/srv/irclogs.ubuntu.com/2018/08/30/#ubuntu-devel.txt

* tsimonq2 is working on a small little project...00:04
tsimonq2https://git.launchpad.net/~ubuntu-dev/+git/pm-stats/00:04
tsimonq2https://pmstats.lubuntu.me/00:04
tsimonq2So far I consider my code "hacky and horrible" because it apparently decides using file storage is a good idea. :P00:04
xnoxtsimonq2, hmm.... we do have private internal dashboard on that.... it's done with grafana dashboard + prometheus backend.02:57
xnoxslangasek, can we publish any parts of our dashboard publically?02:58
xnoxor like mirror/sync it?02:58
tsimonq2xnox: Don't worry, this didn't take too long. ;)03:24
tsimonq2xnox: If that can be publicised, cool, but if it can't I intend on writing my own clone that is. :P03:25
xnoxtsimonq2, well, it all lives in a private instance. most things are based on public data which is crapped. but there are also internal-infra stats which are non-public.03:26
xnox(ie. monitors health of internal infra, our managed mirrors, etc.)03:27
xnoxnot sure if it can be made public, due to possible data leaks03:27
xnoxfoundations piggy-backed on that infra, cause it was nice and easy.03:27
Unit193Sure, but private doesn't really help us, and gives you a slight edge on us. :>03:27
tsimonq2xnox: Is there any hope of mirroring on a separate instance?03:27
tsimonq2Right, I want to crunch some of my own numbers.03:27
xnoxno idea.03:28
tsimonq2xnox: So who does? :P03:28
xnoxi mean it has the p-m stats, which is like one graph. nothing else is of use publically i don't think.03:28
tsimonq2That's all I care about, the public stuff. ;)03:29
tsimonq2xnox: Specifically what I have in mind is statistics similar to http://people.ubuntuwire.org/~stefanor/ubuntu-activity/ but with more -proposed-focused data points like why things are blocked, how quick they're resolved, a per-person breakdown of packages, etc.03:33
xnoxyeah, that would be nice.03:33
xnoxwe have none of that.03:33
tsimonq2So, since the data is public, would you be able to give me/the channel a screenshot or *something* showing only the public data?03:34
tsimonq2It would be cool to see what exactly you get to look at. :)03:34
xnoxomg, i'm such a slacker, i'm not in top 10, i'm #1103:35
tsimonq2hehe03:35
xnoxjamespage has been doing too many uploads =)))))03:35
tsimonq2Unfortunately there's two major bugs (well three, if you include the fact that it uses poor web design); people like LocutusOfBorg have "Gianfranco Costamanga" and "LocutusOfBorg" in separate entries, and the "who works for Canonical?" graph is not only not robust, the actual data for that hasn't been updated in years.03:37
tsimonq2Last time I did a local run, most uploads are done by Canonical employees, pretty much always.03:37
xnoxwho works for canonical is very hard to track. and that info is not public really.03:37
xnoxalso too much churn in people leaving, rejoining, working on non-work stuff with work emails, and vice versa.03:38
tsimonq2How it guesses is based on members of ~not-canonical and people who use canonical.com email addresses in debian/changelog.03:38
tsimonq2So I know it can't always be accurate, but "somewhat accurate" is better than nothing.03:38
tsimonq2I've separately written at least three scripts to collect some or all of this data individually, for my own curiosity.03:39
tsimonq2Here's one that totals uploads per person in -proposed: http://paste.ubuntu.com/p/2rHxrnQSxP/03:40
tsimonq2So I have all of this code lying around in different places, I just want to finally consolidate it and make it into something useful. ;)03:40
slangasekxnox: I would like those grafana bits to be public but it would involve work03:53
mwhudsonhey um i'm working on some of the same things currently...04:14
tsimonq2mwhudson: Got links to your own scripts?04:35
tsimonq2We can probably take the best parts of both of our scripts and merge, or something...04:35
tsimonq2I doubt you're the only one who also has scripts...04:36
abeatosil2100, thanks, will repush shortly06:38
mwhudsontsimonq2: i haven't started working on them at all yet07:35
tsimonq2mwhudson: ah07:39
mwhudsonthere is a trello card with my face on it :)07:39
mwhudsoni see your stuff isn't exactly super sophisticated either :)07:41
tsimonq2¯\_(ツ)_/¯07:43
tsimonq2"It Works"07:43
mwhudsonwell sure07:43
tsimonq2mwhudson: I also don't have access to internal Canonical Foundations Trello, so I'm curious to see what you're actually assigned to. :P07:44
mwhudsonit's making a report that has a per-team view of packages in proposed07:44
tsimonq2Oh, that would be cool.07:44
tsimonq2Team in terms of Canonical teams, packageset teams...?07:45
mwhudsoni'm not actually sure :)07:45
mwhudsoni need to ask someone about that07:45
