no idea I guess I am to stupid to run it 2024-04-18 22:26:29 hello 2024-04-18 22:26:41 what the hell! 2024-04-18 22:38:21 你们的祖宗过的还好吗？ 2024-04-18 22:39:01 在问你们话呢！ 2024-04-18 22:39:37 别以为你们这群养老禁我言，就当老子不发威！ 2024-04-18 22:39:48 english please 2024-04-18 22:40:11 that was a language? i thought it was just unicode garbage 2024-04-18 22:40:20 你们这群洋老子，我不会容忍任何形式的处罚行为！ 2024-04-18 22:40:22 it is simplified chinese 2024-04-18 22:40:24 ok 2024-04-18 22:41:12 i just get a bunch of invalid and confused junk in my terminal 2024-04-18 22:41:24 AndroUser: this is a volunteer help channel, that kind of talk will not be tolerated 2024-04-18 22:42:37 dewdude: what you're missing out on, after using google translate to make it english: https://bpa.st/6MAQ 2024-04-18 22:42:53 they were already banned once 2024-04-18 22:42:55 so this is ban evasion 2024-04-18 22:43:01 oh cool 2024-04-18 22:43:09 triallax: thanks 2024-04-18 22:44:07 actually copy/paste from term in to google worked. but i recgonized that tactic and give the trolls a hard time 2024-04-18 22:44:38 heh 2024-04-18 22:54:10 dwfreed: wait no sorry i was confused 2024-04-18 22:54:13 they were +q'd, not banned 2024-04-18 22:54:19 still evasion 2024-04-18 22:54:21 but still i think banning was the right move 2024-04-18 23:52:51 hello 2024-04-18 23:52:54 I m sorry 2024-04-18 23:54:05 I shouldn't have made that mistake 2024-04-18 23:55:42 You just blocked one of my IP addresses 2024-04-18 23:56:07 Can you remove it first? 2024-04-18 23:56:09 #oftc 2024-04-18 23:58:18 I don't have any ill will 2024-04-18 23:58:38 take it to #oftc please 2024-04-18 23:58:39 I'm sorry to have completely lost face 2024-04-18 23:59:45 I just want to get a real forgiveness 2024-04-19 00:02:28 Can you unblock it? 2024-04-19 00:02:45 My tablet has been denied access by you 2024-04-19 00:03:01 I'm currently unable to connect to your IRC server due to a ban 2024-04-19 00:03:36 If you don't see my message, please respond 2024-04-19 00:05:06 草泥马 2024-04-19 00:05:33 if you can join this channel, you can join #oftc as well. 2024-04-19 00:06:41 I just wanted to ask you to unblock the IP address or hostname that you just blocked 2024-04-19 00:07:07 That one was accessed on my tablet 2024-04-19 00:09:01 Please let me know if you're really just here to give feedback 2024-04-19 00:09:16 I would change the style of chatting 2024-04-19 00:09:44 The ban was administered by network staff. Only they can lift this ban. Hence, go to #oftc and ask them. 2024-04-19 00:10:56 ok 2024-04-19 00:24:53 傻逼 2024-04-19 00:58:23 am i confused, or is there no way to run podman commpose on alpine? 2024-04-19 00:59:29 s/commpose/compose/ 2024-04-19 00:59:53 i don't see podman-compose in latest-stable main, community, nor edge/testing 2024-04-19 01:58:47 不要说脏话(Please don't use foul language in Chinese) 2024-04-19 02:25:15 Are you still alive? 2024-04-19 02:26:55 @qaqland> 你会中文？ 2024-04-19 02:32:58 A person with multiple personality plus schizophrenia plus bipolar personality disorder 2024-04-19 02:33:00 isme 2024-04-19 02:33:07 Is me 2024-04-19 02:33:37 hello 2024-04-19 02:33:39 help me 2024-04-19 02:36:19 操你妈 2024-04-19 02:36:23 操你妈的！ 2024-04-19 02:36:32 我感觉自己就是个傻逼！ 2024-04-19 02:40:16 傻逼！ 2024-04-19 02:51:57 傻逼！ 2024-04-19 02:52:19 I m super man 2024-04-19 02:55:14 a 2024-04-19 08:13:05 Hm 2024-04-19 08:16:00 It seems enabling Preferred Ozone platform from chrome://flags is not enough anyome? I have to manually start chromium with: `chromium --enable-features=UseOzonePlatform --ozone-platform=wayland` flags. Can I somehow automate this process? Some config file I could add those 2 flags? 2024-04-19 08:17:01 I tried to add to: ~/.config/chromium-flags.conf but it's ignored 2024-04-19 08:19:09 Another Issue I found out is why I can't see this strawberry insisde foot terminal? 🍓 I can see it just fin in a browser. Am I missing some fonts? System wide? 2024-04-19 09:19:43 xyhhxtheythem[m]: the podman-compose package is in edge (but not in 3.19) 2024-04-19 10:03:03 if i never install openrc, I will never get *-openrc packages. right?! 2024-04-19 10:09:48 most probably, yes 2024-04-19 10:12:43 awesome 2024-04-19 10:42:51 I've solved the font issue. I've installed font-not-emoji and ~/.config/fontconfig/conf.d/50-emoji.conf with https://termbin.com/a5kw 2024-04-19 10:46:44 do you actually need that config file 2024-04-19 10:46:47 i don't think you do 2024-04-19 11:09:05 hi, i'm thinking of switching to alpine from arch, but i think i'll miss the AUR 2024-04-19 11:09:38 what should i do if i want to use a program that isn't in the official repos? 2024-04-19 11:13:29 look at PKGBUILD and then build it yourself 2024-04-19 11:14:42 the AUR PKGBUILD? 2024-04-19 11:16:25 would i convert it to a APKBUILD? 2024-04-19 11:16:54 or just execute the commands in the PKGBUILD manually 2024-04-19 11:17:59 user's choice 2024-04-19 11:18:40 thanks, switching seems less daunting now 2024-04-19 11:33:18 bluemoon: there's a tool that can help convert pkgbuilds to apkbuilds. You can install it with apk add toapk. 2024-04-19 11:34:56 thanks for the info 2024-04-19 11:45:24 No problem, there are also plenty of folks here who will happily add things to the ecosystem if they're missing and someone asks nicely. Or help out if you're trying yourself and just need a hand. 2024-04-19 12:18:15 triallax: What do you mean if I need that config file? Just the 50-emoji.conf file? And directory should be there? `~/.config/fontconfig/conf.d` ? 2024-04-19 12:31:05 if the directory is empty it doesn't matter if it exists 2024-04-19 12:31:15 but yeah i was referring to 50-emoji.conf 2024-04-19 12:36:38 Before working emojis I didn't have `~/.config/fontconfig/conf.d` dir hmm.. let me test without 50-emoji.conf 2024-04-19 12:38:43 You are right, I don't even need that 50-emoji.conf just the `~/.config/fontconfig/conf.d` 2024-04-19 12:44:06 are you sure? try removing it 2024-04-19 14:43:45 Hello 2024-04-19 14:50:00 dwfreed 2024-04-19 14:50:29 hello 2024-04-19 14:50:57 Hao arelsla yo 2024-04-19 14:51:30 有人吗？ 2024-04-19 15:48:58 https://gitlab.alpinelinux.org/alpine/aports/-/issues/16016 2024-04-20 00:10:46 is it just me or is libreoffice broken in edge right now? 2024-04-20 04:03:54 There is an issue about it 2024-04-20 04:29:59 o.o hi 2024-04-20 04:30:24 how can i go about setting up a ssh server 2024-04-20 04:33:06 apk add openssh-server; rc-service sshd start; rc-update add sshd 2024-04-20 04:35:37 Hello 2024-04-20 04:35:40 specifically how would i set up a ssh server in alpine that i can then login to without requiring any user interaction 2024-04-20 04:35:40 Of course, we remember you. How can I assist you today? 2024-04-20 04:36:05 test 2024-04-20 04:36:08 eg, any linux : ssh .... # logs into alpine linux ssh 2024-04-20 04:36:53 我们还是仇人吗? 2024-04-20 04:37:42 or rather, any OS that has open ssh binary 2024-04-20 04:38:30 anyway how can i set up such in alpine 2024-04-20 04:39:28 androidui: You can add an ssh key without password 2024-04-20 04:39:41 androidui: that's basic ssh 2024-04-20 04:39:56 dont most setups disallow passwordless login tho ? 2024-04-20 04:40:51 I must have said passphrase 2024-04-20 04:41:07 you can also add a passphrase and use an ssh agent to keep it in memory 2024-04-20 04:41:31 ikke: you didnt say passphrase 2024-04-20 04:42:34 its just mose things i looked up have, for example ssh user@hostname and then requiring a password to login, which would fail if the user does not have a password set (eg, passwordless root user) 2024-04-20 04:42:37 most* 2024-04-20 04:43:33 The default setup disallows password login for root anyway 2024-04-20 04:44:30 eg https://serverfault.com/questions/241588/how-to-automate-ssh-login-with-password https://stackoverflow.com/questions/61833713/how-to-login-by-ssh-in-alpine-linux-without-passwords 2024-04-20 04:46:38 androidui: like I said, use an ssh key instead 2024-04-20 04:47:17 alright 2024-04-20 04:47:42 how can i get alpine to automatically log me in to the tty 2024-04-20 04:48:18 eg localhost login: password: Welcome to Alpine! 2024-04-20 04:49:01 how can i skip the login and auto login as root user 2024-04-20 05:05:56 Hello to you too. 2024-04-20 05:06:15 We are Anonymous. 2024-04-20 05:06:23 We are a legion. 2024-04-20 05:06:31 We cannot forgive. 2024-04-20 05:06:39 We cannot forget. 2024-04-20 05:06:47 Expect us. 2024-04-20 05:07:38 bruh 2024-04-20 05:10:29 hmm https://wiki.alpinelinux.org/wiki/TTY_Autologin 2024-04-20 05:12:35 hello 2024-04-20 05:12:44 I m qiaoran 2024-04-20 05:12:56 I feel like an infamous existence in the entire online world. Is there still a need for me to keep going? 2024-04-20 05:13:15 I'm very sorry about that. 2024-04-20 05:17:11 thanks for moderating ikke 2024-04-20 05:20:19 No problem 2024-04-20 05:33:16 ok, so i did apk add agetty and added tty1::respawn:/sbin/agetty --autologin root tty1 linux and then # reboot 2024-04-20 05:39:55 yay it logs in ^~^ 2024-04-20 05:40:38 oh it glitched ;-; 2024-04-20 05:41:49 i broke it ;-; 2024-04-20 05:43:06 luckily i can still use tty 2 2024-04-20 05:43:30 https://imgur.com/a/fyx7FV5 2024-04-20 05:43:34 how can i fix this? 2024-04-20 05:44:26 https://imgur.com/a/xiiZfxP this is my /etc/inittab 2024-04-20 06:10:24 using mintty kinda works but still ends up auto-exiting to login: thing 2024-04-20 06:13:06 hmm https://unix.stackexchange.com/questions/751105/how-to-enable-automatic-login-in-alpine-linux-for-root 2024-04-20 06:50:31 or rather i get login timed out after 60 seconds 2024-04-20 07:14:04 ok i think the problem was having multiple tty1:: lines in inittab so getty was started on tty1 then bash was started on tt1 which interlaces the output of agetty on tty1 and confuses the system 2024-04-20 07:48:37 Autologin is great. Its nice to just enter your encryption password and then go straight to a desktop. 