[00:23] For the sake of py3, https://packages.qa.debian.org/c/calibre/news/20191025T011934Z.html might be great to grab, no? === not_phunyguy is now known as phunyguy [06:25] rafaeldtinoco: rbasak: I've had two MPs of which I said they are not important since they wait on Focal, now that it is open I'd appreciate a timely review if one of you could take a look [06:25] https://code.launchpad.net/~paelzer/ubuntu/+source/qemu/+git/qemu/+merge/374446 [06:25] https://code.launchpad.net/~paelzer/ubuntu/+source/qemu/+git/qemu/+merge/374447 [06:27] well, now that we know Focal is open and no other fixes will be stacked let me combine them into one ... [06:42] rbasak: rafaeldtinoco: https://code.launchpad.net/~paelzer/ubuntu/+source/qemu/+git/qemu/+merge/374770 would be better [06:42] and in a few minutes we also have proper up to date Focal test build builds in a PPA to try if you want [06:53] also I'm still waiting for all 4 haproxy MPs reachable from https://bugs.launchpad.net/ubuntu/+source/haproxy/+bug/1841936 [06:53] Launchpad bug 1841936 in haproxy (Ubuntu Focal) "Rebuild openssl 1.1.1 to pickup TLSv1.3 (bionic) and unbreak existing builds against 1.1.1 (dh key size)" [Medium,Triaged] [06:53] Those are enough MPs to also tag you ahasenack :-) [06:54] ^^ [07:11] rafaeldtinoco: the second and third chapter have a lot of overlap [07:11] not sure if you'd want to merge that into one [07:11] rafaeldtinoco: you also have to add a huge disclaimer at the manual qemu-cmdline that this disables everything that libvirt would do in terms of feature control [07:12] it is a (very) last resort effort to do it that way [07:12] also e.g. mds-no can be set normally, I ask you to only use those in the cmdline examples which really don't work any other way [07:13] page breaks before new chapters would also help readability [07:14] Oh you meant this to be for bionic, that is why you consider mds missing [07:14] still ahuge disclaimer please [07:17] IIRC while mds-no was one of the complex arch_capability ones [07:17] md-clear was a normal bit and can be used in Bionic [07:18] TBH I'd not even tell people to use qemu commandline from libvirt xml, it just leads to so many bugs later on all too often [07:18] but with a big disclaimer it might be ok [07:19] overll the doc surely will help people, so I hope you get not only mine but also other feedback and then can publish it somewhere === cpaelzer__ is now known as cpaelzer [09:46] thx. yep I need to split what can be done on each version, like the msd-no. and will sure look for feedbacks. Greg's presentation today was all about that, impressive.. [09:48] I'll start doing merges today, and code reviews as the doc is finished and repo is opened. [10:17] cpaelzer: I'm still cloning qemu :-/ [10:17] hehe [10:17] rbasak: rafaeldtinoco: glad to see you both, please coordinate that not both of you clean qemu for too long :-) [10:18] I'm getting an eventual 503 Service Unavailable trying to fetch your branch [10:18] So I might move on and see if it's working later [10:19] rbasak: maybe do haproxy and let qemu for rafaeldtinoco ? [10:19] that is smaller [10:19] and you are involved from a second POV of the SRU team [10:19] haproxy isn't cloning either right now [10:19] oh, so it is LP in general [10:20] I wonder if it's related to the archive opening [10:20] The git-ubuntu importer is presumably very busy [10:20] I'm starting little later today, will catch up and sync with what's left (leave some reviews for me :) [10:42] rbasak: The git-ubuntu importer presumably always uses SSH though? [10:42] Or does it use HTTPS to pull? [10:45] I'm not sure. I can look it up if you need? [10:45] Not especially [10:46] I should unpack what I said though: the 503s are because there's a lowish limit on concurrent HTTPS connections at the moment because there are some frequent events that amount to DDoS by HTTPS [10:46] And limiting concurrent connections makes it easier for the service to recover somewhat naturally most of the time without just faceplanting [10:47] That makes sense, thanks [10:51] cpaelzer: looks like haproxy for Focal needs rebasing onto proposed :-/ [10:51] Does that change anything? [10:51] no [10:51] rbasak: I already abandoned the focal MP [10:51] Ah [10:51] and updated the bug [10:52] the sync relaces the Focal MP, but SRU MPs stay as-is [10:52] +p [10:52] Sorry I shouldn't have jumped straight to your branches [10:57] cpaelzer: I don't see the fix in 2.0.8-1? src/ssl_sock.c appears unchanged. [10:57] I was relying on what I found it tagged in the upstream branch - let me check the actual upload that we got synced [10:58] $ git tag --contains d6de151248603b357565ae52fe92440e66c1177c [10:58] v2.0.8 [10:59] just to confirm why I have assumed it would be in there [10:59] Suddenly I get the opposite answer to my query [10:59] git diff pkg/ubuntu/eoan-devel pkg/ubuntu/focal-proposed -- src/ssl_sock.c [10:59] This was previously giving me nothing, but now I see it [10:59] * rbasak is puzzled [11:00] Ah [11:00] the fixes are in on my check to focal-proposed that I just did [11:00] Apparently working directory matters for that [11:00] My mistake, osrry [11:00] np [11:51] cpaelzer: haproxy (three branches) looks good. What's the plan with the bug reference and SRU information? [11:51] Right now the bug reference doesn't trigger a Launchpad-Bugs-Fixed [11:51] And the SRU information is for the previous no-change rebuild I think? [11:52] SRU information is for both [11:52] if you read the bug it has two templates now [11:52] and the upload will be done with an -v to the last released version [11:52] Ah [11:52] that way it will trigger the updates excactly once [11:52] OK [11:53] but atm autopkgtest has so much to