| [06:42] <doko> sergiodj: yes, please drop the Python2 stuff again | |
| [09:50] <santa_> rbalint: hi, I got a crash in unattended-upgrades and I think I have a patch, if you have some time I would like to discuss how to get it fixed in ubuntu | |
| [09:51] <rbalint> santa_, hi, sure, how is it crashing? | |
| [09:51] <santa_> rbalint: https://bugs.launchpad.net/ubuntu/+source/unattended-upgrades/+bug/1883082 | |
| [09:51] <ubottu> Launchpad bug 1883082 in unattended-upgrades (Ubuntu) "Crash when using Package-Whitelist-Strict" [Undecided,New] | |
| [09:51] <santa_> just reported a min ago | |
| [09:51] <rbalint> santa_, that's fresh :-) | |
| [09:51] <santa_> it always crashes with strict whitelist | |
| [09:52] <santa_> I think I know the cause and I have a possible patch for this | |
| [09:52] <rbalint> santa_, ok, please attach it to the bug | |
| [09:53] <santa_> sure | |
| [10:00] <rbalint> santa_, thanks, the fix seems simple and if you attach it i'm including it in the next upload + sru and give the credit since you found the fix first :-) | |
| [10:01] <santa_> rbalint: ok, that was what I was going to ask - SRU for focal; what would be the version you are going to use? | |
| [10:02] <santa_> because I'm going to build a custom package in a PPA ;) | |
| [10:02] <santa_> would be 2.3-0ubuntu0.1 ? | |
| [10:03] <santa_> 2.3-0ubuntu0.1 for focal SRU I mean | |
| [10:03] <rbalint> 2.3.0ubuntu0.1 i think, since it is a native package | |
| [10:04] <rbalint> santa_, out of curiosity why do you use strict whitelist? | |
| [10:05] <santa_> rbalint: I'm working for a company which requested this | |
| [10:05] <santa_> we used to have a home-made script to do this, but I'm rewrking the thing, if possible, with u-a | |
| [10:06] <santa_> * reworking | |
| [10:07] <santa_> rbalint: the thing needed is having just a small set of packages being upgraded automatically (firefox, openssh server, ...) and the rest manually if I'm not mistaken | |
| [10:08] <rbalint> santa_, thanks, make sense, i'd just would like to understand how u-u is being used in the wild | |
| [10:09] <santa_> sure np | |
| [10:21] <Laney> juliank: on a different note ;-) | |
| [10:21] <Laney> can you help me quickly to make https://salsa.debian.org/release-team/britney2/-/blob/master/britney2/policies/policy.py#L1032 work with xenial's python3-apt? | |
| [10:21] <Laney> looks like that has changed signature since | |
| [10:37] <juliank> Laney: you're backporting? | |
| [10:37] <Laney> juliank: nope, trying to run britney on snakefruit which is xenial | |
| [10:37] <juliank> Laney: The architecture option does not exist on xenial | |
| [10:38] <juliank> Laney: So I'm not sure this can be run on xenial, because that does the [...] architecture filtering | |
| [10:39] <Laney> hmm | |
| [10:39] <juliank> you might have to implement architecture wildcards yourself and pre-mangle input | |
| [10:39] <Laney> was hoping for a way to do it anyway :P | |
| [10:39] <juliank> Laney: Ah | |
| [10:39] <juliank> Laney: You can set APT::Architecture before calling it on xenail | |
| [10:40] <juliank> Laney: But like restore it afterwards | |
| [10:41] <juliank> native_arch = apt_pkg.config["APT::Architecture"] | |
| [10:41] <juliank> apt_pkg.config["APT::Architecture"] = arch | |
| [10:41] <juliank> parse_src_depends(block, False) | |
| [10:41] <juliank> apt_pkg.config["APT::Architecture"] = native_arch | |
| [10:42] <Laney> ah, that's nice if it works, let me try it | |
| [10:42] <Laney> in an except:, maybe upstream will take it | |
| [10:42] <Laney> silly ancient machines | |
| [10:54] <Laney> neat, seems to be working | |
| [10:54] <Laney> merci juliank | |
| [11:33] <paride> Hi! Groovy live-server ISO images are successfully being built for all the 4 archs: https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/groovy/ubuntu-server-live | |
| [11:33] <paride> but the ppc64 image is not published here: http://cdimage.ubuntu.com/ubuntu-server/daily-live/pending/ | |
| [11:34] <paride> It use to be up to May 28. | |
| [11:35] <paride> any idea of what could be wrong? | |
| [11:35] <cjwatson> https://people.canonical.com/~ubuntu-archive/cd-build-logs/ubuntu-server/groovy/daily-live-20200611.log | |
| [11:35] <cjwatson> absurdly hard to read but I think the relevant error is: | |
| [11:35] <cjwatson> xorriso : FAILURE : -as mkisofs: Unsupported option '--netatalk' | |
| [11:36] <cjwatson> I think xnox was working on a xorriso upgrade at some point? | |
| [11:37] <paride> yeah I saw his branch | |
| [11:40] <paride> https://code.launchpad.net/~xnox/debian-cd/use-xorriso/+merge/384480 | |
| [11:40] <paride> the date of the latest comments matches the last day ppc64 images were produced | |