2024-04-20 07:55:06 thats AI working for you 2024-04-20 08:10:29 ok i got it working 2024-04-20 08:11:08 tty1::respawn:/sbin/agetty --autologin root 38400 tty1 2024-04-20 08:11:42 tty2::respawn:/bin/bash -i tty2 also works 2024-04-20 08:14:08 tty2::respawn:/bin/bash -i also works * as does tty3::respawn:/sbin/getty -n -l /bin/bash 38400 tty3 (also works) 2024-04-20 08:15:14 tho im not sure if such are the correct way to login (eg, allocate a tty so bash job control works) 2024-04-20 08:22:18 i want to automatically decrypt a luks-encrypted drive partition on boot and mount an lvm volume. i achieve this with an entry in /etc/conf.d/dmcrypt and it works if i specify the partition with "source='/dev/sda1'" in /etc/conf.d/dmcrypt, but specifying the uuid of the partition doesn't work. what am i missing? 2024-04-20 08:22:44 udev is in the sysinit runlevel so i'd assume the uuid is already available then 2024-04-20 08:23:06 UUID is an fs-level attribute; unless you mean PARTUUID 2024-04-20 08:23:51 i meant the UUID. don't know too much about them or their difference tbh 2024-04-20 08:24:00 is PARTUUID what I'm looking for, then? 2024-04-20 08:24:05 most likely 2024-04-20 08:24:12 i'll give it a shot, thanks! 2024-04-20 08:24:16 use 'blkid' to print all known attributes 2024-04-20 08:24:41 right 2024-04-20 08:25:13 ah, UUID doesn't seem to make sense here, right. now i know the difference lol 2024-04-20 08:28:32 hm no that doesn't seem to solve the problem either. giving the PARTUUID gives the same result as UUID, namely the message 'source "..." for ... missing, skipping...' 2024-04-20 08:33:23 paste your source= line? 2024-04-20 09:09:50 source='27abc22e-5d53-4417-851c-e4c0ecc36a41' 2024-04-20 09:42:03 hi. i'm on sway. since my upgrade to chromium 124, "it" doesn't start anymore: https://0x0.st/XoVl.txt 2024-04-20 09:43:07 don't know why "it want's X" 2024-04-20 09:46:11 oh, i need the flag --ozone-platform=wayland / but we already have --ozone-platform-hint=auto in /etc/chromium/chromium.conf ... 2024-04-20 09:55:37 seems like the flag in /etc/chromium.conf --ozone-platform-hint=auto doesn't work anymore 2024-04-20 11:37:15 something is amiss with https://pkgs.alpinelinux.org/package/edge/community/x86_64/libreoffice-common -- it shows contents but claims zero bytes, and it looks like an empty package on my machine 2024-04-20 12:21:32 sounds like #16005 2024-04-20 12:26:43 anyone know an alternative to the fmt(1) command? 2024-04-20 12:51:07 what's wrong with fmt? 2024-04-20 12:53:41 dwfreed: requires fetching GNU coreutils 2024-04-20 12:53:52 yes it can be standalone but i'm wondering if there are alternatives 2024-04-20 12:54:50 I don't know of any 2024-04-20 14:43:45 how do i get ssh to accept my ssh key 2024-04-20 14:43:56 im doing ssh -v -o HostBasedAuthentication=yes -o UserKnownHostsFile=ssh_key.pub -o IdentityFile=ssh_key -o PubkeyAuthentication=yes 2024-04-20 14:44:03 but it still asks for a password 2024-04-20 14:44:39 and i cant use sshpass/expect since ssh -t wants a psuedo tty 2024-04-20 14:45:09 (eg, ssh bash will work but no input will be entered) 2024-04-20 14:51:25 if i cat ssh_key.pub >> ~/.ssh/authorized_keys then it still asks for password 2024-04-20 14:52:36 debug1: Offering public key: ssh_key ED25519 SHA256:Ru/hbc0HWjbIvOxUSf4Ghgnh8EwdNF2LDokQXrQh/WI explicit debug1: Authentications that can continue: publickey,password,keyboard-interactive debug1: Next authentication method: keyboard-interactive 2024-04-20 15:04:35 witcher: is lvm started before the mount is attempted? 2024-04-20 15:18:22 androidui: which ssh server is running ? 2024-04-20 15:18:33 sshd 2024-04-20 15:18:36 openssh 2024-04-20 15:19:00 if i set a password in the key then ssh asks for the key pashphrase 2024-04-20 15:19:08 if i set an empty password in the key then ssh asks for the username password instead 2024-04-20 15:19:37 if you type the key passphrase, does it then login, or asks for the user password instead? 2024-04-20 15:19:46 it logs in 2024-04-20 15:19:57 interesting 2024-04-20 15:22:52 im using ed25519 2024-04-20 15:46:29 I've made a complete mess of my dnsmasq config LOL 2024-04-20 15:48:45 welp i needed to use ssh-agent to auto-input it 2024-04-20 15:49:33 I'm pretty confused on what alpine is supposed to be like. Is it trying to stay close to upstream while fixing CVEs and making software work on musl with downstream changes? Or does it also patch things which could be considered unintended behavior? 2024-04-20 15:51:52 lassebq: we try to avoid customizing packages 2024-04-20 15:52:43 Good. So generally, if the software contains unintended behavior you're waiting for upstream to fix? 2024-04-20 15:53:04 And what about cases where software doesn't patch some bug? 2024-04-20 15:53:09 for a long time 2024-04-20 16:42:46 I am looking at options for DNS over TLS from my Alpine firewall (dnsmasq). Does the DNSCrypt-Proxy replace my dnsmasq DNS server, or would I need to run both? My goal is to run DNS over tls when forwarding requests to my controld setup. 2024-04-20 16:51:57 I can't speak to dnsmasq, but adguard-home has built-in support for DoH/DoT/DNSCrypt 2024-04-20 17:02:05 iggy: this is while decrypting the disk with LUKS, before LVM is relevant 2024-04-20 18:28:09 is abseil now a runtime component dep for chromium? 2024-04-20 18:29:44 abseil-cpp-debugging-internal and others pulled in for chromium update 2024-04-20 21:46:40 god dammit i just accidentally deleted my folder ;-; 2024-04-21 00:05:04 time for trash-cli then 2024-04-21 01:01:30 looks like a bunch of abseil stuff is depended on via re2 for chromium now, yeah 2024-04-21 04:12:28 Howdy all. Just need a quick hand. I've installed `mailx` ==( # apk add mail )== and well, no mailx appears to exist on the system.. 2024-04-21 04:12:55 nothing in /usr/bin, `which mailx` command returns with nothing.. 2024-04-21 04:13:40 apk apparently pulled in mailx-8.1.2_git20220412-r1 and `apk info` confirms it's installed 2024-04-21 04:14:31 also, typo in my first line above... I did type mailx (not just mail) 2024-04-21 04:15:19 I think mailx gets installed as sendmail? 2024-04-21 04:15:30 "apk info -L mailx" says it's usr/bin/mail 2024-04-21 04:16:36 raven523: thanks. I was under the impression `mail` comes as standard 2024-04-21 04:17:02 I guess I'll uninstall it and see if `mail` dissapears to confirm 2024-04-21 04:17:11 -L ... I knew there was a flag to show contents, just couldn't find it fast enough 2024-04-21 04:18:01 cool, that was esily solved. thanks 2024-04-21 04:19:21 I haven't had internet service I could actually send mail from in like a decade... I only vaguely remembered what mailx was 2024-04-21 04:19:41 (way to make me feel old) 2024-04-21 04:21:47 I just set up an smtp server last week because I hate myself 2024-04-21 04:24:15 haha I have been using heirloom-mailx on my NetBSD machine for quite some time and made a hand wrapper script to make using it a little more interactive 2024-04-21 04:25:18 Unfortunately it seems some editing is required for it to work on Alpine but it's much nicer than having to open a browser or even use mutt.... 2024-04-21 04:25:43 I suppose I should just use the alpine mail client while on Alipne... 2024-04-21 04:26:18 I got tired of not being able to send email from cron jobs/etc, so I wrote a sendmail replacement that just dumps messages into slack 2024-04-21 04:27:56 I reckon if I had a powerful/modern machine Id likely just use a browser or a GUI client. But I working with a intel core2 with 2gb ram so 2024-04-21 04:29:21 Slack is not doing well in chromium 124, hm 2024-04-21 04:32:46 switching channels seems to hang its renderer 2024-04-21 07:14:52 ok i managed to recreate my folder and qemu/ssh script files :) 2024-04-21 08:38:32 Hi! What's your experience with self-updating alpine machines (similar to debian's unattended-upgrade packages/daemons)? Does that break sooner or later? 2024-04-21 08:38:32 I'm about to deploy a new git server running nginx, gitea, postgres and some ruby scripts (asciidoctor!). Currently I use Gentoo w/ my own binpkg host, but I'd prefer not to be liable for the binpkgs of that new host, too. 2024-04-21 08:39:19 (Gentoo runs fine, but not really unattended, and my next approach would be Arch linux, which breaks no-matter-what every week :-P) 2024-04-21 08:43:04 holgersson: I run auto updates on 'edge' and 99% of the time I don't have problems, so running an stable version should be very stable. The only problem is taking care if you need to restart some service or reboot for kernel upgrades. 2024-04-21 08:47:56 donoban: Sounds promising, thanks. I'll test that! 2024-04-21 08:47:56 Actually I abuse a minimal debian host and let the system run inside a systemd-nspawn. As it's an office-only machine I'll shut the container down every night, copy it at file level and restart. That way I've easy-to-restore backups (that I can shove off the host, ofc), do not need to care about deamon restarts nor kernel upgrades. 2024-04-21 08:47:56 The debian host is in unattended-upgrades mode and every half year or so I do reboot that one, too. 2024-04-21 11:15:59 Is anyone able to use mangohud? 2024-04-21 11:18:00 It crashes with an assertion failure for me 2024-04-21 12:22:55 known issue: https://gitlab.alpinelinux.org/alpine/aports/-/issues/15933 2024-04-21 12:23:36 #15933 2024-04-21 12:29:51 Yeah I am the reporter lol 2024-04-21 12:31:00 I was wondering why it's only an issue on alpine. I don't know much about c++, but is it related to standard lib implementation? 2024-04-21 12:31:29 I researched this issue a little more and reported upstream https://github.com/flightlessmango/MangoHud/issues/1304 2024-04-21 12:31:55 Apparently it's related to a line being parsed without being checked for being empty? 2024-04-21 12:41:32 naive question: do c++ programm actually link against libcs (may it be glibc or musl)? 2024-04-21 12:43:38 they do 2024-04-21 12:58:32 Okay, there's almost certainly a problem with packaging mangohud 2024-04-21 12:58:58 there's \$LIB in /usr/bin/mangohud 2024-04-21 13:56:09 @ikke ty 2024-04-21 16:49:11 iggy, thanks for the suggestion on addguard-home. I am looking for something more broad to use on the alpine fw. It looks like the DNSCrypt-proxy package is no longer available.? 