tsimonq2ah :D07:45
mwhudsonah actually reading the card the idea is to do it via subscriptions07:46
tsimonq2The latter would be useful, and if I do continue writing things from scratch, I could tackle, but I don't know about Canonically things... :P07:46
tsimonq2ohhh07:46
tsimonq2At first that sounds like bug 1683749 but we're talking adding support to Britney for subscriptions?07:46
ubottubug 1683749 in britney "Please list bugs tagged as "update-excuse" on update_excuses" [Undecided,Confirmed] https://launchpad.net/bugs/168374907:46
tsimonq2That would be *sweet*.07:46
mwhudsonum no, more like show packages that are in proposed where e.g. ~ubuntu-desktop is subscribed to bugs on the package07:47
tsimonq2oh :P07:48
tsimonq2So a little less overengineering; got it. :)07:48
mwhudsonat least that's my reading of it07:48
mwhudsontouching britney at all is definitely out of scope :)07:48
tsimonq2Internal Canonical Trello cards are internal, so I guess I'll have to trust your interpretation. :P07:48
mwhudsonHave a different page that lists only packages with a team subscribed to them and then it is broken into sections for different teams.07:49
mwhudsonMinimum details to include for each team:07:49
mwhudson - name of package07:49
mwhudson - age in proposed07:49
mwhudson - is it a candidate07:49
mwhudsonthere you go :)07:49
tsimonq2ah :)07:49
tsimonq2mwhudson: So something like how the sponsorship queue works, ish?07:50
mwhudsonyes i guess so07:50
tsimonq2I don't want to steal your thunder if that's your obligation working for Canonical, but that's probably doable with Not A Lot Of Python...07:51
tsimonq2Anyway, thanks mwhudson, I should probably look at going to bed :)07:53
mwhudsonyeah it shouldn't be a lot of work really07:53
mwhudsontsimonq2: good night07:54
abeatosil2100, I have updated https://github.com/CanonicalLtd/ubuntu-image/pull/15508:38
abeatotests for xenial passing08:38
sil2100abeato: excellent, let me take a look and merge it in in a minute08:43
sil2100abeato: thanks o/08:43
abeatonp, happy to get it merged :)08:44
sil2100abeato: (currently bionic and cosmic snap.sh tests are likely to fail due to the nature of classic snaps, glibc mismatch and how we had to build it in the end)08:44
abeatosil2100, I have created https://github.com/CanonicalLtd/ubuntu-image/pull/160 to follow-up the discussion on grub/classic09:04
dokoseb128: did you forward the issues that sarnold pointed out in https://bugs.launchpad.net/ubuntu/+source/libfprint/+bug/1745454 ?09:21
ubottuLaunchpad bug 1745454 in libfprint (Ubuntu) "[MIR] libfprint" [High,Fix committed]09:21
dokoand please don't set MIR issues to committed state yourself09:22
seb128doko, he did and upstream picked up some others as well, see https://gitlab.freedesktop.org/libfprint/libfprint/issues09:31
seb128doko, sorry, I though that security team ack meant it was good to go and that they just forgot to update the status09:31
seb128doko, it had been sitting like that untouched for a while I was trying to get things unblocked...09:31
seb128doko, we need to sort out the MIR situation at some point, we need more people doing revie. we keep having features missing full cycles because MIR just don't move (e.g https://bugs.launchpad.net/ubuntu/+source/tracker-miners/+bug/1770877)09:34
ubottuLaunchpad bug 1770877 in tracker-miners (Ubuntu) "[MIR] tracker-miners" [Undecided,New]09:34
dokocoreycb: https://bugs.launchpad.net/ubuntu/+source/zvmcloudconnector/+bug/1787640/comments/110:35
ubottuLaunchpad bug 1787640 in zvmcloudconnector (Ubuntu) "[MIR] zvmcloudconnector" [Undecided,New]10:35
seb128doko, so what's the status for libfprintf/fprinfd, do you want want to change the status back to something else? they are currentlu on component-msimatch since before ff can we get them promoted and that mismatch resolved?10:39
dokoI'll look at these today10:40
seb128thx10:51
mwhudsontsimonq2: http://people.canonical.com/~mwh/team-summary.html11:10
tjaalton"canidate"11:16
seb128mwhudson, interesting page, how is it generated? do you have the script/source available somewhere?11:19
mwhudsonseb128: it's generated by a terrible terrible python script that i've been writing in the last hour or so11:20
mwhudsontjaalton: heh thanks11:20
didrocksmwhudson: output very useful info for a terrible script! (btw s/canidate/candidate/)11:33
seb128didrocks, timo pointed that typo a few lines earlier here :)11:35
seb128mwhudson, I would still welcome you script if you want to email me? I would like to consolidate similar info on http://people.canonical.com/~platform/desktop/desktop-packages.html which is one of the report desktop is using11:36
didrocksok, I must admit I only backlogged the last few lines, not the paragraph before :)11:36
