[00:19] <xnox> nacc, go nacc go! =)
[00:31] <nacc> xnox: heh
[00:50] <tsimonq2> xnox: Now I have to import your qtbase upload into VCS manually :/
[00:51] <tsimonq2> xnox: Also, block-proposed still belongs on bug 1749472 because mesa migrating will break Kubuntu and Lubuntu Next completely, so imho "because it's blocked by other things" isn't good enough :/
[00:53] <xnox> tsimonq2, i was meant to put the tag back in, post qtbase migrating
[00:54] <xnox> tsimonq2, unfortunately we do not have block-proposed tags that block only some packages =/
[00:54] <xnox> tsimonq2, thanks for puttin ghte tag back it.
[00:54] <xnox> *in
[00:55] <tsimonq2> xnox: I wonder if there's a way for me to implement syntax like block-proposed-qtbase-opensource-src
[00:55] <tsimonq2> xnox: In fact, I'll look into that now and see if it's as simple as adding a few lines of code
[00:57] <tsimonq2> xnox: Either way I'd appreciate if you could throw me a PR here merging your upload in: https://salsa.debian.org/qt-kde-team/qt/qtbase/tree/ubuntu+1
[00:57] <xnox> tsimonq2, well.... one could have two bugs; one bug per source package; with tags and say "blocks bug #" whatever in description......
[00:57] <xnox> lovely
[00:57] <xnox> i'm yet to do a PR on salsa /O\
[00:58] <tsimonq2> xnox: Right but I think that's far from ideal :)
[00:58] <tsimonq2> xnox: I did my first PR on Salsa yesterday, there should be a fork button somewhere on https://salsa.debian.org/qt-kde-team/qt/qtbase
[00:59] <tsimonq2> xnox: To be precise: https://salsa.debian.org/qt-kde-team/qt/qtbase/forks/new
[01:08] <jbicha> xnox: I believe all you have to do is mark a specific package as Fix Released for block-proposed to no longer apply to it
[01:12] <tsimonq2> jbicha: But it *isn't* Fix Released
[01:12] <tsimonq2> jbicha: imho that's *super* hacky
[01:16] <nacc> xnox: xevious: ok, just uploaded the last two php-horde*. I need to be afk tonight, but will retry what i see tmrw morning
[01:18] <rizonz> xnox: hi! from nacc I understood you are maybe the one buidling the janus package ?
[01:23] <rizonz> xnox: I have some issues with stopping the service, killing does not work and restarting has pidfile issues and never writes one on bionic
[01:24] <xnox> rizonz, literarly no idea what janus is
[01:25] <xnox> rizonz, i did a no change rebuild of it; as part of openssl1.1 transition, but that was 100% automated mechanical bot upload.
[01:25] <xnox> rizonz, file a bug about it?
[01:25] <xnox> rizonz, on launchpad, and maybe on Debian BTS too.
[01:25] <rizonz> xnox: https://launchpad.net/ubuntu/+source/janus/0.2.6-1build1
[01:25] <rizonz> ah ok!
[01:25] <xnox> rizonz, as, I said, that was a blind upload. the source package is unmodified from debian.
[01:25] <rizonz> your name was there
[01:25] <rizonz> okay!
[01:26] <rizonz> then I ask them
[01:26] <rizonz> xnox: sorry to bother you :)
[01:26] <xnox> no problem
[01:27] <rizonz> heh, you make me hungry btw, you remind me about unox
[06:29] <coolfish> Hi, in bug #1749790 a bigger autopkgtest-VM is needed to move the ganeti package from bionic-proposed to release (pass the test). Is there someone willing to set a bigger VM?
[09:08] <Unit193> nmap (7.60-1ubuntu4) bionic; urgency=medium
[09:08] <Unit193>   * debian/control: Add dependency menu for zenmap. (LP #126331)
[09:08] <Unit193> Not exactly ideal..
[09:46] <sil2100> rbasak: hey! Just so you know, I will be doing and approving some uploads to xenial today as part of preparations for the point-release - just letting you know so that you don't get confused that I'm doing an SRU shift today
[09:46] <sil2100> I will not get in your way with normal SRUs
[09:56] <rbasak> sil2100: OK. Thank you for syncing with me!