2024-04-21 16:52:03 Nevermind.... somehow I missed enabling the community repo 2024-04-21 18:29:56 alpine edge has issues with updating python 3 for 2 days now, many conflicts 2024-04-21 18:30:39 voidastro[m]: make sure you don't have any orphaned packages installed that keep python on 3.11 2024-04-21 18:30:44 ^, `apk list -O` 2024-04-21 18:34:13 doesnt appear to be the case. only some fonts and av1 libs, grimshot (sway screneshot util) 2024-04-21 18:35:17 https://pastebin.com/raw/4HsshN3p 2024-04-21 18:37:37 are you on riscv ? 2024-04-21 21:04:05 for some reason, 'docs' didn't pull in wireguard-tools-doc for me. 2024-04-21 21:44:10 invoked: do you have wireguard-tools installed (and not just wireguard-tools-wg and/or wireguard-tools-wg-quick) 2024-04-21 22:13:59 I can confirm apk repo (at least on princeton.edu) is currently a bit inconsistent 2024-04-21 22:14:04 but only package I had to uninstall was boost-dev 2024-04-21 22:15:52 apparently lots of stuff still depends on boost 1.82 but boost-dev metapackage is 1.84 (but I know that if I needed it urgently, I could add boost1.82-dev) 2024-04-21 22:35:59 dwfreed: good catch. i didn't pay attention and thought wireguard-tools-{wg,wg-quick} would pull everything in. it's funny because if you run wg-quick without args, it tells you to refer to wg-quick(8) 2024-04-21 22:36:18 heh 2024-04-22 01:11:42 Hello 2024-04-22 01:14:30 I am not aware of how significant the consequences might be, but I hope you can lift the blacklist restrictions on certain names. I won't do it a third time. 2024-04-22 01:44:18 hello 2024-04-22 01:54:59 hi 2024-04-22 02:00:42 "are you on riscv ?" <- x86 64 2024-04-22 02:01:26 is your apkindex up to date? I see several revisions in your list that are not latest 2024-04-22 02:02:38 apk update; apk upgrade 2024-04-22 02:03:23 what mirror are you pulling from? 2024-04-22 02:04:29 https://dl-cdn.alpinelinux.org/alpine/edge/main https://dl-cdn.alpinelinux.org/alpine/edge/community @testing https://dl-cdn.alpinelinux.org/alpine/edge/testing 2024-04-22 02:06:43 out of curiosity, if you `host dl-cdn.alpinelinux.org`, what do you get back? wondering what mirror you have 2024-04-22 02:07:58 elly: dl-cdn.alpinelinux.org is an alias for dualstack.j.sni.global.fastly.net 2024-04-22 02:08:17 huh, ok, same one for me 2024-04-22 02:09:10 i update every day. if there was briefly something broken which got fixed i may have gotten into an undefined state perhaps 2024-04-22 02:09:58 OK: 10355 MiB in 1830 packages 2024-04-22 12:42:11 Hi! 2024-04-22 12:48:33 Can anyone reply for testing purposes? 2024-04-22 12:48:58 gothic-riverboat, hi! test succeeded 2024-04-22 12:50:08 Thanks, Habbie! 2024-04-22 13:06:01 I have a problem setting up anacron. After installing cronie (since it includes anacron and I am not aware of any other way of installing it from the default repos) commands from `/etc/anacron` do not run. Everything works properly when running `anacron -fnd`. Which way to troublesoot? Is there any additional steps when setting up cronie? Cronie service is running btw. 2024-04-22 13:06:20 `/etc/anacrontab`* 2024-04-22 13:24:32 you could maybe start with the cronie service file 2024-04-22 13:29:34 lopid, It runs `crond -n` in the background. Crond manual says that crond should check anacrontab file. 2024-04-22 13:32:57 it also accepts -x … for debugging 2024-04-22 14:07:30 lopid, yes it is! Right now I am trying `sch,proc,misc,load,pars` and can see that crond runs `root` crontab and loading stuff from `/etc/periodic`. Since I already checked, there are no anacron-related tasks in there. Should there be one? 2024-04-22 14:08:11 Did anybody call me? `;-) 2024-04-22 14:09:09 tallies with https://pkgs.alpinelinux.org/contents?branch=v3.19&name=cronie&arch=x86_64&repo=community 2024-04-22 14:19:38 lopid, does it? I can see no /etc/periodic in there. Anyway, does it mean that anacron does not work by design, and I should add it? 2024-04-22 14:22:22 Because I can see `0anacron` file in `cron.hourly` here: https://docs.rockylinux.org/guides/automation/anacron/#anacron-configuration-file 2024-04-22 14:22:47 /etc/periodic is from the default busybox cron. did you indeed replace it with cronie? 2024-04-22 14:23:45 I installed cronie alongside busybox, disabled default cron service and enabled cronie service. 2024-04-22 14:24:50 i am not much help since i'm using only busybox cron 2024-04-22 14:25:50 Well, in that case I will add anacron task to 15min periodic tasks and see what will happen :) 2024-04-22 14:26:36 does rc-status show cronie or crond? 2024-04-22 14:28:15 Shows cronie. 2024-04-22 14:48:41 Yep it absolutely works. Still have no idea why it is not enabled by default tho. 2024-04-22 15:50:30 I just realized I have over 40GB of stuff in /var/tmp/abuild.* 2024-04-22 15:50:38 is this just from times I ran `abuild -K` or what? 2024-04-22 16:00:53 ah, looks like it is 2024-04-22 16:26:58 where do sysctl parameters get injected from? sysctl -a shows them all, but /etc/sysctl.conf and /etc/sysctl.d/ have no entries... 2024-04-22 16:28:47 they'll be built into the kernel already 2024-04-22 16:29:09 built a kernel w/o ipv6 and getting warnings at boot regarding invalid ipv6.net.xxxxx sysctl entries. 2024-04-22 16:30:13 net.ipv6? 2024-04-22 16:30:15 i see. so abuild adds in a default list, regardless of whether the sysctl kernel setting applies to the kernel actually being built? 2024-04-22 16:30:49 i don't know about abuild, but the linux kernel has its defaults 2024-04-22 16:35:11 this is just odd. disabling ipv6 in the kernel should remove all net.ipv6 (corrected) entries, but looks as if some remain. no big deal. thanks. 2024-04-22 16:36:00 i wonder if it's procfs making them available regardless 2024-04-22 17:37:32 hey, quick question: if i'm setting up fake raid with mdadm and wanna configure uefi, but the docs about it are a bit unclear (to me at least). do i just create one partition in linux raid and point setup-disks to that? or do i create two partitions (one for boot) and if so which one do i point setup-disks to? 2024-04-22 19:29:15 I'm missing some packages I think: libz, libssl and libcrypto: https://termbin.com/d2w7 But I have installed: libssl3 and openssl-dev and libcrypto. Not sure about libz, I have installed zlib-dev 2024-04-22 19:31:05 do you have the relevant -dev packages installed? 2024-04-22 19:31:31 Not sure which ones are relevant :{ 2024-04-22 19:32:57 What are you trying to do? 2024-04-22 19:33:35 Trying to build crystal lang app with release and static flag 2024-04-22 19:34:31 Works fine if I don't use --static flag. But would like to try with static so I can just transfer bin to server and try to run it there 2024-04-22 19:37:44 Shard that's complaining is https://github.com/kostya/lexbor https://termbin.com/zbi6 2024-04-22 19:41:48 You need the -static variants in that case 2024-04-22 19:41:52 zlib-static 2024-04-22 19:42:14 openssl-libs-static 2024-04-22 19:44:48 Aaa ok that... makes sense 2024-04-22 19:45:35 And it works. Thank you ikke! 2024-04-22 19:45:57 np 2024-04-22 19:46:21 Maybe I should use dockers for that too.. so I won't mess too much my laptop with all the dev and static pkgs 2024-04-22 21:35:18 Hey to all,... (full message at ) 2024-04-22 23:51:01 for some reason this fixed the python3 update issue i was having: apk del boost boost-dev libtorrent-rasterbar-dev boost-build 2024-04-22 23:52:10 apk add boost now generates conflicts however 2024-04-22 23:58:36 https://pastebin.com/raw/BQYQVgss 2024-04-23 02:45:19 Hello, Is there anyone who could share with me a working x86 install of alpine with a working desktop gui please? VM is fine 2024-04-23 03:11:00 plebotomisty[m]: what do you mean? it's not difficult to install 2024-04-23 10:41:06 Something is strange with libreoffice 2024-04-23 10:41:12 john-1.9.0-r7 2024-04-23 10:41:12 % apk search cmd:libreoffice 2024-04-23 10:41:36 And there is no libreoffice command anymore from any of the libreoffice packages 2024-04-23 10:42:13 #16005 2024-04-23 15:07:07 anyone try the fortify-headers update from https://github.com/jvoisin/fortify-headers ? is there a blocker whereby alpine hasnt moved to it? 2024-04-23 15:11:02 probably nobody has rolled it? 2024-04-23 15:12:25 https://dl.2f30.org/releases/ (the source we use) seems to have a 2.1 and we're using 1.1 2024-04-23 15:12:36 you or someone else could send an MR for main/fortify-headers to update it 2024-04-23 15:14:10 also noted the default abuild flags in /usr/share/abuild/default.conf but no -fstack-protector-strong explicity declated. is musl patched to automatically apply ssp without having abuild declare -fstack-protector-strong in any flags? 2024-04-23 15:22:36 will roll and test fortify-headers 2.x locally and go from there 2024-04-23 15:22:40 good luck! 2024-04-23 15:22:45 `abuild bump` is the tech :) 2024-04-23 16:49:11 anyone build chromium with cfi true? any idea how long it would take with 64 gb system ram? 2-4 hrs? longer? just wanted to make some time for it... 2024-04-23 16:50:13 does it even build on alpine with cfi (default is off)? 2024-04-23 16:50:49 if you have 64 GB of RAM and lots of cores (like, 64 or more) then it'll take a couple of hours, but I would be surprised if CFI works 2024-04-23 16:51:05 chromium only barely works in the config the APKBUILD uses, any other config isn't likely to work 2024-04-23 16:51:08 but you could try it! 2024-04-23 16:57:41 yeah, chromium is very fragile overall 2024-04-23 16:57:52 if you look at it wrong, it can fail to build 2024-04-23 16:58:15 will give it a go. fyi fortify-headers 2.2 seems to work fine with gcc/x86_64/edge. no idea with other arches. more warnings when building but seems fine. will continue testing... 2024-04-23 17:00:33 going back to ssp: does the toolchain auto apply fstack-protector-strong like gentoo? hence, no need to declare it in the default abuild cflags/cxxflags? 2024-04-23 18:55:45 what happened to missing tree-sitter grammars? eg diff/meson/yaml? 2024-04-23 18:57:15 build issues or just not contributed yet? would have thought they would all be available... 