[09:42] <nteodosio> It seems that if you write any inexisting URL for autopkgtests results, it gets an unauthorized message, e.g. https://autopkgtest.ubuntu.com/results/autopkgtest-mantic.
[09:42] <nteodosio> This is a quite confusing when the page doesn't yet exist (simply because the tests haven't finished running).
[09:43] <nteodosio> What is the right department to file a bug -- or is there already one?
[09:51] <seb128> nteodosio, https://bugs.launchpad.net/auto-package-testing/+filebug
[11:00] <ginggs> @pilot in
[13:12] <rbasak> Eickmeyer: o/ bug 2033992 is showing in pending-sru as ready for Jammy. Is that status the status you expect, or should it be verification-failed now?
[13:12] -ubottu:#ubuntu-devel- Bug 2033992 in ubuntustudio-menu (Ubuntu Lunar) "[SRU] Change Ubuntu Studio Chat from Matrix to IRC" [High, Fix Committed] https://launchpad.net/bugs/2033992
[13:14] <rbasak> enr0n: regarding https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/2036358/comments/28, would it be appropriate to retry these with migration-reference/0 then? I'm not supposed to release without the failures being clean, one way or another.
[13:14] -ubottu:#ubuntu-devel- Launchpad bug 2036358 in systemd (Ubuntu Lunar) "systemd wait-online now times out after jammy and lunar upgrade" [Critical, Fix Committed]
[13:18] <enr0n> rbasak: yes, I think that would be okay then
[13:25] <rbasak> OK, all submitted, thanks. Let's see if that works.
[13:26] <enr0n> rbasak: thanks!
[14:17] <Eickmeyer> rbasak: That one is good for Jammy. It's bad for lunar and needs to be re-done for lunar due to another SRU that was attempted. I've been eyeballs-deep in fixing mantic bugs before final freeze.
[14:38] <rbasak> Thanks. Released then, with a comment in the bug.
[15:00] <jbicha> @pilot in
[15:01] <ginggs> @pilot out
[18:39] <tsimonq2> juliank (cc rbasak): I have a local patch building now for apt to suggest git-ubuntu on Ubuntu machines, when using apt source. Is this an agreeable approach?
[18:42] <tsimonq2> If I don't hear anything otherwise, I'll submit it to Salsa for a review. Any feedback, including how to get the vendor information in the cleanest way possible, would be appreciated.
[18:48] <juliank> tsimonq2: This is going to need a new field in the Release file to specify a default URL for git repositories; similar to Changelogs and Snapshots
[18:48] <juliank> You could set a default based on Origin though as both of these features do
[18:49] <juliank> See apt-pkg/init.cc for the funny options to configure those
[18:49] <tsimonq2> Cool. What about usage of the `git ubuntu clone` command instead of a bare Git clone, does that seem like a
[18:49] <juliank> And you can look at the 2.4.y branch for a clean backport of the snapshots feature for inspiration
[18:49] <tsimonq2> ..."hackier" approach?
[18:49] <juliank> Hmm
[18:50] <juliank> How about Vcs-Clone command
[18:50] <juliank> Vcs-Clone: git ubuntu clone @SOURCEPACKAGE@
[18:50] <juliank> And apt includes "Run <Vcs-Clone here> command to get the source code for this package"
[18:51] <juliank> I think this probably is the nicest approach
[18:51] <tsimonq2> That could certainly work :) if there is an existing Vcs-Git, would Vcs-Clone auto-populate with the appropriate `git clone` command?
[18:51] <juliank> I have not formed an opinion on that matter
[18:52] <juliank> Also what happens with missing git-ubuntu repos
[18:52] <juliank> Some packages are not imported
[18:52] <juliank> Maybe we can have launchpad figure out which packages have repos and add Vcs-Clone fields to the package entries
[18:52] <juliank> :D
[18:53] <tsimonq2> That's a really good point; I'd almost like to ping Robie and Colin to ask for their input, but part of me thinks this belongs on ubuntu-devel@?
[18:54] <juliank> Go for the email!
[18:54] <tsimonq2> Sounds good! I'll CC you :)
[18:58] <rbasak> It's...complicated
[18:59] <rbasak> git-ubuntu is convenient in that it'll give you _actual_ source used to build a package, but there are still some rough edge cases.
[19:00] <rbasak> But if you are submitting a fix, then the team that maintains the package may prefer that you base your changes from a commit from Vcs-Git (or even somewhere else that no metadata points to) with which the git-ubuntu branch has no common ancestor at all.
[19:00] <rbasak> OTOH, I'd like for those teams to accept git-ubuntu MPs since rebasing is trivial and that way it's the same workflow for every package, which is much easier for first time contributors.
[19:01] <jbicha> @pilot out
[19:08] <tsimonq2> juliank, rbasak: https://lists.ubuntu.com/archives/ubuntu-devel/2023-October/042813.html
[23:02] <mwhudson> what's the easiest way to answer "is this binary package in main" in a script i wonder
[23:07] <rbasak> rmadison -cmain -aamd64 -sfocal mosh
[23:07] <rbasak> No answer = not in main
[23:07] <rbasak> rmadison -cmain -aamd64 -smantic mosh
[23:07] <rbasak> Answer = in main
[23:07] <rbasak> You could use -cuniverse to confirm for certain
[23:07] <rbasak> (rather than just a nonexistent package)
[23:08] <rbasak> I don't know how well that'll handle all errors though
[23:08] <rbasak> Another way might be chdist + grep-dctrl
[23:08] <rbasak> and looking at the Section field
[23:08] <rbasak> Or APT-Sources
[23:09] <rbasak> That's probably more reliable but requires setup
[23:11] <mwhudson> yeah i cooked up something with chdist and indextargets