[11:04] <DarkTrick> **Q** @ Someone familar with I18n for C programs: I'm currently upgrading the source for package `gjiten` to GTK 3. I have some trouble with internationalization: tutorials say I should `#define _(str) gettext (str)` etc. in my source. However I could find it nowhere in the origial source. Should I suspect something wrong, or are there configurations (through configure.ac,...) that make these defines unnecessary?
[11:04] <DarkTrick> Source btw: https://github.com/DarkTrick/gjiten
[11:09] <cjwatson> DarkTrick: some glib headers define that macro
[11:09] <cjwatson> which will be included indirectly by gtk headers
[11:10] <DarkTrick> cjwatson, so if I use gtk, I won't need the defines specifically in my code
[11:10] <DarkTrick> cjwatson, thank you!
[11:10] <cjwatson> https://developer.gnome.org/glib/stable/glib-I18N.html
[11:10] <DarkTrick> cjwatson, a reference! thank yoU!
[11:10] <DarkTrick> you
[11:11] <cjwatson> np
[11:29] <seb128> bdmurray, hey, any idea why https://errors.ubuntu.com/problem/bbaeb2a51fa715f17ffcac958fecad4e400436aa is failing to provide a backtrace on the retracers?
[12:15] <DarkTrick> **Q** @ configure.ac: is it a good thing to not use `AM_MAINTAINER_MODE` in a configure.ac?
[12:39] <eoli3n> Hi
[12:40] <eoli3n> xnox i use ansible to install some snap package at the end of a preseed installation
[12:41] <eoli3n> my tasks fails because of a package not found
[12:41] <eoli3n> see https://x0.at/VAO.txt
[12:41] <eoli3n> but when i first boot my system and then use "snap install pdftk" it works
[13:43] <dbungert> seb128: thanks for the heads up on LP: #1926298, I put that in my queue
[13:44] <seb128> dbungert, thanks, I didn't spend much time on it but I sounds like using g_find_program_in_path could simplify your code and maybe fix the issue as a side effect (if the bug is your custom implementation of a similar function)
[13:44] <dbungert> sounds great, I'll look into it
[13:45] <seb128> 👍
[13:46] <eoli3n> is there a way to tell ubuntu installer to not create "ubuntu" default efi entry ?
[13:46] <eoli3n> any preseed key ?
[13:50] <xnox> eoli3n: snap install will not work in a chroot, with /run mounted from host. instead you should use /usr/lib/snapd/snap-preseed --reset to reset the /target system, and then use `snap download` to get any extra snaps & assertions, drop them in /var/lib/snapd/seed and modify the yaml file there.
[13:50] <xnox> eoli3n:  this is the correct way to download snaps, and make snapd "install them on first boot on a classic system fully offline/noninteractive"
[13:50] <xnox> eoli3n:  that's how i.e. we install snap-store snap, lxd snap, etc. on various classic installers.
[13:51] <eoli3n> so the problem comes from "snap" ansible module
[13:51] <eoli3n> i will simlpy "when: not ansible_is_chroot" on this
[13:51] <eoli3n> what about the efi entry creation ?
[13:51] <xnox> eoli3n:  yes and no, maybe. ansible normally does "ssh as root into the host" and in that case "snap install does work"
[13:52] <xnox> eoli3n:  efi entry it is always created and it is harmless. You can use efibootmgr to delete it, but the disk image, will always create it on first boot and on upgrades.
[13:52] <eoli3n> xnox not really, for exemple, the module "systemd" skip tasks about services, when chroot is detected
[13:52] <xnox> (ie. boot/bootx64.efi will launch boot/fbx64.efi which will create it on first boot)
[13:52] <eoli3n> xnox, how to keep pxe boot as default entry then ?
[13:53] <xnox> eoli3n:  good question, in some servers pxe boot can be set as always-preferred which overrides any distro attempts to set local boot as first one.
[13:53] <xnox> or use efibootmgr to _reorder_ but do not _delete_ ubuntu entry.
[13:53] <eoli3n> lets try this
[13:53] <xnox> ie. efibootgr you want to reorder to make pxe first.
[13:53] <eoli3n> thanks for your help
[13:54] <xnox> eoli3n:  re systemd / chroot => i don't know if ansible does it or systemctl. Cause systemctl itself has chroot detection and doesn't do anything then.
[13:54] <eoli3n> i tried, and as i remember, reordering with efibootmgr is pain
[13:54] <eoli3n> ah maybe its systemctl's detection
[13:55] <eoli3n> so why snapd doesnt detect the chroot before trying packages ?
[13:55] <xnox> eoli3n:  well you can always copy and install BootOrder variable as a file, if you know which order things are in =)
[13:55] <eoli3n> that's not even snapd, which is not running
[13:55] <eoli3n> i miss something here
[13:55] <xnox> i.e. one can copy the binary /sys/firmware/efi/efivars/BootOrder-8be4df61-93ca-11d2-aa0d-00e098032b8c file around if the bootorder of entries in the same on all machines.
[13:56] <eoli3n> i remember that you worked on the chroot detection for the chromium package
[13:56] <xnox> eoli3n:  "worked" => open a bug report to fix the postinst.
[13:56] <eoli3n> good to know thanks
[13:56] <xnox> eoli3n:  but i kind of agree that `snap install` should be sane when called from a chroot.
[13:57] <eoli3n> snap install should first test if chroot, before doing anything
[13:57] <eoli3n> doesn't it ?
[13:57] <eoli3n> lets open a bug report
[13:57] <xnox> no idea, you need to ask snapd people
[13:57] <eoli3n> hm
[13:57] <eoli3n> so bug report on github/snapd
[14:01] <eoli3n> if i snap install an existant package, i get a drop because of the chroot. Here, the problem is that the package exists, but snap can't find it in post install, and early after the first boot
[14:06] <eoli3n> https://bugs.launchpad.net/snappy/+bug/1926301 : done
[23:16] <DarkTrick> *Q* @ internationalization: are gettext and intltool opposing technologies?