2024-04-23 19:31:42 nvm. helix on Alpine is using upstream tree-sitter grammars of which there is no diff/meson/etc... 2024-04-23 23:45:38 im trying to install alpine on a secondary partition and dualboot with windows 2024-04-23 23:45:56 and i cant figure out how to install grub properly 2024-04-23 23:46:58 i mounted the esp with mount /dev/sda1 /boot/efi and the system partition with mount /dev/sda4 /mnt, then ran setup-disk -m sys /mnt and get errors 2024-04-23 23:47:09 i tried following what the wiki says but it doesnt work 2024-04-24 03:02:17 Hello world. I just got a wacom tablet. It shows in dmesg but doesn't map to /dev/input/wacom* or show in sway inputs. Anyone here using a wacom tablet on alpine? 2024-04-24 03:07:37 I have a wacom tablet, and it works with sway, but I didn't test it with alpine specifically 2024-04-24 03:14:51 Ermine: Good start. So, it seems there isn't a wacom driver package. 2024-04-24 03:15:37 sway doesn't need any special packages iirc 2024-04-24 03:15:54 https://pkgs.alpinelinux.org/packages?name=*wacom*&branch=edge&repo=&arch=x86_64&maintainer= 2024-04-24 03:16:19 https://github.com/linuxwacom/input-wacom/wiki/Installing-input-wacom-from-source 2024-04-24 03:17:43 Ermine: lsmod | grep -i wacom ? 2024-04-24 03:18:05 I don't have the wacom.ko 2024-04-24 03:42:35 Correction. Wacom tablet works perfectly, when you have the right kernel modules for the kernel to load. ;-) 2024-04-24 03:42:38 Just had to reboot. 2024-04-24 03:51:20 So... is there any way to tell apk to just, not delete the current kernel's modules dir besides just manually backing it up? 2024-04-24 04:26:17 Sofia: not really no 2024-04-24 05:14:16 Huh. Just searched for something involving flatpak and stumbled onto this. Alpine's flatpak is out of date apparently. https://linuxiac.com/flatpak-patch-addresses-major-sandbox-escape-flaw/ 2024-04-24 05:15:00 https://github.com/flatpak/flatpak/security/advisories/GHSA-phv6-cpc2-2fgj 2024-04-24 09:24:33 es most megy a nagy panik 2024-04-24 09:24:36 whoops 2024-04-24 09:24:45 wrong chat 2024-04-24 09:45:59 hello, on a new edge VM I have this apk error: ERROR: kactivities5-libs-5.115.0-r1: BAD signature 2024-04-24 09:46:26 are you using dl-cdn.a.o as your apk mirror? 2024-04-24 09:49:20 hum, I was using my own private cache, seems better now I changed to official mirror 2024-04-24 09:51:30 also I don't seem to be able to install linux-edge 2024-04-24 09:52:05 of course I added community repo 2024-04-24 10:49:22 IDENTIFY ahmadraniri Pensivers1994! 2024-04-24 10:49:40 oops sorry.... 2024-04-24 10:50:16 gotta change your password now 2024-04-24 10:50:36 and also that's quite a poor password 2024-04-24 10:50:48 i advise you to use a password manager that generates passwords for you 2024-04-24 10:59:30 why, what's yours? 2024-04-24 11:00:11 hunter2 2024-04-24 11:02:16 triallax (f378378b91@000306d2.user.oftc.net) has joined #alpine-linux 2024-04-24 11:02:35 suspicious indeed 2024-04-24 11:02:45 O:) 2024-04-24 15:01:50 IDENTIFY nanon ********* 2024-04-24 15:02:16 oops, good thing IRC hides passwords for you! 2024-04-24 15:03:08 ok, hunter2 2024-04-24 15:03:36 I'm glad people who use IRC are also old enough to know the jokes :P 2024-04-24 15:04:02 i'm not that old 2024-04-24 15:04:05 the joke is fairly well-known 2024-04-24 15:05:27 we keep spreading the meme 2024-04-24 15:06:50 nanon: lol 2024-04-24 15:09:23 #bitlbee has a bot that goes "welcome to the PUBLIC bitlbee channel" if you try to identify 2024-04-24 15:10:10 nice 2024-04-24 15:10:37 I loved bitlbee when facebook messenger still had xmpp 2024-04-24 16:11:34 does shadow-4.15.1-r0 support changing the pass has in pam.d/base-password from sha512 to yescrypt? 2024-04-24 16:12:52 with shadow-login, dont want to rehash and not be able to log back into the system at next sys restart... 2024-04-24 16:13:26 anyone replace sha512 with yescrypt? 2024-04-24 16:45:05 straydog: the `yescrypt` argument to the pam_unix.so for password changes will change which algorithm is used when changing a password. The /etc/shadow- has indications as to which hash algo was used, so it shouldn't be a problem if you have different hash algos in use. `$6$` indicates sha512 and `$y$` indicates yescrypt. 2024-04-24 16:46:19 as always, try it in a test system if you're not sure. Even after years of working with systems, I still constantly run little tests to validate nuance or details in behavior. 2024-04-24 16:46:33 especially with components I only deal with every once in a while (like pam/shadow) 2024-04-24 17:13:10 changed sha512 to yescrypt > ran passwd: memory buffer error. rebooted > ran passwd: memory buffer error. 2024-04-24 17:14:12 switched back to sha512 > ran passwd: password rehashed $6$ with sha512. will look into this... 2024-04-24 18:50:27 anyone using a thinkpad p14s gen2? having some suspend resumption issues , not sure if i need a specific driver or something 2024-04-24 18:51:08 these suspend issues didnt happen on arch fwiw 2024-04-24 18:54:02 how do i write to a socket from a shell script? 2024-04-24 18:54:38 socat 2024-04-25 03:09:13 anyone having issue building rust apps under rootbld? was it the latest abuild update that broke things? or is it likely something on my end? using vanilla apkbuilds for fd/rg/oxipng/helix/etc... 2024-04-25 03:09:32 "note: some `extern` functions couldn't be found; some native libraries may need to be installed..." 2024-04-25 03:10:43 c/c++/go apps are fine (the ones i tested), just rust apps (checksec.rs builds fine though)... 2024-04-25 03:11:28 nothing on my build box has changed for the past few days save for 1-2 abuild updates... 2024-04-25 03:14:55 at first i thought abuild rootbld was pulling in a wrong dependency from REPODEST but nope 2024-04-25 03:16:01 was able to build fd/ripgrep/etc just a few days ago...odd... 2024-04-25 03:26:55 everything builds fine until link time it seems. using the default toolchain w/ default flags. everything save for REPODEST and PACKAGER_PRIVKEY is default... 2024-04-25 06:38:19 straydog: 'abuild rootbld' on community/fd works for me; last updated my system (on edge) about a week ago. 2024-04-25 06:49:11 straydog left 2024-04-25 06:49:26 housecat right 2024-04-25 06:54:45 :) 2024-04-25 08:27:21 Hm. I've tried to pass interpreter=i to build crystal section of APKBUILD: https://git.alpinelinux.org/aports/tree/community/crystal/APKBUILD?h=3.19-stable ran abuild checksum and abuild -r but I got a bunch of py3 errors: https://termbin.com/aiwp I don't see where python package is required from APKBUILD ? 2024-04-25 08:52:58 Seems like you still have things installed that deoend on python 3.11? 2024-04-25 08:53:50 yeah. whynothugo: think its the most recent abuild built on 4/21 that may have broken something. else, i am running into this: https://github.com/helix-editor/helix/discussions/9558 2024-04-25 08:54:23 may just install rustup and give it a go... 2024-04-25 08:55:22 straydog: i could build something yesterday 2024-04-25 08:55:27 ripgrep 2024-04-25 08:55:28 any plans to update the repo rust to say 1.77.2? is rust only updated every major version 2024-04-25 08:55:32 Without issue 2024-04-25 08:56:05 strange. must be on my end. abuild-3.13.0-r3 ? 2024-04-25 08:57:15 I did a system upgrade before building, so should be katest available version 2024-04-25 08:58:03 ikke: python --version -> Python 3.12.3 2024-04-25 09:11:22 ikke: hm but llvm16-test-utils-16.0.6-r8 is'nt installed? There is no sign of it in /etc/apk/world 2024-04-25 09:23:24 Hm while read p; do apk info -R "$p"; done < /etc/apk/world |grep python shows nothing 2024-04-25 09:29:11 Woops 2024-04-25 09:29:33 apk list -I shows much more stuff than /etc/apk/world 2024-04-25 09:31:27 world only shows explicitly installed packages 2024-04-25 09:31:32 Not dependencies 2024-04-25 09:31:56 apk list --orphaned 2024-04-25 09:32:17 Thank you 2024-04-25 10:39:13 dayum...rustup is able to build everything rust manually, so its just my abuild or rootbuild env. dunno what it going on. loosing too much sleep... 2024-04-25 10:40:57 going to change/unset some flags..see if that helps... 2024-04-25 11:18:20 resolved. rust apps build again under rootbld. it was my env. 2024-04-25 12:12:21 Hi 2024-04-25 12:12:34 I installed alpine on proxmox and I'm having issues with installing community packages 2024-04-25 12:13:02 I enabled the community repos in the /etc/apk/repositories file and updated the package list 2024-04-25 12:13:06 but it doesn't seem to be working 2024-04-25 12:13:44 Any community package I try to install returns "Unable to select packages, no such package" 2024-04-25 12:14:10 what is the output of apk update? 2024-04-25 12:14:48 OK: 22989 distinct packages available 2024-04-25 12:15:28 It should also show you lines starting with fetch: 2024-04-25 12:15:37 ´yeah it does that too 2024-04-25 12:15:57 there's 4 mirrors 2024-04-25 12:16:28 Does it show a line with something like edge/community/x86_64/APKINDEX.tar.gz? 2024-04-25 12:17:12 fetch http://alpine.mirror.far.fi/latest-stable/main/x86_64/APKINDEX.tar.gz and fetch http://alpine.mirror.far.fi/latest-stable/community/x86-_64APKINDEX.tar.gz 2024-04-25 12:17:31 and then 2024-04-25 12:18:10 What package are you trying to install? 2024-04-25 12:18:25 greenbone-security-assistant 2024-04-25 12:18:32 and other openvas things 2024-04-25 12:19:09 There is no such package 2024-04-25 12:19:14 wellll ummm 2024-04-25 12:19:42 ohhh 2024-04-25 12:19:48 the last ones vwere on v3.15 2024-04-25 12:20:07 do I need to install an older version of alpine for that? 2024-04-25 12:20:25 opr can I just fetch older package lists and install em 2024-04-25 12:20:25 yes, you would need to install 3.15, but that one is no longer supported 2024-04-25 12:20:28 No 2024-04-25 12:20:31 rip 2024-04-25 12:20:31 Very unlikely 2024-04-25 12:20:57 welp Ig I'm installing another OS 2024-04-25 12:20:59 thx 2024-04-25 16:01:21 Hi alpine community, i'm having trouble finding some guidance on how to create wlan AP interface on rpi zero running alpine. I see /sbin/iw command on rasbian but cant find what can be used on alpine armhf 2024-04-25 16:09:48 badday: iw is a package that you can install 2024-04-25 16:11:47 thanks i will try iw package, just wanted to skip it if something already exists which achieves the same thing 2024-04-25 16:19:08 if you want any kind of security, you're going to need hostapd 2024-04-25 16:38:49 https://wiki.alpinelinux.org/wiki/How_to_setup_a_wireless_access_point 2024-04-26 13:50:45 mandoc and man-pages does not bring up the man pages for me. what can be the reason? 