[11:26] <rbasak> jamespage: in bug 1724622, shouldn't there be a test output summary attached to the bug?
[11:27] <jamespage> rbasak: uh yes you are quite correct - juggling to many things
[11:27] <jamespage> lemme dig those out
[11:54] <rbasak> bdmurray: should ubuntu-release-upgrader in the Xenial SRU queue have a bug reference? Or is it special somehow?
[14:10] <coolfish> Laney: do you remember me, asking for a bigger autopkgtest-VM for package ganeti? It seems ganeti just got default 1536MB RAM. In the mean time I opened bug #1749790. Could you please have a look at it?
[14:18] <tsimonq2> xnox: Did you get a chance to look at proposing a PR to qtbase on Salsa quite yet?
[14:18] <xnox> let's do this now.
[14:19] <tsimonq2> OK
[14:22] <xnox> tsimonq2, ehhhm that repository is crazy
[14:24] <xnox> tsimonq2, it appears to be missing 5.9.3+dfsg-0ubuntu1 5.9.3+dfsg-0ubuntu2 5.9.3+dfsg-0ubuntu3 5.9.3+dfsg-0ubuntu4 uploads.
[14:24] <xnox> tsimonq2, what would you like me to do? rebase the ssl patch against ubuntu+1 branch, aka 5.9.4 tree?
[14:24] <tsimonq2> Please
[14:25] <xnox> tsimonq2, but what about other changes? =/
[14:26] <tsimonq2> Merge those in if possible please
[14:26] <xnox> tsimonq2, but it's like your upload =/ did you not push latest to salsa?
[14:32] <xnox> tsimonq2, that repository is very weird usage of git.
[14:34] <jamespage> rbasak: pasted my test results as you pasted or IRC conversation :-)
[14:37] <Laney> coolfish: I messed up and didn't put it in one of the required files, fixing now
[14:39] <xnox> tsimonq2, test building
[14:39] <rbasak> jamespage: :)
[14:39]  * rbasak looks
[14:48] <Laney> coolfish: See if the next lot of tests for 2.16.0~rc2-1 are any better
[14:55] <acheronuk> xnox: message from tsimonq2 via Telegram. He apologises, but he had to "abruptly leave for class".
[14:59] <xnox> acheronuk, tsimonq2 - no worries, i should have something proposed on salsa soon.
[15:01] <xnox> tsimonq2, https://salsa.debian.org/qt-kde-team/qt/qtbase/merge_requests/1
[15:01] <xnox> no idea if this is right, or wrong.
[15:01] <xnox> it builds
[15:16] <mitya57> xnox: tsimonq2 wrote me in Telegram: “Please pass on to xnox that I had to abruptly leave for class”
[15:16] <mitya57> oh, acheronuk already told the same :P
[15:16]  * mitya57 is slow
[15:17] <mitya57> I can look at your merge request though
[15:17] <acheronuk> :)
[15:18] <tsimonq2> Boy oh boy I want to graduate already
[15:20] <tsimonq2> xnox, acheronuk, mitya57: Thanks,
[15:20] <acheronuk> tsimonq2: sadly, won't the 1st time you wish that. high school -> college -> more quals :P
[15:20] <acheronuk> *last time
[15:21]  * mitya57 commented on xnox's merge request
[15:26]  * xnox struggles to find notification of any sort from salsa
[15:26]  * acheronuk suggests tsimonq2 does a few circuits to relax
[15:26] <xevious> nacc: Updated the gist a little while ago... https://gist.github.com/iammattcoleman/88013cb5f92105b15a66ee2ada442a16#file-2018-02-19_124034_php-defaults-txt-L171
[15:26] <xevious> nacc: Whenever you're ready, let's split up the remaining work.
[15:31] <tsimonq2> grr acheronuk :P
[15:34] <coolfish> Laney: thanks a lot, I'll watch the autopkgtest status.