| [12:25] <paride> cjwatson, thanks for pointing me to the right point in the build log. fiy this is the problem: https://bugs.launchpad.net/ubuntu-cdimage/+bug/1883103 | |
| [12:25] <ubottu> Launchpad bug 1883103 in Ubuntu CD Images "Switch to xorriso causes failure in ppc64 image generation" [Undecided,New] | |
| [12:25] <paride> xnox, vorlon, ^^ | |
| [12:26] * cjwatson nods | |
| === ben_r_ is now known as ben_r | |
| [13:46] <xnox> tah | |
| [13:47] <xnox> cjwatson: paride: i am in progress redoing ppc64el using cd-boot-images. I'll try to push a branch for that asap then. | |
| [13:51] <paride> xnox, which does not use --netatalk | |
| [13:52] <paride> so maybe it's not needed after all | |
| [14:05] <paride> could it be a legacy thing needed to support Apple PPC hardware? | |
| [14:19] <xnox> it is | |
| [14:20] <xnox> the new cd-boot-images for ppc64el just has a couple of files and that's it, without any magic options | |
| [14:20] <xnox> and i did test it locally and it does boot | |
| [15:02] <santa_> rbalint: I've just attached a patch for a u-a SRU, the first one was cheesy so please ignore it ;) | |
| [15:03] <santa_> I plan to use it in the next few days, let's hope it works as expected | |
| [15:04] <Laney> xnox: oh yay for submodules | |
| [15:04] * Laney grinds teeth | |
| [15:04] <Laney> :) | |
| [15:11] <xnox> Laney: ooooooh is that good or bad? | |
| [15:12] <Laney> xnox: hahah well I've not enjoyed using them up to now | |
| [15:12] <Laney> but still, thanks for the work | |
| [15:12] <cjwatson> Better than what was there before, probably worse than subtrees :) | |
| [15:12] * xnox played a lot for with --prefix subtree & submodules | |
| [15:12] <Laney> did you fork all of the repos into Launchpad or something | |
| [15:12] <cjwatson> The main problem with submodules is that the UI is awful even for git | |
| [15:12] <Laney> sorry for asking instead of checking | |
| [15:13] <xnox> cjwatson: so subtree wouldn't allow to rebase things nicely. Unless something like git-dpm to do the "rebase, but actually git merge" | |
| [15:13] <xnox> with subtree integration. | |
| [15:13] <cjwatson> And likes to sometimes leave submodules mysteriously out of date | |
| [15:13] <xnox> Laney: cjwatson: yes, i did setup git->git for all of the d-i projects, and then added ~ubuntu-installer team repos based off that, with ubuntu delta, on top of the right debian tag. | |
| [15:14] <xnox> Laney: cjwatson: so i can setup a "subtree" based git repo with all of those things too, I guess for evaluation. | |
| [15:14] <cjwatson> It can absolutely be handled, it's just one of those things where you need to pay closer attention than average at the client end | |
| [15:15] <cjwatson> (Also doesn't work with recipes and it's not clear how to make it work, though you may or may not care about that - see https://code.launchpad.net/~tintou/git-build-recipe/+git/git-build-recipe-1/+merge/351699) | |
| [16:49] <xnox> right, in the worst case scenario of subtree "rebase" one just removes a subtree and merges in a new one. | |
| [17:08] <bdmurray> vorlon: looking at https://wiki.ubuntu.com/PlusOneMaintenanceTeam/Status and "Things to pick from" nautilus-python has migrated right? I'm wondering how up to date this is | |
| [18:09] <vorlon> bdmurray: only as up-to-date as the last +1 maint person who updated it :) | |
| [19:37] <arnatious> I'm trying to figure out how to make this SRU: flake8 released an update supporting python3.8 only after we released focal. Should I just make the issue "This valid python code is failing <insert 3.8 language feature>, flake8 3.8 addresses this?" | |
| [19:58] <Odd_Bloke> arnatious: I'm not 100% sure I understand the question; could you explain the issue in a little more detail? | |
| [19:59] <arnatious> Yeah - the python3-flake8 package doesn't support all of python 3.8 as-is | |
| [20:00] <arnatious> Odd_Bloke flake8 released the update adding 3.8 support in early May, after we released focal. | |
| [20:01] <arnatious> We're on python3-flake8 3.7.9, PyPi has 3.8.x now. flake8 is supposed to keep minor version in sync with python | |
| [20:01] <Odd_Bloke> Aha, right, I hadn't understood you were talking about SRUing flake8 itself, thanks for the clarification! Do you have a bug for this? | |
| [20:02] <arnatious> I'm filing one, I was wondering if saying ^ that would be enough for a bug report | |
| [20:06] <Odd_Bloke> I'm not a member of the SRU team, so I'm not the best person to be answering that specifically I'm afraid. | |
| === lfaraone_ is now known as lfaraone | |
| [23:32] <RAOF> arnatious: I *am* a member of the SRU team, and https://wiki.ubuntu.com/StableReleaseUpdates#SRU_Bug_Template is a good start. | |