[00:20] Unit193: i use the package import git for casper [00:21] The only thing I found current just had imported dscs, not development commits. :/ [00:31] Unit193: well yes [00:33] mwhudson: I'm trying to narrow down why certain ISOs I'm playing with drop me in the initramfs, d/changelog and Launchpad diffs are nice but sometimes actual commits have more reasoning and much more narrow changes. The importer is just as good as reviewing d/changelog and diffs. :/ [00:34] Unit193: yeah, it's exactly as good as that [00:34] Unit193: i don't know of anything more finegrained though [00:36] Eg https://git.launchpad.net/ubiquity/log/ is quite useful. Ah, OK. Well as long as it's "It doesn't exist" rather than "I can't find it", that'll do. Thanks, mwhudson. === ricab is now known as ricab|brb === ricab|brb is now known as ricab [11:43] hum, https://people.canonical.com/~ubuntu-archive/pending-sru.html hasn't been updated for more than a day ... who is the right person to ping about that? [11:44] I can look at that [11:44] this happened last week too... [11:45] yeah, stuck process [11:46] Does it use lots of launchpadlib? git-ubuntu tends to hang when Launchpad is struggling. I never got to the bottom of it. [11:47] Laney, thx [11:49] rbasak: Yeah, it's doing more or less what you'd expect to get a list of published packages in -proposed, check they've built, things like that - a lot of API usage [11:49] I can't see what it was doing [11:50] Not ideal that there's no monitoring [11:50] Maybe it could be run through 'timeout' at least === tmhoang1_ is now known as tmhoang1 === ricab_ is now known as ricab|lunch === ricab|lunch is now known as ricab [15:57] vorlon or tobikoch (found you in 'git log'): can you please comment on https://code.launchpad.net/~ken-vandine/ubuntu-seeds/+git/ubuntu/+merge/370062 ? (cc kenvandine) [15:57] I thought that livecd-rootfs was supposed to be able to figure this out (hence sil2100 removing that entry), BICPW [15:57] BW* [16:00] possibly something around this area https://git.launchpad.net/livecd-rootfs/tree/live-build/functions#n485 [16:09] Laney: the functionality is missing from livecd-rootfs in Xenial, afaict. [16:16] tobikoch: kenvandin_e said it was broken for eoan too (that MP is for the eoan seeds) [16:20] Laney: It should work in eoan. [17:17] Laney, tobikoch: does mvo's livecd-rootfs MP fix this? https://code.launchpad.net/~mvo/livecd-rootfs/+git/livecd-rootfs/+merge/370065 It seems to only add validation, so maybe that just turns the runtime failure into a debuggable build failure [17:17] vorlon: we should get that in I think [17:18] Laney: k. I'm off today, feel free to merge & upload or get someone from Foundations to do it [17:18] kenvandine is saying (#ubuntu-desktop) that core is always needed - snapd is saying "cannot proceed without seeding "core"" which certainly sounds fairly definitive [17:18] ('core' is missing from seed.yaml) [17:19] livecd-rootfs is added snapd if core isn't seeded [17:19] * Laney is off, please talk amongst yourselves :) [17:19] and according to the snapd folks, we still need core for the classic case [17:19] vorlon: ^^ [17:19] vorlon: mvo's MP only adds a check to prevent us from getting broken images [17:20] right, so we have to add core back into livecd-rootfs's handling to work around snapd not being ready for a core-free system (and now enforcing this at seed time instead of failing later) [17:20] yeah [17:21] might be better to seed core explicitly, which livecd-rootfs will see and not add snapd? [17:21] then we don't need to touch livecd-rootfs when we can have a core free system [17:21] just by changing the seeds [17:21] this way we can control this in the seed [17:21] * kenvandine prefers to be explicit [17:22] I think that's inconsisent with the design principle of seeds, which is generally that you declare the top level / leaf packages that you care about and expect the machinery to spit out the right result [17:23] yeah, but without a dependency system [17:23] we *don't* want the core snap as part of the definition of Ubuntu 19.10, but we have to pull it in because ugs [17:23] bugs [17:23] so I think the machinery should handle this, and not declare it in the seed [17:24] ok [17:24] mvo suggested they should really provide a tool for us to use in livecd-rootfs that takes the seed and figures out what needs to be added to pass validation [17:24] instead of having the logic in livecd-rootfs itself [17:25] that doesn't exist today, but would be nice to have in the future [17:28] vorlon: go back to enjoying your time off and I'll handle this [17:28] vorlon: thanks for your feedback [17:29] * vorlon waves :) [20:05] /query tobijk