[15:39] <coolfish> Laney: regarding the version: our understanding is, that ubuntu syncs with debain-testing/buster (ganeti-2.15), ganeti-2.16rc2 is in debian-unstable/sid and AFAIK has open regressions.
[15:41] <bdmurray> rbasak: Its special because the mirror file is just being updated. See https://launchpad.net/ubuntu/+source/python-apt/1.1.0~beta1ubuntu0.16.04.1 Maybe we should document that.
[15:44] <rbasak> Ah. I had assumed that there's always a tracking bug, but looking at tzdata that isn't always the case.
[15:45] <rbasak> bdmurray: how should I approach reviewing this?
[15:45] <rbasak> Just a sanity check and accept?
[15:45] <rbasak> And does sru-review work as normal?
[15:46]  * rbasak goes afk for a bit
[15:50] <bdmurray> rbasak: I'd check with sil2100 since he's leading the 16.04 point release
[15:52] <sil2100> rbasak: yeah, leave that one for me, I'm not including tracking bugs for releasey things (if it doesn't make sense at least)
[15:52] <sil2100> rbasak: I checked all the ADT failures and will be taking care of releasing it before the candidates are being created
[15:52] <sil2100> After some sanity checking
[15:53] <sil2100> There will be a few other uploads like that
[16:05] <rbasak> ack
[16:12] <nacc> xevious: i'm back now
[16:12] <xevious> nacc: Good morning!
[16:13] <xevious> nacc: Which packages are you working on? I'll take a handful that you're not currently looking at and add them to LP#1750041.
[16:14] <xevious> ...and, ya know, work on them.
[16:15] <xevious> Or, is there one in particular you'd like me to work on?
[16:15] <nacc> xevious: the ones i don't know how to fix are php-horde-kolab-storage, php-horde-kronolith, php-horde-sessionhandler,
[16:16] <nacc> xevious: all regressed because of php7.2 changes in how session works, i think
[16:16] <nacc> xevious: it's the 'headers sent after...' messages
[16:16] <nacc> xevious: it seems like in php <7.2, it was silently doing nothing
[16:16] <nacc> now it is noisily warning
[16:19] <nacc> xevious: i'll retry the others now
[16:32] <nacc> xevious: would you be able to investigate those? or even a few of them (maybe just sessionhandler to start, as that is the root cause of the regression in kronolith, iirc)
[16:32] <nacc> xevious: i need to do some MIR work today, but can pivot back to php in a few hours, i think
[16:33] <nacc> well, that's a new one (running an autopkgtest.u.c retrigger): "A server error occurred.  Please contact the administrator."
[16:33] <nacc> Laney: --^ would that be you?
[16:35] <xevious> nacc: Yeah, I'll start with sessionhandler.
[16:35] <nacc> xevious: thanks!
[16:35] <nacc> xevious: iirc, the tests actually aren't failing
[16:36] <nacc> but since php emits a warning, and stderr messages aren't allowed by default, the tests fail
[16:36] <nacc> so there are two solutions, 1) silence the warnings or 2) fix the warnings :)
[16:36] <nacc> i'd prefer the latter, but i couldn't figure them out
[16:39] <nacc> slangasek: would you know who the administrator of autopkgtest.ubuntu.com is?
[16:42] <slangasek> nacc: present
[16:42] <slangasek> nacc: (Laney and myself)
[16:43] <Laney> coolfish: We sync from unstable
[16:43] <Laney> nacc: link?
[16:44] <nacc> slangasek: Laney: https://autopkgtest.ubuntu.com/request.cgi?release=bionic&arch=amd64&package=php-horde-serialize&trigger=php-defaults%2F60ubuntu1&all-proposed=1
[16:45] <nacc> i've not tried any others, since i don't want to just hammer the server with failures :)
[16:45] <nacc> it was working fine for the first 3 or 4 packages I ran against
[16:45] <nacc> (i'm working through the php-defaults queue that i fixed last night)
[16:45] <slangasek> hmph
[16:45] <slangasek> Laney: I'm on calls right now; do you have time to dig into this?
[16:48] <Laney> yeah
[16:49] <Laney> can't connect to the amqp server seems to be the problem
[16:49] <Laney> :(
[16:51] <xevious> nacc: Do you already make an LP issue for sessionhandler?