2024-04-26 13:51:25 are colours compiled into busybox ls? I cant find a .profile alias which says that directory listings should be colored. 2024-04-26 13:54:38 apk add mandoc 2024-04-26 13:54:58 i doubt busybox supports colours 2024-04-26 13:56:08 lopid: that is what i did: "apk add mandoc" and "apk add man-pages". still manpages are not there 2024-04-26 13:56:38 are any installed? 2024-04-26 13:56:52 lopid: i also doubt busybox supports colors. this is the question, from the the colored directory listing come from if ls is from busybox 2024-04-26 13:57:10 lopid: i suppose man-pages are the pages. i had also man-db installed 2024-04-26 13:57:59 a package's man pages are often in separate -doc package 2024-04-26 13:58:15 how come "apk add irssi" installs qtwebengine as a dependency at first? 2024-04-26 13:58:46 lopid: i see, is there no way to install the man pages of installed packages/programs all at once? 2024-04-26 13:59:13 for ones installed before mandoc? there will be a small scriptable way, at least 2024-04-26 14:00:22 very rudimentary: while read -r p; do apk add "$p"-doc; done < /etc/apk/world 2024-04-26 14:00:40 with some false negatives 2024-04-26 14:01:25 apk add docs 2024-04-26 14:01:35 then you will get them automatically without having to install them explicitly 2024-04-26 14:01:46 nice 2024-04-26 14:06:48 lopid: ikke: thankss 2024-04-26 14:07:48 then where do the colors in "ls" come from? usually they have to be called for with "ls --color"... 2024-04-26 14:08:31 which ls? 2024-04-26 14:08:54 the command ls, which supposedly is part of busybox 2024-04-26 14:09:12 you may have install coreutils' ls 2024-04-26 14:09:16 ed 2024-04-26 14:10:50 right, thanks. still, how do colors get there? in /etc/profile etc there is nothing 2024-04-26 14:11:09 ls outputs them 2024-04-26 14:11:55 only if it is asked for, with the option --color, which i cant find 2024-04-26 14:12:07 which version are you using? 2024-04-26 14:14:06 alias 'ls --color=auto' 2024-04-26 14:15:29 ikke: yes, exactly, this is what i have not done, still the colors are there, and i wonder how? 2024-04-26 14:15:57 as there is no such alias in any profile 2024-04-26 14:16:01 busybox will automatically do it, coreutils ls not 2024-04-26 14:16:15 ikke: ok, interesting, thanks 2024-04-26 14:17:15 ikke, do you by chance have an idea why "apk add irssi" installed qtwebengine as a dependency before installing irssi? 2024-04-26 14:18:49 It does not do so for me 2024-04-26 14:20:20 what happens if you try `apk del qt5-qtwebengine` (or qt6-qtwebngine) 2024-04-26 14:30:53 apk info -R irssi 2024-04-26 14:31:54 lopid: ah, the -R option i was looking for, thanks! 2024-04-26 14:32:30 It doesn't show recursive dependencies though 2024-04-26 16:30:07 Hm. I start chromium like this: `chromium --enable-features=UseOzonePlatform --ozone-platform=wayland` to actually see GUI in sway but how can I ommit this or add somewhere in the config of browser 2024-04-26 16:40:37 try Preferred Ozone platform in chrome://flags ? 2024-04-26 17:42:47 alternatively, something with a .desktop file? i recall doing that for Firefox+wayland (but that was on Debian) 2024-04-26 18:08:56 ikke: thanks for "apk add docs". could be corrected in the FAQ of the wiki. installing irssi required qtwebengine because a previous installation error. what is still unclear: why busybox-doc does not give the man pages for ls etc, while coreutils-doc does. 2024-04-26 18:09:50 patanga: busybox does not have man-pages 2024-04-26 18:12:18 yes, i see. seems the package busybox-doc is empty or refers just to busybox itself .... 2024-04-26 18:17:37 had the idea to add this, just today 2024-04-26 20:37:09 irssi, qtwebengine... what kind of dependency is that?! when I was this years old, irssi only required perl... 2024-04-26 20:40:42 satmd: I think they are unrelated 2024-04-26 20:40:50 phew 2024-04-26 20:41:25 I was worried for a moment 2024-04-27 00:18:50 I'd rather dep on qt than perl these days 2024-04-27 00:18:53 ACTION runs 2024-04-27 07:07:25 yo 2024-04-27 07:07:31 whoops, wrong chat 2024-04-27 08:13:50 anjan: it's ok, yo back 2024-04-27 11:52:09 Hello, i have installed GNOME on my Alpine Linux install, and when i try to open GNOME Software, it gives me the following error: `The name dev.Cogitri.apkPolkit2 was not provided by any .service files`. I have enabled the apk-polkit-server service and started it, and my system is up-to-date. 2024-04-27 12:03:50 can you run the apk-polkit-server manually and check if it doesn't crash? 2024-04-27 12:03:58 alright 2024-04-27 12:04:21 like the command or the service 2024-04-27 12:05:27 like `doas /usr/libexec/apk-polkit/apk-polkit-server` in the terminal 2024-04-27 12:05:33 k 2024-04-27 12:06:07 it segfaults 2024-04-27 12:06:28 its the same thing that happened in a postmarketOS issue 2 weeks ago 2024-04-27 12:06:58 what 2024-04-27 12:07:28 it triggers on issue\ 2 2024-04-27 12:07:37 oh 2024-04-27 12:07:40 https://gitlab.com/postmarketOS/pmaports/-/issues/2739 2024-04-27 12:09:34 Hello. Is there anyone who uses GDM and is able to run xorg sessions with it? 2024-04-27 12:09:51 On Alpine 2024-04-27 12:09:58 gdm should already be able to run x.org sessions 2024-04-27 12:10:07 Well, it doesn't 2024-04-27 12:10:11 I get tty permission error 2024-04-27 12:10:18 and have to chown it prior to runnin session 2024-04-27 12:10:39 have you ran `setup-xorg-base`? 2024-04-27 12:10:40 Does not happen with wayland sessions for some odd reason 2024-04-27 12:10:55 Probably not 2024-04-27 12:11:17 then go for it 2024-04-27 12:12:38 That did absolutely nothing 2024-04-27 12:12:44 It ran some apk commands but I already had necessary packages 2024-04-27 12:13:09 is this an old install or you just installed alpine 2024-04-27 12:13:40 Less than a month old 2024-04-27 12:13:43 edge branch 2024-04-27 12:13:51 hmm 2024-04-27 12:13:59 lassebq: is your user in the correct groups? do you have udev properly ( 4 services in `rc-update` ) set up? 2024-04-27 12:14:16 s/4/3/; udev, udev-postmount, udev-trigger 2024-04-27 12:15:22 What does s/4/3/ mean? 2024-04-27 12:15:28 runlevels? 2024-04-27 12:15:37 oh, it's just sed syntax because i made a mistake in the previous message 2024-04-27 12:15:40 it means "replace 4 with 3" 2024-04-27 12:15:45 oh 2024-04-27 12:15:58 Well services are there 2024-04-27 12:16:14 groups are users tty lp wheel audio input video plugdev seat abuild 2024-04-27 12:16:43 hm. that looks correct 2024-04-27 12:16:54 how are you running gdm? 2024-04-27 12:17:18 gdm service 2024-04-27 12:17:35 For reference, lightdm is able to run sessions, but that's purely because xorg runs as root 2024-04-27 12:17:55 lxc launch images:alpine/edge a2 ... failed instance creation image could not be found 2024-04-27 12:18:02 gdm runs it as user and for whatever reason that requires me to manually chown a tty 2024-04-27 12:18:43 i dont see why you need to do this 2024-04-27 12:18:46 i installed the gnome package 2024-04-27 12:18:51 which does include gdm 2024-04-27 12:19:00 added gdm, polkit and dbus to the default runlevel 2024-04-27 12:19:07 added my user to the input group 2024-04-27 12:19:20 (installed dbus and polkit too) 2024-04-27 12:19:27 ran setup-xorg-base as well 2024-04-27 12:19:32 reboot and it works 2024-04-27 12:19:36 with xorg sessions as well 2024-04-27 12:19:42 well 2024-04-27 12:19:44 Gnome on Xorg 2024-04-27 12:19:48 Damn, I guess I'm reinstalling then 2024-04-27 12:33:09 my issue should be fixed when alpine repositories re-build apk-polkit-rs 2024-04-27 12:33:57 a fix has already been commited to the repository 2024-04-27 12:35:45 should have already been built, then 2024-04-27 12:35:49 please guys 2024-04-27 12:35:54 no, 2024-04-27 12:35:56 alpine does not show 2024-04-27 12:36:01 give me a moment to send a screenshot 2024-04-27 12:36:02 in lxc image list 2024-04-27 12:37:33 pxrb: I suppose you mean that the pkgrel needs to be bumped for apk-polkit-rs 2024-04-27 12:39:11 https://0x0.st/XH4K.png 2024-04-27 12:39:20 build date is 2023-04-17 2024-04-27 12:40:54 pxrb: what fix were you referring to that was already been pushed? 