=== chihchun_afk is now known as chihchun [07:06] hello! [07:07] I'm testing a script that interacts with launchpad and before messing up the production database, I would like to test it on staging area [07:07] I mean staging launchpad [07:07] in launchpad help page it says: "You can't create a new account on staging — instead, create one in Launchpad's production environment and then wait up to 24 hours for your account to be available on staging." [07:07] I have an account that was created and activated a long time ago and yet I can't login [07:09] I see "Incorrect email/password combination" when I try to log with the same info as the one I use on production environment... does anyone know the problem? [07:16] I have 2 e-mail addresses attached to my launchpad account, but it looks like it fails with both === ePierre is now known as ePierre-afk [17:13] what's going on with LP builders? 15hour long queue for amd64? [17:15] clivejo: there is a infra rebuild going on, meaning limited capacity, and the archive recently opened for bionic, so it's probably loaded [17:16] why they private builds? [17:17] clivejo: ? [17:17] Building private job ? [17:18] clivejo: sorry, I'm not sure I understand your question still [17:18] all the jobs are currently showing as "Building private job" for the amd64 builders [17:19] or cleaning [17:19] iirc, private jobs get prioritized higher, but i might be wrong [17:19] just wondering why they are private [17:24] Security builds [17:24] And yes, the lcy01 cloud is being redeployed at the moment, as part of making it possible to add extra capacity to it [17:25] Having these two things happen at the same time is unfortunate, but oh well ... [17:25] cjwatson: it's amazing how quickly people noticed :) [17:25] *three, in fact, counting auto-sync [17:25] yeah [17:25] a perfect storm! [17:25] Any one of those would have produced a backlog [17:26] And yeah, I'm sorry, I can't go into more detail about private builds in public. Hopefully they'll clear out in an hour or two [17:32] a bit over a hundred builds to go in the relevant security queue, I think [22:40] Hello, I need help with creating a new branch for my project. After executing bzr push lp:gnome-paint/stable I get bzr: ERROR: Permission denied: "Cannot create 'stable'. Only Bazaar branches are allowed." [23:03] Could you help me? [23:12] Etua: A branch needs an owner, so you need to push to lp:~USER_OR_TEAM/gnome-paint/stable [23:12] lp:gnome-paint/stable is the branch for gnome-paint's "stable" series, which doesn't exist. [23:14] cjwatson: wgrant: I think I'm missing something in the definition of git_target in the launchpad API. I'm working on the repointing script that will, as we import source packages, change the default for a given source package to the imported repository. Is the target somethig like ubuntu/+source/ ? current script is here: http://paste.ubuntu.com/25868780/. I'm also not sure how to test it. Can I [23:14] unset the default repository to None? [23:16] I usually spell it with a leading slash, but yes, a distribution_source_package (whose URL is /ubuntu/+source/) is an example of a git_target [23:16] and the relevant example here [23:17] you can unset the default repository for a given target by passing repository=None, yes [23:17] cjwatson: cool, that helps :) [23:18] PS you should pass the package names through urllib.parse.quote() [23:18] when you're substituting them into URL paths, anyway [23:18] cjwatson: yep, thans as always! [23:18] LP will sometimes let you get away without that, but should quote anyway [23:18] yep [23:26] cjwatson: great! thanks again, script works! [23:42] cjwatson: if a source package publishing record does not have a published date, does that typically mean it just hasn't been published yet, but will eventually? e.g. https://api.launchpad.net/devel/ubuntu/+archive/primary/+sourcepub/8444072 and a bunch of the stuff that's stuck right now. I wonder what our script should do for keeping up with the publisher. It seems like we should wait until we start [23:42] seeing published dates again? (we don't want to lose our "place" in the publisher's history) [23:42] rbasak: --^ fyi, there's a bug in my script until we fix that loop :) [23:44] correct - date_published is set when the publisher processes a given publishing history entry [23:44] it's normal for some recent entries to have date_published=None [23:45] yeah, i think once we've 'caught up', we need to just tread water in our watcher script [23:45] cjwatson: thanks! [23:50] (date_published is set when the status transitions from Pending to Published) [23:50] wgrant: ah also good [23:50] thanks to you both