do, I'm not sure when the 2.0.8 sync will release anyway [11:54] but then -proposed is enough (in terms of prereq) to get the SRUs uploaded [11:54] so if you approved them I can go on [11:55] rbasak: I agree that the later ones which already had openssl need updated bug links [11:55] consider that done [11:56] Well you just convinced me that with a -v in the upload you don't need them? :-P [11:56] which is correct for Bionic [11:56] but D/E never had the openssl upload [11:56] Ah [11:56] there your feedback was correct [11:57] I'm just proactively fixing your way to be green [11:59] and the 2.0.8 would already be released if the opening of the sync wouldn't put that much pressure on autopkgtest [11:59] things might need a while to settle there [11:59] I think it's OK for you to mark Fix Committed for Focal and for me to then accept the SRUs. [12:00] done [12:00] now uploading [12:03] rbasak: all tagged and uplaoded, the MPs are marked approved and the bug on Focal "Fix Committed" [12:03] thanks for the review [12:14] thanks for the accept rbasak, that way I can do the tests right away later today [12:15] giving it an hour or two to build and show up in proposed [12:17] Is the expected output of "dpkg --verify" mostly made up of '?' aside from the MD5 check? Docs say it is using RPM format which would normally be '.' for a pass - can I assume that '?' means "this check not done on Debian" ? [12:39] xnox you mind if i merge mdadm, or are you working on it [13:25] xnox: mind taking a look at the following MR whenever you have some time (https://code.launchpad.net/~rafaeldtinoco/ubuntu/+source/systemd/+git/systemd/+merge/374027), pls ? systemd-networkd related. === ricab is now known as ricab|lunch [14:21] Hey! === ricab|lunch is now known as ricab === Wryhder is now known as Lucas_Gray [15:37] ahasenack: postgresql-common 208 showed up, but needs a while to be syncable [15:37] as I said in my mail, maybe you could kick the sync at your EOD [15:37] then it should be available for a sync I'd hope [15:45] slashd: hey, have I been granted PPU for LibreOffice yet? I noticed I'm still not allowed to trigger autopkgtests against LibreOffice in a PPA [15:45] "You submitted an invalid request: You are not allowed to upload libreoffice or libreoffice to Ubuntu, thus you are not allowed to use this service." [15:47] marcustomlinson: IIRC that is an action that only core-dev's can do atm [15:47] I think it lacks an ACL finer than that [15:47] :( [15:49] marcustomlinson: I think cpaelzer is mistaken [15:49] bileto uses archive.checkUpload to determine whether the requester can upload [15:50] sounds promising cjwatson thanks [15:50] if you can upload I think you should be able to trigger tests [15:51] Indeed. I checked and marcustomlinson hasn't yet been granted any kind of PPU [15:51] cjwatson: that would make sense yeah. slashd was gonna do that for me, I just suspect he hasn't gotten around to it yet. [15:52] marcustomlinson, I thought it was sil2100 task, I'll get back to you soon. Sorry for the delay. [15:52] marcustomlinson, just share retry urls on #ubuntu-desktop meanwhile if needed (or sent in query, I'm happy to click on those for you) [15:52] It needs to be done by a member of the TB [15:53] slashd: ah sorry, you did the team additions, cyphermox was gonna do the PPU [15:54] thanks seb128, I don't actually need one at this moment, was just giving it a test [15:54] marcustomlinson, I'll figure this out with the other member of the dmb and get back to you soon. [15:54] thanks slashd, sorry for bugging you (the wrong person) [15:55] marcustomlinson, not problem at all, it's dmb matter so all good [15:55] Not really happy with testing access by trying to create unnecessary jobs [15:55] We are slammed at the minute and libreoffice isn't exactly trivial [15:56] Laney: well I don't NEED to test is what I'm saying, as I can for amd64 locally first [16:00] Great, that's what everyone should be doing :> [16:02] Also you can test for access using edit-acl (lp:ubuntu-archive-tools) [16:02] edit-acl -p marcustomlinson query [16:04] cjwatson: useful thanks :) [16:24] slashd: I thought only the PPU package addition for Gunnar was my action item? [16:24] Did I forget something? [16:24] uh! [16:27] sil2100, it's cyphermox's action [17:24] bryce: https://code.launchpad.net/~racb/usd-importer/+git/usd-importer/+ref/importer-add-tests [17:25] More refactoring and cleanups to come, but that should give you an idea of what's coming. [17:25] You'll want to spend some time understanding the structure [17:25] Diff against master please - the individual commits probably won't be helpful === jdstrand_ is now known as jdstrand [20:12] marcustomlinson: PPU is done. [20:39] That's what I tought but what would be the rationale behing installing the whole ubuntu desktop when apt installing xorg on a netinst? I mean wouldn't it be better to move those dependencies in "suggested packages" insteand of "recommended"? [20:40] Thanks but that's exactly what I tried to avoid since I'll need to install each packages I need individually. === Wryhder is now known as Lucas_Gray [22:32] There's no difference in the dependencies between netinst and other install methods; they're backed by the same archive. The solution suggested - `apt install xorg gnome-shell-` should get you xorg without gnome-shell. Check if that set of packages is what you're after. If not, repeat, excluding another likely-looking package with `$PACKAGE-` on the end, until the set of packages is [22:32] to your liking. [22:45] RAOF: Well, I'll try that solution even if there's a wall of packages to check. I guess I'll have some fun --dry-run'ing a few times until I find all the packages I need to exclude XD.