2024-04-27 12:41:36 https://gitlab.alpinelinux.org/Cogitri/apk-polkit-rs/-/merge_requests/48 2024-04-27 12:42:14 so waiting for a release 2024-04-27 12:42:23 (or backport) 2024-04-27 12:42:28 yes 2024-04-27 12:42:33 no, im on edge 2024-04-27 12:42:49 I mean backport from master to the current release 2024-04-27 12:42:57 of apk-polkit-rs 2024-04-27 12:43:06 oh 2024-04-27 12:43:16 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/64678 2024-04-27 12:43:20 but it fails to build 2024-04-27 12:44:32 the test failed 2024-04-27 12:44:35 not the compilation 2024-04-27 12:45:01 The package fails to build 2024-04-27 12:45:12 That includes passing tests 2024-04-27 12:45:53 ``` 2024-04-27 12:45:57 ERROR: Unable to open root: No such file or directory 2024-04-27 12:45:57 ERROR: 2 errors updating directory permissions 2024-04-27 12:45:57 Caused by: 2024-04-27 12:45:57 error: test failed, to rerun pass `-p apk-polkit-rs --bin apk-polkit-rs` 2024-04-27 12:45:57 process didn't exit successfully: `/builds/pabloyoyoista/aports/community/apk-polkit-rs/src/apk-polkit-rs-2.1.1/output/debug/deps/apk_polkit_rs-cf27705b9b8203a8` (signal: 11, SIGSEGV: invalid memory reference) 2024-04-27 12:45:59 + kill %1 2024-04-27 12:46:01 ``` 2024-04-27 12:46:10 was where the error happened 2024-04-27 12:46:16 Yes, I can see that in the logs 2024-04-27 12:46:34 the compiler isn't whats giving the error 2024-04-27 12:46:44 one of the tests is giving an error failing the build 2024-04-27 12:46:47 I'm not saying it's the compiler 2024-04-27 12:47:58 misunderstanding 2024-04-27 12:47:59 sorry 2024-04-27 13:02:18 Hummm, after an update I'm unable to mount my Alpine/zfs anymore 2024-04-27 13:02:31 At boot, init says: /init: line 373: zpool: not found 2024-04-27 13:03:30 quinq: check /etc/mkinitfs.conf if the zfs feature is still present 2024-04-27 13:03:39 Right, thanks :) 2024-04-27 13:03:40 /etc/mkinitfs/features.conf 2024-04-27 13:04:22 /etc/mkinitfs/mkinitfs.conf* 2024-04-27 13:04:50 It is :/ 2024-04-27 13:05:03 features="ata base keymap kms mmc scsi usb virtio nvme zfs zfshost" 2024-04-27 13:05:24 And I was able to chroot into the fs and re-generate the initrd already, without apparent error 2024-04-27 13:06:36 (with Linux 6.6.28-0-lts) 2024-04-27 13:06:55 (and mkinitfs 3.9.1-r0) 2024-04-27 13:07:46 oh 2024-04-27 13:07:50 sh: zpool: not found 2024-04-27 13:07:52 wth 2024-04-27 13:08:13 apk version zfs 2024-04-27 13:08:34 I have zfs-lts-6.6.28-r0 installed though 2024-04-27 13:08:44 that's just the kernel modules 2024-04-27 13:08:54 you need zfs for the userspace tools 2024-04-27 13:09:00 Then the utils are not installed anymore :/ 2024-04-27 13:09:51 Ah :D 2024-04-27 13:09:58 But I have zfs-lts-6.6.28-r0 installed 2024-04-27 13:10:05 wait 2024-04-27 13:10:10 going full circle ^^ 2024-04-27 13:10:16 cat /etc/mkinitfs/features.d/zfs.files in the croot 2024-04-27 13:10:18 chroot* 2024-04-27 13:10:28 No but zfs is definitely absent 2024-04-27 13:10:31 That's the problem 2024-04-27 13:20:30 And I can't boot from USB on that machine :/ 2024-04-27 13:20:59 pxrb: Just did a clean system install and wayland session doesn't start now 2024-04-27 13:21:07 xorg fails with the same tty permission error 2024-04-27 13:21:43 xf86OpenConsole: Cannot open virtual console 7 (Permission denied) 2024-04-27 13:22:11 Oh btw gdm works somehow 2024-04-27 13:22:16 anyone running alpine container ? 2024-04-27 13:22:23 lassebq: maybe you didnt install wayland 2024-04-27 13:22:25 as in, it's the only graphical env which works 2024-04-27 13:22:32 what are you trying to run 2024-04-27 13:22:34 or start 2024-04-27 13:22:37 I try to add alpine 3.17 image not found 2024-04-27 13:22:55 I just want to use desktop, no particular intention 2024-04-27 13:23:03 what DE/WM 2024-04-27 13:23:07 gnome 2024-04-27 13:23:15 thats extremely wierd 2024-04-27 13:23:19 what gpu do you use 2024-04-27 13:23:22 amd 2024-04-27 13:23:27 hmm 2024-04-27 13:23:34 I literally run gnome on another distro 2024-04-27 13:23:35 send me your .ash_history 2024-04-27 13:23:39 no problems there 2024-04-27 13:24:03 also, did you install alpine using setup-alpine? 2024-04-27 13:24:14 No, alpine-chroot-install 2024-04-27 13:25:21 send me your .ash_history 2024-04-27 13:25:45 on the alpine install 2024-04-27 13:26:13 using a pastebin 2024-04-27 13:28:39 https://paste.gnome.org/d3fM8MeET 2024-04-27 13:28:55 I get 403-ed by pastebin 2024-04-27 13:30:31 nothing seems out of the ordinary 2024-04-27 13:30:56 I know, alpine just seems to be broken 2024-04-27 13:41:09 everything is ok 2024-04-27 13:41:14 all packages are installed correctly 2024-04-27 13:41:18 i dont see why it doesnt work 2024-04-27 14:07:42 Finally fixed, thanks for the hint ikke :) 2024-04-27 14:33:06 Update, I back to the same system as before, for the most part 2024-04-27 14:33:19 Still, xorg cannot be started from gdm 2024-04-27 14:33:24 no matter how hard I tried 2024-04-27 14:34:05 lassebq, I haven't followed the conversation, but Xorg is already running if you can see a graphical session manager (here gdm) 2024-04-27 14:34:06 this has to be an issue with gdm interacting with elogind, or whatever manages VTs 2024-04-27 14:34:26 quinq. It doesn't, gdm runs on wayland by default 2024-04-27 14:34:47 But then you're runnin wayland, not xorg 2024-04-27 14:34:49 g 2024-04-27 14:35:09 No? DM is able to start both xorg and wayland sessions 2024-04-27 14:35:24 why do you need GNOME for Xorg 2024-04-27 14:35:25 It's supposed to start a separate xorg server if the DM itself is using wayland 2024-04-27 14:35:44 I doubt that can work 2024-04-27 14:35:54 It does 2024-04-27 14:35:59 on other distros 2024-04-27 14:42:34 Anyway, the only way to make GDM work with xorg sessions is to chown /dev/tty* to my user 2024-04-27 14:43:47 That's pretty stupid but it works. And for some reason wayland sessions don't need it, they chown /dev/tty themselves 2024-04-27 14:51:48 In arch it works with a tty group, ls -la /dev shows that tty* is root:tty 2024-04-27 14:51:55 weird thing is that im not in the tty group 2024-04-27 14:56:06 maybe it doesnt work with the tty group 2024-04-27 14:56:11 its just that the tty group owns it 2024-04-27 15:06:30 pxrb: even ones that have active sessions? 2024-04-27 15:06:32 are root:tty 2024-04-27 15:06:54 because on alpine gdm user owns /dev/tty1 and /dev/tty7 is owned by logged in user 2024-04-27 15:07:03 that is when running a wayland session 2024-04-27 15:07:20 yes 2024-04-27 15:07:36 when running xorg session /dev/tty7 stays owned by root, and xorg quits unexpectedly 2024-04-27 15:08:05 tty7 over here is still owned by root:tty 2024-04-27 15:08:18 but user isn't in tty group right? 2024-04-27 15:08:29 nope 2024-04-27 15:08:32 im not in the tty group 2024-04-27 15:08:34 even gdm isnt 2024-04-27 15:08:55 pxrb: note that arch runs graphical sessions somewhere within /dev/tty1 - /dev/tty3 2024-04-27 15:09:04 depends on which ones are free and which DM is used 2024-04-27 15:09:34 do they also have unchanged owner? 2024-04-27 15:09:36 every tty* file is owned by root:tty 2024-04-27 15:09:39 Ok 2024-04-27 15:09:44 That is weird then 2024-04-27 15:10:32 crw--w---- 1 gdm root 4, 1 Apr 27 18:10 /dev/tty1 2024-04-27 15:10:49 lol did that trigger the bot? 2024-04-27 15:11:01 send it again 2024-04-27 15:11:04 crw--w---- 1 gdm root 4, 1 Apr 27 18:10 /dev/tty1 2024-04-27 15:11:07 ok 2024-04-27 15:13:02 algitbot: \o/ 2024-04-27 15:14:00 wait 2024-04-27 15:14:05 tty2 is owned by user:tty 2024-04-27 15:16:07 Alright, then it matches alpine behavior, except for the fact that xorg sessions dont change owner? 2024-04-27 15:16:41 let me check on that 2024-04-27 15:18:15 Oh, lmao 2024-04-27 15:18:29 Went through irc logs and saw that highFreqSurfer has the same issue 2024-04-27 15:18:45 I am not alone in this, finally 2024-04-27 15:21:43 ye 2024-04-27 15:21:53 in X11 only tty2 is modified to user:tty 2024-04-27 15:21:58 as well 2024-04-27 15:43:56 Hello all. I am running Alpine in diskless mode from a flash drive, and would like to know how to configure the RAM split, which seems to be around 50/50 at the moment. Thank you. 2024-04-27 18:44:58 how do you folks do hard disk encryption on your laptops? 2024-04-27 18:54:46 luks with detached header on a bootable usb stick, so the laptop has no bootpartition or anything at all except one big cryptoluks device 2024-04-27 18:58:02 I thought about doing the detached header thing. Only really necessary if you want the plausible deniability. 2024-04-27 19:03:48 Hi 2024-04-27 19:04:49 oGEosFMtUw8, so you permanently need a stick to boot your machine? 2024-04-27 19:05:31 They would have to plug it in to boot, from my understanding. 2024-04-27 19:05:34 yes but only until intram pivots 2024-04-27 19:06:06 no, detached header allows you to keep your disk keys always under physical control 2024-04-27 19:06:39 Are you doing autologin? 2024-04-27 19:06:43 gi 2024-04-27 19:06:48 hji 2024-04-27 19:06:51 no, haha 2024-04-27 19:07:04 hurra 2024-04-27 19:07:06 Then I still dont understand the point of it 2024-04-27 19:07:16 am i visible 2024-04-27 19:07:22 Yes 2024-04-27 19:07:25 nice 2024-04-27 19:07:32 i have a question 2024-04-27 19:07:57 can i install brcm43142 driver which isnt supported by alpine and not in b43 2024-04-27 19:08:48 Oh my, detached LUKS header 2024-04-27 19:08:55 what 2024-04-27 19:09:16 if the header is attached any attacker having a copy of your raw disk can bruteforce your password (perhaps even knowing it beforehand via cctv e.g.) 2024-04-27 19:09:26 First time I hear about someone even mentioning it, as well as I hear of someone using it 2024-04-27 19:09:38 if there is no header there is no practical way of bruteforcing the keys 2024-04-27 19:09:41 Yes, I am aware. I been thinking about detaching headers 2024-04-27 19:09:58 However, I don't have a safe solution. I have had too many USBs die on me 2024-04-27 19:09:59 i'm not talking to you. 2024-04-27 19:10:08 Oh 2024-04-27 19:10:10 gamingirc was asking 2024-04-27 19:10:17 but you are welcome to chime in 2024-04-27 19:10:43 uhm 2024-04-27 19:10:49 What do you use for USB that you trust not to fail? 2024-04-27 19:11:08 bruteforceing a luks2 slot with the default argon2id set to 2 seconds would be very hard tho 2024-04-27 19:11:23 hwy 2024-04-27 19:11:25 hey 2024-04-27 19:11:42 babz: Today yes. 2024-04-27 19:11:45 again, what if your password is compromised through cctv? 2024-04-27 19:11:57 like you know happened to ola 2024-04-27 19:12:12 😔, 2024-04-27 19:12:14 😔 2024-04-27 19:12:44 all you need is a backup of your detached header, and then failures of usbsticks is not an issue 2024-04-27 19:13:04 Yea, true 2024-04-27 19:13:10 but over the last 15 years (afaicr) no usb stick died on me 2024-04-27 19:13:22 no boot stick, and neither any other 2024-04-27 19:14:00 Did you loose them after 1 year? I have had sticks die within one year (most survived around 4 years tho') 2024-04-27 19:14:03 but then bootsticks usually are dedicated sticks, with very little data on them, an initrd and nothing else 2024-04-27 19:14:11 fuck 2024-04-27 19:14:37 i also don't loose sticks, especially bootsticks i make sure i always maintain physical contact 2024-04-27 19:14:40 I flashed my boot sticks every other day 2024-04-27 19:14:59 ah, i do that much more rarely, maybe 3-4x a year. 2024-04-27 19:15:25 hello 2024-04-27 19:15:55 I'll ask a real question now. Why my wifi show up as eth1 and refuses to scan? iwctl just spits out "Operation not permitted" 2024-04-27 19:16:40 use with sudo 2024-04-27 19:16:48 Yea I done 2024-04-27 19:16:49 you use networkmanager right? 