[16:51] <xevious> *Did
[16:52] <Laney> nacc: seems to be some maintenance going slightly awry
[16:52] <nacc> xevious: no, you can just add tasks to the php-defaults bug, i think, if you want
[16:53] <nacc> Laney: ok, thanks -- I'll hold off on further triggers for now, but others will probably hit it :)
[16:54] <Laney> nacc: not a lot we can do, and it won't really cause a problem other than irritating you
[16:54] <Laney> assuming britney is doing something sane like not thinking it has submitted requests ...
[16:55] <nacc> Laney: ack -- do you have any idea when I should retry again? I know these will all pass once retriggered, just need to through them so we can get php migrated
[16:55] <Laney> crashing
[16:55] <Laney> that'll do
[16:55] <nacc> heh
[16:56] <Laney> nacc: dunno, I'm watching IS people try to fix it
[16:56] <nacc> Laney: ah ok :)
[16:56] <Laney> ah
[16:56] <Laney> works now I think
[16:56] <Laney> nacc: just clicked that link for you, feel free to run others
[16:56] <nacc> Laney: thanks!
[16:57] <nacc> Laney: yep, working again
[16:57] <Laney> networking is useful, it turns out
[17:05] <nacc> Laney: slangasek: thank you to both
[17:07] <nacc> xevious: fwiw, i was reading other upstream code's changes to deal with the same messages with php7.2
[17:08] <nacc> xevious: one of them had a commit message like 'do not call ini_set before session_start'
[17:09] <nacc> Laney: hrm, i submitting a bunch of retriggers now, and none have showed up in http://autopkgtest.ubuntu.com/running .. usually they show up relatively quickly
[17:09] <nacc> even if in the backlog
[17:09] <nacc> "a bunch" = 30-40
[17:18] <Laney> nacc: I can see a handful of tests being run by you there
[17:19] <Laney> are there some missing? if so, which package/arch?
[17:23] <Laney> (probably the /running page got a bit confused when its amqp connection went away)
[17:27] <xevious> Laney: Which AMQP server are you using?
[17:29] <Laney> rabbitmq
[17:56] <nacc> Laney: sorry, let me see
[17:59] <Laney> nacc: I've got https://paste.ubuntu.com/p/srVp6vtxg5/
[18:00] <nacc> Laney: ok, that looks fine
[18:00] <nacc> Laney: i think I was just impatient :)
[18:05] <nacc> bdmurray: can you tell why LP: #1750842 was marked private?
[18:05] <nacc> (the submitting user is asking)
[18:05] <nacc> xevious: any luck?
[18:06] <xevious> Still debugging. This is a doozy.
[18:06] <nacc> xevious: yeah
[18:06] <nacc> xevious: i got fairly deep into the package code and hte upstream bug reports about this change
[18:06] <bdmurray> nacc: I can't see it with my lp account. Can you give me some more details?
[18:06] <nacc> and then decided i really didn't know how to fix it :)
[18:14] <nacc> bdmurray: i can subscribe you if that would help (the user subscribed me)
[18:14] <nacc> bdmurray: apport filed it for them
[18:14] <nacc> bdmurray: but they don't know why it's marked private
[18:15] <bdmurray> nacc: is it tagged needs-amd64-retrace?
[18:15] <nacc> bdmurray: it was, but is no longer
[18:15] <nacc> bdmurray: does that make it private?
[18:17] <bdmurray> nacc: It starts of as private w/ only the retracer subscribed, then after retracing the retracer leaves it private and subscribes ubuntu-bugcontrol. Its stays private because the Stacktrace might have sensitive information.
[18:17] <bdmurray> nacc: Or maybe not bugcontrol but some intermediate team.
[18:19] <bdmurray> nacc: yeah - crash bug triagers for universe or main should be subscribed by apport
[18:20] <bdmurray> Well its really only one team but still named ubuntu-crashes-universe
[19:24] <nacc> bdmurray: ah ok
[19:25] <nacc> bdmurray: thanks for the explanation
[19:39] <nacc> bdmurray: in this particular case, i don't see any team subscription being added after the retracer tag was removed?
