[09:24] <rbasak> We managed to upload a broken git-ubuntu snap to the edge channel yesterday
[09:25] <rbasak> Possibly due to some non-determinism in snapcraft, or because the Launchpad build environment uses a different snapcraft from our CI or something (it had passed CI).
[09:25] <rbasak> Our snap has a self-test that succssfully detects the problem as soon as the snap is installed, and it looks like that's working correctly.
[09:26] <rbasak> Is there any way we could use this to have Launchpad detect a problem prior to upload?
[09:26] <rbasak> "After build, install the snap, and if <test> fails, treat the build as failed"
[09:32] <wgrant> rbasak: Launchpad's unlikely to define its own format for specifying how to test snaps (particularly since you usually can't just install, you have to autoconnect etc. as well). It might be worth discussing with snapcraft people whether some unified mechanism there makes sense.
[09:33] <rbasak> wgrant: are you suggesting that snapcraft could provide a hook (or just do it by default) and then Launchpad would use that?
[09:34] <wgrant> rbasak: I think snapcraft could define a way to specify a test in snapcraft.yaml, and then LP would invoke it to run that, yeah.
[09:35] <wgrant> Similar to how there's a common interface for autopkgtests in Debian source packages
[09:35] <rbasak> OK, thanks. I'll ask on the forum.
[09:45] <rbasak> https://forum.snapcraft.io/t/launchpad-post-build-pre-upload-testing/5545?u=rbasak
[16:12] <nacc> rbasak: there is a step that can be done in the clients
[16:12] <nacc> rbasak: there's already a bug filed for it with instructions
[16:12] <nacc> so even if the LP -> store upload occurs, the users won't ever get a broken snap
[17:37] <ahasenack> hi, while checking out some bugs and why incomplete ones have expired, I came across this: https://bugs.launchpad.net/bugs/1677541/+watch/123799
[17:37] <ahasenack> "Recent activity" is full with oopses
[17:37] <ahasenack> the github url looks normal, and loads just fine: https://github.com/systemd/systemd/issues/3374
[17:37] <ubot5-ng> systemd bug 3374 in systemd "udevd: `Could not generate persistent MAC address for $name: No such file or directory`" (comments: 23) [Bug 🐛, Network, Udev, Open]
[17:37] <ahasenack> last oops is OOPS-2909d29f2613ea5e39a7fdb97f6279ac
[17:38] <ahasenack> it's a KeyError
[19:26] <zleap> hi