2024-04-27 19:17:13 I can turn off and do other things, but specifically scan is not permitted on eth1, and it is a wireless interface 2024-04-27 19:17:35 gprbuild is outdated on edge, what could i do to update it? 2024-04-27 19:17:35 Mostly asking incase someone knows about that weird issue. Old thinkpad 2024-04-27 19:21:33 is ifconfig eth0 up? 2024-04-27 19:22:28 Yes. eth0 is cable and eth1 was wifi 2024-04-27 19:22:58 But I gave up, and installed OpenBSD since the wifi just werked on that combo 2024-04-27 19:23:09 But I am willing to give alpine a try again 2024-04-27 19:24:13 wine now has wow64 which means lib32 is not needed, can it be compiled to run on musl? 2024-04-28 13:58:46 Do you already know anything about when 3.20 is going to be released? I am asking because I am following edge and I plan to stay on 3.20. I do not want to risk upgrading past 3.20 by mistake. 2024-04-28 14:00:46 engblom: somewhere next month is the planning, but the builders are not finished yet 2024-04-28 14:01:46 ikke: I do not need any exact date, but do you think it will be in the begining, middle or end of the month? 2024-04-28 14:02:12 probably more along the end of the month 2024-04-28 14:02:39 Okay, thanks! Then I know it is low risk to continue to upgrade the next few weeks. 2024-04-28 14:29:58 I have the very awkward issue of not being able to login to my root account with `su` in a Gnome session. In the other, GUI-less tty it works fine. I might sure 20 times that the password is correct 2024-04-28 14:30:27 I use a different keyboard layout in gnome, but I switched that to US already, too, to rule it out 2024-04-28 14:38:53 ah nevermind, i just reset with US 2024-04-28 17:06:06 i upgraded from 3.18 to 3.19 and ended up with an broken apk on i586 2024-04-28 17:06:37 is it possible that alpine accidentally introduced an requirement on i686 for the x86 arch? 2024-04-28 17:07:04 not accidentially 2024-04-28 17:07:07 accidentally 2024-04-28 17:07:29 i dont see it in the release notes 2024-04-28 17:07:41 https://www.alpinelinux.org/posts/Alpine-3.19.0-released.html 2024-04-28 17:16:14 It missed the release notes apparently, but https://gitlab.alpinelinux.org/alpine/aports/-/commit/8f820025bcf606a5adc1bcc970b5a67b33c236c6 2024-04-28 17:16:27 well duh, my apk is now bricked 2024-04-28 17:17:02 from the issues at least two other people ran into this 2024-04-28 17:17:08 this should get added to the release notes 2024-04-28 17:19:28 https://gitlab.alpinelinux.org/api/v4/projects/5/packages/generic//v2.12.9/x86/apk.static will probably still work 2024-04-28 17:20:13 rip 2024-04-28 17:23:00 rip https://pomf2.lain.la/f/dc1kbmq.jpg 2024-04-28 17:24:26 ikke: if that platform is not supported by alpine anymore, i'll have to find a new OS anyways or give up the machine 2024-04-28 17:24:57 Sounds all a bit awkward – this is a very major change. Usually distros discuss this intensively before applying it 2024-04-28 17:26:04 was discussed in https://gitlab.alpinelinux.org/alpine/tsc/-/issues/35 2024-04-28 17:26:08 just not communicated 2024-04-28 17:33:57 nero, guess you'll become a Slackware user then soon ;) 2024-04-28 17:36:18 jemius: thats a bingo for me: http://www.slackware.com/zipslack/sysreq.php 2024-04-28 17:38:21 I just failed miserably setting up an Alpine machine with Gnome and working sound 2024-04-28 17:39:01 I too had to retire a a lot of elderly hardware is extremely energy-inefficient :( 2024-04-28 19:47:20 alpine-curious here, anyone daily-drive a diskless system with lbu? 2024-04-28 20:41:37 "alpine-curious here, anyone..." <- I used to daily drive alpine, but not diskless 2024-04-28 20:41:49 I daily drive openbsd now 2024-04-28 20:56:09 I know that nvidia proprietary driver aren't usable on musl, but is it possible to "bypass" it by using flatpak ? like installing steam with flatpak and then install the nvidia driver in flatpak, will it work ? 2024-04-28 22:24:47 vova: try installing an ubuntu or debian chroot and run your stuff in there. 2024-04-28 22:33:13 Yeah I think it'll be simpler, I know that you can use xauth stuff to run GUI stuff in a chroot, do you know any alternative with Wayland ? I thought about using vnc but I plan to play game with steam (and so nvidia driver) from the chroot so... 2024-04-28 22:38:28 no idea about wayland, sorry. not my cup of tea 2024-04-28 22:38:55 you can most likely use bwrap 2024-04-28 22:39:06 i don't do any of the nvidia stuff but i do run stuff under wayland using it 2024-04-28 22:39:27 ah wait, you're talking about flatpak 2024-04-28 22:39:32 but well, flatpak's using bwrap underneath... 2024-04-28 22:39:40 nah i diverted to ubuntu/deb chroot 2024-04-28 22:39:52 i'm talking to vova 2024-04-28 22:40:24 yes, me too, i diverted vova to deb chroot 2024-04-28 22:47:14 I'll try setuping a Debian or arch chroot and first installing the proprietary nvidia driver in it 2024-04-28 22:50:23 btw i have amdgpu on alpine and it works perfectly with steam in flatpak. bg3, dragons dogma 2, and other recent games all work well. maybe a reason to ditch proprietary nvidia and go amd? 2024-04-28 22:51:26 I want to ditch nvidia don't worry lol, but unfortunately my laptop has a nvidia card so... 2024-04-28 22:52:46 and tbh I don't want to buy another laptop, this one is working absolutely fine, I just didn't use the nvidia but only the integrated Intel GPU since years and everything is running smoothly, but recently I wanted to play some game (CS2) and the integrated GPU is just too weak for it lol 2024-04-28 22:53:01 have you tried nouveau 2024-04-28 22:53:16 it might not be as performant as the proprietary driver though 2024-04-28 22:53:29 and i'm unsure of its compatibility with all nvidia cards 2024-04-28 22:53:31 but you can check it out 2024-04-28 22:56:40 tried it, the performances are really really bad (for my card at least) 2024-04-28 22:56:48 i see 2024-04-29 00:24:39 Hi! I've got an old notoriously troublesome laptop model (Asus X205TA) that, surprisingly nearly everything works out of the box with on Alpine these days. 2024-04-29 00:24:39 That said, I tried luks disk encryption, and the built-in keyboard does deliver keypresses in the boot phase in which the encryption key needs to be supplied. It looks like I need the hdi-asus kernel module to be available to the initramfs. The keyboard works fine in the live usb, and I've confirmed that an rmmod of hdi-asus does indeed disable keyboard input on that machine. 2024-04-29 00:24:39 I'm not all that familiar with configuring the bootstrap process, and have been unsuccessful so far. Does anyone happen to have a link to a guide for including specific modules in either setup-disk or via mkinitramfs, or has anyone dealt with a similar issue? 2024-04-29 00:24:41 fwiw, the mkinitramfs usb "feature" appears to be included by default via setup-disk, and the usb feature does list the hid-asus.ko file that I think I need. 2024-04-29 00:25:33 * the built-in keyboard does *not* deliver keypresses during that boot phase 2024-04-29 05:29:42 ikke: the apk.static you linked is not execv'able on that system 2024-04-29 05:29:54 unfortunately, i also got no strace so i don't think i can see why 2024-04-29 05:35:29 19:23:00 nero$ rip https://pomf2.lain.la/f/dc1kbmq.jpg 2024-04-29 05:35:32 nice term :D 2024-04-29 05:36:39 its the thing i use in the morning before the real hardware gets booted up 2024-04-29 05:51:08 yo! is this place alive? I have a few questions 2024-04-29 05:51:55 well seems alive enough... 2024-04-29 05:52:07 question: whats the stance on rust apps in alpine linux? 2024-04-29 05:52:32 are there packages for common libs? like rust stdlib, to link dynamically to it? 2024-04-29 06:24:46 no, it's being discussed in https://gitlab.alpinelinux.org/alpine/tsc/-/issues/76 2024-04-29 06:31:40 I just want my own app (which I will soon ask to be taken into alpine repos) to be small as possible, I am willing to help with that assumption that I as a developer will do the best on my side to dyn link if possible 2024-04-29 06:32:03 the question is whether alpine is interested in shipping dyn linking versions of common rust packages 2024-04-29 06:33:10 can I reply to the thread using email 2024-04-29 06:59:48 afaik you have to register a gitlab.alpinelinux.org account 2024-04-29 12:08:13 I installed alpine on a GPT hard drive. The system installed correctly. After restarting the computer, unfortunately it does not want to start - there is no visibility of the disk from which the system should boot. 2024-04-29 12:08:17 During installation, I chose how to partition the disk: 2024-04-29 12:08:20 crypted, sys 2024-04-29 12:08:22 After installation, the following partitions are visible on the disk: 2024-04-29 12:08:25 FAT32 300 MB Flags: esp, boot Other 101 GB Other 17 GB 2024-04-29 12:08:27 UEFI settings 2024-04-29 12:08:30 - AMD AHCI BIOS Rom - ENABLED 2024-04-29 12:08:32 - SATA IDE Combined mode - DISABLED 2024-04-29 12:08:35 - Compatibility Support Module - DISABLED 2024-04-29 12:08:37 - Secure Boot - DISABLED 2024-04-29 12:08:40 System/Hardware 2024-04-29 12:08:42 Alpine Extended 3.19.1 2024-04-29 12:08:45 HDD: 1TB 2024-04-29 12:08:47 If the disk was in the MBR format, the operating system started correctly after installation. 2024-04-29 12:08:50 How to fix it? 2024-04-29 12:09:27 Partitions on the disk: 2024-04-29 12:09:38 1) FAT32300 MBGPT, EFI System Partition 2024-04-29 12:09:42 2) Other101 GBGPT 2024-04-29 12:09:46 3) Other17GBGPT 2024-04-29 12:23:41 dawciobiel: where does it fail in the startup sequence? 