[19:52] <xevious> nacc: PHPUnit's output is breaking the test. `headers_sent()` returns FALSE before this line and TRUE after it: https://github.com/sebastianbergmann/phpunit/blob/master/src/TextUI/TestRunner.php#L283-L285
[19:53] <xevious> Still working on a fix, but I at least tracked down the line where we go from working to broken.
[19:54] <nacc> xevious: weird, but is that really a header send to PHP?
[19:54] <xevious> As soon as any output occurs, PHP considers the headers sent.
[19:54] <xevious> ...since you can't change HTTP headers after you start sending the page content.
[19:55] <nacc> xevious: ah interesting
[19:55] <nacc> xevious: but PHPUnit has always done this, afaict
[19:56] <nacc> xevious: or, i guess i would think phpunit would be exec'ing php to run the test, so it'd be a new session?
[19:57] <nacc> dunno exactly waht i mean, but it's weird to me that the runner's env would affect the runnee's
[19:58] <xevious> Yeah, I agree. Still trying to figure out *why* that's breaking the test now.
[19:58] <nacc> xevious: if you are able to get sessionhandler working, i hav ethe kronolith fixes for phpunit6 locallyu, and i can test against your updated version
[20:04] <bdmurray> nacc: well subscribe me and I'll have alook
[20:24] <nacc> bdmurray: done
[20:25] <bdmurray> nacc: Oh, I think there was an LP issue overnight
[20:28] <nacc> bdmurray: ah ok
[20:29] <bdmurray> nacc: That doesn't seem to be it though so I'll look into more
[20:31] <nacc> bdmurray: thanks, i'll let them know
[20:31] <nacc> xevious: any luck?
[20:32] <nacc> xevious: i wonder if this is related? https://github.com/yiisoft/yii2/pull/14959
[20:33] <xevious> I think I found the fix
[20:33] <xevious> Give me a few...
[20:33] <nacc> xevious: per the comment there, do we need to add an output buffer?
[20:33] <nacc> xevious: ok cool!
[20:35] <nacc> xevious: it seems like a bunch of projects have added freeze/unfreeze logic upstream?
[20:49] <xevious> nacc: output buffering didn't help. I got it passing but it's still complaining about the headers being sent.
[20:50] <xevious> All its assertions pass, though.
[20:52] <xevious> nacc: Which LP issue should I upload the patch to?
[20:53] <xevious> n/m found it
[20:54] <xevious> nacc: This issue is about removing packages unused, is there a separate issue about PHP 7.2 compatibility? https://bugs.launchpad.net/ubuntu/+source/php-phpdocumentor-reflection/+bug/1749783
[20:59] <nacc> xevious: it's about anythiing blocking php-defaults (per the title)
[20:59] <nacc> xevious: so it's fine to attach there
[20:59] <xevious> Ok
[21:04] <xevious> I've built up a strong distaste for Horde users.
[21:07] <nacc> xevious: :)
[21:07] <nacc> xevious: fwiw, a bunch of stuff should start migrating from proposed after i retriggered something like 400 tests :)
[21:11] <xevious> Splendid
[21:12] <nacc> xevious: and a bunch more are wedged because there are still some horde packages i need to hit
[21:12] <nacc> i'm d/ling those now
[21:21] <nacc> xevious: i don't have my mail open right now (having to build 20 or so more horde packages). let me know when you attach the patch and i'll take a look
[21:22] <xevious> Will do. I got the tests to pass while inside the adt shell after one of the failures. However, when I applied those changes to the package and retested, a new issue cropped up: dealing with it now.
[21:23] <nacc> xevious: thanks
[21:48] <xevious> Woops
[21:50] <xevious> nacc: https://bugs.launchpad.net/ubuntu/+source/php-horde-sessionhandler/+bug/1749783/comments/16
[21:51] <xevious> nacc: I don't feel great about that patch because I supressed a few warnings to eliminate stderr output, but all of the assertions in the unit tests pass.