seb128trying to turn into a sort of dashboard11:36
mwhudsonseb128: of course i've hacked more and broken it again :)11:36
mwhudsonseb128: i'll push it to lp before i go to bed and send you the url11:36
mwhudson(i'm going to be working on it a bit more over the next week)11:36
seb128thx11:37
seb128would be nice to consolidate those reports/dashboards a bit11:38
mwhudsonseb128: https://git.launchpad.net/~mwhudson/+git/pm-stats/12:11
mwhudsonupdated http://people.canonical.com/~mwh/team-summary.html too12:11
seb128mwhudson, thx12:12
mwhudsonseb128: dashboard proliferation is a bit of a problem, hope i'm not making this worse!12:12
jbichaslashd: are you interesting in backporting https://launchpad.net/ubuntu/+source/gdm3/3.28.2-0ubuntu1.4 to xenial?12:24
jbichahave you ever done security updates before?12:25
coreycbdoko: uploaded a new zvmcloudconnector with your requested change13:03
bdmurraydidrocks: Hi, did you see this comment? https://bugs.launchpad.net/ubuntu/+source/apport/+bug/1778694/comments/814:43
ubottuLaunchpad bug 1778694 in apport (Ubuntu Bionic) "In autoreport mode in whoopsie-preferences API, reports are not sent by whoopsie" [High,Fix committed]14:43
didrocksbdmurray: yes, I didn't understand about your question, it's declaring:14:46
didrocks[Install]14:46
didrocksWantedBy=paths.target14:46
didrocksso, it's started on all systems, via systemd14:46
didrocksas this target should be always be ran14:47
didrocksif it's not, it's a systemd bug I guess14:47
didrocksbdmurray: if it's only when you upgrade the package, indeed, I think nothing is starting it until next reboot14:47
didrockswhich was an issue only during bionic sru upgrade (until next reboot)14:47
didrocksor new installs14:48
didrockswe can force starting it in cosmic if you feel that being necessary14:48
didrocks(in the postinst script)14:48
didrocksbut I would rather investigate why the systemd debhelper helper doesn't do it by default14:48
didrockssounds like the fix should be there (and so more on systemd debian/ubuntu maintainers)14:49
didrocksrather than workaround that around in one package14:49
bdmurrayMy point is that in gnome-control-center when you choose automatic the service should be started so people don't keep getting crash dialogs.14:50
bdmurrayIsn't that the goal?14:50
didrocksno, the .path should be always activated14:51
didrocksunconditionnaly14:51
didrocksthen, the script check for the automated report condition14:52
didrockssounds like the systemd postinst helpers should do that anyway (IMHO)14:52
* didrocks needs to step out for a bit, bbl14:52
=== chrisccoulson_ is now known as chrisccoulson
dokocoreycb: I had it already accepted. but I know we changed the defaults in other packages as well15:01
coreycbdoko: fwiw we're still defaulting the core openstack packages to py2 in cosmic15:03
coreycbdoko: we're planning to get function testing done in cosmic and flip them to default to py3 in D.15:04
dokocoreycb: ugh, because we want to switch to 3.7 when D opens ...15:09
coreycbdoko: we've just not had the time to function test py315:09
coreycbjamespage: fyi ^15:10
www2hi i wand to report that liblilv have problems with multi architect setups (x86/x86-64) a spesial with gstreamer-bad16:56
nacc!bug | www216:56
ubottuwww2: If you find a bug in Ubuntu or any of its derivatives, please file a bug report at: http://bugs.ubuntu.com/ IRC is not a good medium to report bugs and this channel is for development coordination.16:56
GunnarHjHi slangasek, did you receive an email from sil2100 the other day about bug #1778041? You input in the matter would be much appreciated.17:49
ubottubug 1778041 in freshplayerplugin (Ubuntu Xenial) "browser-plugin-freshplayer-pepperflash broken" [High,In progress] https://launchpad.net/bugs/177804117:49
slangasekGunnarHj: I did not18:07
slangasekGunnarHj: correction, I found it now18:12
GunnarHjslangasek: Ok.18:24
tsimonq2mwhudson: ooh19:43
GunnarHjslangasek: Thanks for your reply. I think it's sufficient guidance for sil2100 to decide on those pending SRUs.20:06
=== sergiusens_ is now known as sergiusens
scientesis there a reason gnome xorg doesn't work?23:08
naccscientes: this is not a support channel23:09
naccscientes: you want #ubuntu23:09
scientesoh my bad ubuntu+123:09
naccor that23:09
bluesabreIs there any way to blacklist a package from showing on gnome-software? For example, exo-utils (part of Xfce) would never be installed directly and can cripple the desktop when uninstalled (see https://bugzilla.xfce.org/show_bug.cgi?id=14588)23:48
ubottubugzilla.xfce.org bug 14588 in General "Deleting "Mail Reader" or "Web Browser" crashes the computer" [Critical,New]23:48

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!