2024-04-29 13:34:12 @nero just on after the POST. UEFI just dont see from what disk it shoot boot-up. 2024-04-29 13:34:59 nero: just on after the POST. UEFI just dont see from what disk it shoot boot-up. 2024-04-29 15:11:14 it seems like the license for catfish is wrong in the APKBUILD 2024-04-29 15:11:30 looking through the codebase the correct license seems to be GPL-2.0-only, but the template uses GPL-2.0-or-later 2024-04-29 15:14:53 Interesting, most packages mark it as GPL-2.0-or-later: https://repology.org/project/catfish/information 2024-04-29 15:15:29 ah, hmm 2024-04-29 15:15:32 okay i see what's happening 2024-04-29 15:15:49 the headers seem to indicate GPL-2.0-only: https://gitlab.xfce.org/apps/catfish/-/blob/master/setup.py?ref_type=heads#L8 2024-04-29 15:16:08 but then below it also declares that the license is GPL-2.0-or-later: https://gitlab.xfce.org/apps/catfish/-/blob/master/setup.py?ref_type=heads#L263 2024-04-29 15:16:32 maybe good to ask for clarification 2024-04-29 15:16:38 yeah i'll bring it upstream 2024-04-29 15:16:38 but I think legally the headers would prevail 2024-04-29 15:16:45 most likely 2024-04-29 15:16:46 (ianal) 2024-04-29 15:16:48 (not a lawyer) 2024-04-29 15:16:50 heh 2024-04-29 15:20:52 another thing, is this really still the right url for gigolo? https://git.alpinelinux.org/aports/tree/community/gigolo/APKBUILD#n8 2024-04-29 15:21:22 doesn't seem like it to me because it's hosting an old version of the thing 2024-04-29 18:11:03 I installed netcat-openbsd 2024-04-29 18:11:19 nc command not found bat 2024-04-29 18:11:38 nm 2024-04-30 09:18:11 hey, when i do `lspci -v` it segfault 2024-04-30 09:19:03 because of a strlen call on an invalid pointer 2024-04-30 09:20:44 do you know how i can get the debug symbols for libpci.so.3 ? 2024-04-30 09:47:07 dok, seems like it's not packaged 2024-04-30 09:49:15 other topic, it seems that gtk (at least claws-mail) isn't able to load icons anymore 2024-04-30 09:49:19 it fail in a subcall of pci_fill_info 2024-04-30 09:49:33 (claws-mail:5286): GdkPixbuf-WARNING **: 11:49:22.871: Error loading XPM image loader: Image type “xpm” is not supported 2024-04-30 09:49:38 stock_pixmap.c:722 Condition pix_d->pixbuf != NULL failed 2024-04-30 09:49:40 which calls to a another function through a function pointer 2024-04-30 09:49:57 dok, which arch? 2024-04-30 09:50:11 x86_64 2024-04-30 09:50:12 it works here, on amd64 2024-04-30 09:50:38 Do you know on which driver/device it crashes? 2024-04-30 09:50:51 maybe thats the device's fill_info methode that's broken 2024-04-30 09:50:52 Maybe it's missing a dependency 2024-04-30 09:51:00 i do not know yet 2024-04-30 09:51:51 I mean in order, by comparing the last one that works and one that doesn't ^^ 2024-04-30 09:53:32 dok, now that you talk about it 2024-04-30 09:53:39 I think I had this problem a couple days ago 2024-04-30 09:54:01 On a new install, trying to reorder the audio devices on the snd_hda_intel driver, with “bad parameters” 2024-04-30 09:54:05 It did crash lspci 2024-04-30 09:54:42 (the problem in question was specifying indexes of 1 and 2, without a default 0 device) 2024-04-30 09:54:43 it seems to only work for the first two devices 2024-04-30 09:54:46 (for me) 2024-04-30 09:55:22 And what's the third, without the -v? 2024-04-30 09:56:21 > 00:01.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Family 17h-19h PCIe Dummy Host Bridge (rev 01) 2024-04-30 09:58:16 strange 2024-04-30 09:59:07 yeah i don't really get it 2024-04-30 11:00:04 hi 2024-04-30 11:00:16 any news on the transition from openrc to s6 on alpine ? 2024-04-30 11:01:19 no 2024-04-30 11:28:14 quinq: the device uses "linux_sysfs" as the backend methods to fill the information (not sure if other uses different methods) 2024-04-30 11:29:12 i know some quirks of using musl's printf on sysfs 2024-04-30 11:29:22 i wonder if this is related 2024-04-30 11:29:38 (using readv/writev on sysfs) 2024-04-30 14:26:35 what, can it be caused by a signed extension on a pointer ? 2024-04-30 14:27:38 the result of a basename is passed to another function as the third argument `pci_set_property(struct pci_dev *d, u32 key, char *value)` 2024-04-30 14:27:48 and i see the following instrucion just before the call: 2024-04-30 14:28:01 3> 0x7ffff7f4f080 movslq %eax,%rdx 2024-04-30 14:28:38 wtf, the string pointer is: 2024-04-30 14:28:44 0x7ffff7ffed20: "/sys/kernel/iommu_groups/0" 2024-04-30 14:28:51 and after base name I have: 2024-04-30 14:28:57 0x7ffff7ffed39: "0" 2024-04-30 14:29:05 which is turned into: 2024-04-30 14:29:17 0xfffffffff7ffed39: 2024-04-30 14:37:49 .... could it be because the basename symbols and thus the return value is considered to be an int ? (hypothesis) 2024-04-30 14:40:57 yeah that's it 2024-04-30 14:41:19 i compiled main/pciutils/ and it showed a warning about basename not being defined ! 2024-04-30 14:42:15 gg 2024-04-30 14:43:22 ACTION feels like a genious 2024-04-30 14:43:34 Well, that's a pretty good debug session you did :) 2024-04-30 14:43:34 ACTION also spent the entire day on this 2024-04-30 14:43:51 i can rest in peace now 2024-04-30 14:43:54 :> 2024-04-30 14:44:01 Fixes are usually simple 2024-04-30 14:44:10 when the bug is undertood 2024-04-30 14:44:18 The hard part is being able to reproduce the issue, tracking down where it happens, and why it does 2024-04-30 14:44:41 yeah in this case it was easily reproductible 2024-04-30 14:44:54 also i learn about the ignore cmd in gdb 2024-04-30 14:45:17 so i could skip the 50 first time the breakpoint was hit 2024-04-30 14:45:43 :D 2024-04-30 14:45:55 You can also refine a breakpoint to match certains conditions 2024-04-30 14:46:03 but yeah 2024-04-30 14:46:28 i also did that, but it was too deep to understand what was going on 2024-04-30 14:46:43 debuging without much debug info is a pain tho 2024-04-30 14:46:50 indeed 2024-04-30 14:47:25 good thing addresses where not randomized each run 2024-04-30 14:49:01 “You can ask questions and report bugs on the linux-pci mailing list running on Vger, but please Cc the author (mj@ucw.cz).” 2024-04-30 14:50:21 thanks 2024-04-30 14:50:35 ACTION writting the fix 2024-04-30 15:20:28 you can also use rr 2024-04-30 15:20:31 reverse debugger 2024-04-30 15:20:40 run until thing breaks, then run backwards 2024-04-30 15:20:48 its gdb plus fancy stuff 2024-04-30 15:38:33 ive used r2 and gdb 2024-04-30 15:38:41 i dont know much about rr 2024-04-30 16:07:15 oh, https://rr-project.org/ looks very interesting 2024-04-30 16:12:45 quinq: maybe i should also cc musl, what do you think ? 2024-04-30 16:14:01 dok: there is #musl on irc.libera.chat 2024-04-30 16:14:22 yeah i was going to ask them 2024-04-30 16:14:34 thanks for the remainder, no worries 2024-04-30 16:22:41 dok, I don't think so 2024-04-30 16:22:47 It has nothing to do with musl 2024-04-30 16:23:13 It's a (POSIX) C programming error 2024-04-30 16:23:29 quinq: well... i agree but let's see what pciutils maintainers will say 2024-04-30 16:23:52 It's not an opinion ^^ 2024-04-30 16:23:58 I am asking because man basename(3) mention something about _GNU_SOURCE being defined before including string.h 2024-04-30 16:24:18 which should ? also define basename 2024-04-30 16:24:18 Ah, that's specific to GNU libc 2024-04-30 16:24:22 yeah 2024-04-30 16:24:34 but maybe that's something than musl could do, no ? 2024-04-30 16:24:36 :3 2024-04-30 16:24:39 But string.h is not the header for basename 2024-04-30 16:24:49 It's libgen 2024-04-30 16:25:10 true, but why tho ? 2024-04-30 16:25:14 Why what? 2024-04-30 16:26:27 why it is an issue being defined in libgen.h and not in string.h? 2024-04-30 16:27:02 Because string.h isn't the header defining basename, it's libgen.h that does 2024-04-30 16:27:20 It's not an issue, it's just how it's specified by the standard 2024-04-30 16:27:41 By the way, you can directly ask Hello71 here who wrote that it ;) 2024-04-30 16:28:25 I don't think it's an intentional omission, just a bug like most of them happen, forgot to add something while adding something 2024-04-30 16:29:33 s/that it/that in/ 2024-04-30 16:34:53 okay 2024-04-30 16:52:41 quinq: patch sent, thanks a bunch ! 2024-04-30 17:04:14 dok: to add a little bit of context, GNU libc has two definitions of basename(), the one from string.h being non-POSIX ( and apparently also more commonly used ) behaves a little differently 2024-04-30 17:04:52 which is also why you can't always blindly replace the #include to point to libgen.h, because it might break on some edge cases 2024-04-30 17:04:58 Nice, though the problem isn't really sign extention, but value truncation 2024-04-30 17:05:34 ptrc, good point 2024-04-30 17:08:44 Though that makes it non-portable 2024-04-30 19:35:37 Hu hu 2024-04-30 19:36:45 Does alpine support tl-wn725n wifi adapter 2024-04-30 19:36:58 İ want to buy it but idk if it supports 2024-04-30 19:45:37 Searching online gives mixed results with built-in support 2024-04-30 19:47:14 uh 2024-04-30 19:47:25 i happen to own the exact same device, i will test right now 2024-04-30 19:47:37 rtl8818eu 2024-04-30 19:47:41 seems to be the module 2024-04-30 19:51:10 r8188eu: module is from the staging directory, the quality is unknown, you have been warned. 2024-04-30 19:51:19 aside from that, requires linux-firmware-rtlwifi firmware package, but works 2024-04-30 19:51:57 provided the recent wn725n devices are the same that i have right here and not some other chipset 2024-04-30 19:56:31 Nero 2024-04-30 19:56:40 So it's working right? 2024-04-30 19:57:01 yes 2024-04-30 19:57:10 but my wn725n is at least 8 years old 2024-04-30 19:57:22 Np 2024-04-30 19:57:27 As long as it works 2024-04-30 19:57:28 sometimes vendors put in new chips and sell it as the same product 2024-04-30 19:57:30 I'm okay 2024-04-30 19:59:02 How did you install it 2024-04-30 19:59:35 My current adapter is brcm43142 and because it is dkms it isn't supported by alpind 2024-04-30 20:00:27 i only installed the linux-firmware-rtlwifi package 2024-04-30 20:00:37 Okay thank you so much 2024-04-30 20:00:49 Finally i will be able to install alpine 2024-04-30 20:01:36 np 2024-04-30 20:19:44 Hi, is there a way to apply new interfaces added to /etc/network/interfaces without just restarting networking service and dropping connection? 2024-04-30 20:20:07 ifdown ; ifup 2024-04-30 20:20:23 hmm that works, thanks! 2024-04-30 20:20:39 I really need to learn the "classic" linux tools not just systemd ones :/ 2024-04-30 20:45:10 ifup/ifdown are interfaces to iproute2 if you want to scratch deeper