[22:02] <nacc> xevious: do you want to ping the upstream about it? i know it's just in the tests, so the underlying code is unchanged, i just wonder if they would take it
[22:06] <xevious> A large part of the fix came from them.
[22:06] <xevious> Hold on though
[22:06] <xevious> I think I might have figured out how to avoid supressing the warnings.
[22:09] <xevious> nacc: Here's the majority of the fix: https://github.com/horde/SessionHandler/commit/d1a72d266117894be11f10645486cb13c7b1b40e
[22:10] <nacc> xevious: oh great!
[22:10] <nacc> xevious: can you add dep3 headers for that?
[22:10] <nacc> (cf what `dpkg-source --commit` does)
[22:10] <xevious> Some of my changes are definitely upstreamed.
[22:11] <xevious> ..and some of the other changes are upstreamable.
[22:11] <xevious> Let me try out this change, then we'll circle back to dep3 headers.
[22:15] <nacc> xevious: thanks! it just lets us track where things come from and such
[22:51] <xevious> nacc: Ok, I have a better patch that doesn't require supressing warnings.
[22:51] <xevious> How do I do the dep3 stuff?
[22:54] <nacc> xevious: so i'm looking for something like
[22:54] <nacc> http://dep.debian.net/deps/dep3/
[22:54] <nacc> specifically the Description, Origin, Author fields
[22:55] <nacc> xevious: you can then use dep3changelog to genrate the changelog entry using that data :)
[23:03] <nacc> xevious: let me know if you need more
[23:03] <xevious> I'm splitting it up into two patches: one to pull in the file from upstream, and one with my changes to that file.
[23:04] <nacc> xevious: ah great, that's a good way of doing it
[23:04] <xevious> After reading the Debian documentation, it seemed like the right thing to do.
[23:05] <xevious> Should I put myself or the upstream committer as the author for the first patch?
[23:05] <nacc> xevious: if it's fully from upstream, i tend to use the upstream commitere/author
[23:15] <nacc> xevious: is that working ok for you?
[23:16] <xevious> Yeah, just making the debdiff now
[23:16] <xevious> My local test has to finish, too.
[23:17] <nacc> xevious: nice
[23:23] <xevious> nacc: I've been keeping track of all my changes and plan on at least attempting to upstream them all. Some of the projects specifically say they're a PHP5 library, so I don't know if they'll accept PRs that make them PHP7-only.
[23:25] <nacc> xevious: yeah, it's worth doing still, even then
[23:26] <xevious> nacc: Here you go: https://bugs.launchpad.net/ubuntu/+source/php-horde-sessionhandler/+bug/1749783/comments/17
[23:26] <nacc> xevious: reviewing now
[23:27] <xevious> Thanks. I've got to head out to dinner soon, but I'd like to make sure it's good before I leave.
[23:28] <nacc> xevious: yep, pulling it down now the patch looks good, i'll make sure it builds and passes locally
[23:29] <nacc> and i'll test and push the dependent packages if it does
[23:29] <nacc> xevious: fixed the LP entry (it needs to be LP:  to be seen)
[23:31] <xevious> Oops! Thanks for pointing that out.
[23:33] <nacc> xevious: np
[23:34] <nacc> xevious: looks good
[23:34] <nacc> xevious: i'll test the dependent packages and then upload
[23:34] <xevious> Excellent.
[23:35] <xevious> I'm heading out for the night. We'll see what's left, if anything, tomorrow.
[23:36] <nacc> xevious: thanks again for all your help!
[23:40] <nacc> xevious: ok, that didnt' solve the two related packages (afaict)
[23:40] <nacc> xevious: definitely kronolith is related
[23:40] <nacc> https://paste.ubuntu.com/p/CvSQBvRVf6/
[23:41] <nacc> and for php-horde-kolab-storage with my local changes, i get: https://paste.ubuntu.com/p/ztZpBqrPbW/
[23:42] <nacc> xevious: i'll update the bug, because it hink sessionhandler itself needs changes in order to work
[23:45] <nacc> xevious: the reason i think the kolab-storage stuff is related, is that synchronizeList uses the session, i believe