14:03:09 <Stskeeps> #startmeeting Mer bug triage 2/1/2012 14:03:09 <MerBot> Meeting started Mon Jan 2 14:03:09 2012 UTC. The chair is Stskeeps. Information about MeetBot at http://wiki.merproject.org/wiki/Meetings. 14:03:09 <MerBot> Useful Commands: #action #agreed #help #info #idea #link #topic. 14:04:28 <Stskeeps> welcome to the first mer bug triage :) the procedure is fairly simple, we walk through https://bugs.merproject.org/buglist.cgi?priority=Undecided&query_format=advanced&emailassigned_to1=1&order=Bug%20Number&bug_severity=critical&bug_severity=major&bug_severity=normal&bug_severity=trivial&bug_severity=enhancement&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&email1=need-triage&emailtype1=substring , set priority, ... 14:04:34 <Stskeeps> ... discuss a bit about the bug and issues. If you see a bug you're volunteering to fix, feel free to indicate this 14:06:32 <Stskeeps> we can vary on severity=critical,major,normal,trivial,enhancement and task, and priority=low,normal,high 14:06:43 <Stskeeps> task means it's a bite-size task someone can do 14:06:46 <Stskeeps> okay, let's start started 14:07:06 <Stskeeps> er, get started :) 14:07:24 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=3 - qt-mobility does not build QtMessaging module 14:08:05 <Sage> do we have qmf in mer core already? 14:08:06 <Stskeeps> i propose this to become a task, and changing it into "enable pkgconfig(qmf) as buildrequires" - priority high 14:08:09 <Stskeeps> yes, it's there 14:08:15 <Stskeeps> that was bug 10 :) 14:08:46 <Sage> ok. Then this is quite easy however I would prefer someone to try the qt messaging before it is submitted to mer core as a patch. 14:09:16 <Sage> qmf is already release or currently in pre release? 14:09:19 <Stskeeps> any takers for fixing it or any objections (if you want, assign to yourself later)? 14:09:21 <Stskeeps> prerelease, i think 14:09:37 <Sage> I can take this one for packaging point of view but someone else needs to test it. 14:09:49 <Stskeeps> ok 14:10:03 <Stskeeps> #info task, high, assigned to sage, someone else needs to QA 14:10:05 <Sage> I might be the one who removed it originally :) 14:10:29 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=4 - Bug 4 - process needs a way to wait-for other changesets to be merged for testing 14:10:55 <lbt> I think I should take that one 14:10:56 <Stskeeps> in Mer, we use a workflow system called BOSS, and this bug deals with developers being able to indicate "please first merge this when another change has been made" 14:11:18 <Stskeeps> i'd say medium priority as it's not critical, but with the latest toolchain work, it is definately needed 14:11:34 <Stskeeps> #link http://gitweb.merproject.org/gitweb?p=mer/boss-workflow.git;a=tree 14:11:40 <lbt> good - that's what I wondered 14:11:53 <Stskeeps> any objections to medium priority? 14:12:16 <Stskeeps> #info assign to lbt 14:12:40 <Stskeeps> #info medium 14:12:51 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=6 - Toolchain bug occurs on armv7hl architecture builds 14:12:58 <lbt> mmm 14:13:13 <Stskeeps> #info has been fixed with prerelease new compiler, to my knowledge 14:13:20 <Sage> yes 14:13:44 <Sage> at least someone said so at IRC IIRC 14:14:16 <Stskeeps> i propose RESOLVED FIXED, commenting in bug this fact 14:14:53 <Sage> agreed 14:14:56 <Stskeeps> #info RESOLVED FIXED, fixed in prerelease new compiler 14:15:28 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=7 - DNS address not set with static IP 14:16:05 <lbt> Is this a Mer bug? 14:16:12 <Stskeeps> isn't connman always listening on 127.0.0.1 and forwarding to the correct DNS server? 14:16:32 <Sage> we should test newer connman 14:16:33 <Stskeeps> or whatever it perceives as the correct DNS server 14:16:53 <Sage> the one that fixes the grps DHCP/DNS query problems might fix this one 14:16:58 <wagi> Stskeeps: yes, ConnMan has a dnsproxy 14:17:25 <Stskeeps> #info connman has dnsproxy, DNS setting may be reliant on what connection is being set as 'default' 14:17:38 <lbt> FWIW the file should contain a comment about usage in Mer 14:17:44 <wagi> Yep, ConnMan keeps a list of DNS servers 14:17:49 <Stskeeps> agreed, ie, "AUTOGENERATED"? 14:17:58 <Stskeeps> resolv.conf is created by connman so, afaik 14:18:09 <wagi> You can ask with the python test scripts what DNS server is active 14:18:09 <lbt> "do not edit this file as Mer uses conman on 127.0.0.1" 14:18:40 <wagi> resolv.conf forwards all DSN requests to ConnMan 14:18:55 <Stskeeps> ok, so i propose NEEDINFO - wagi: could you elaborate to the submitter on what python test scripts to use to gain more accurate info? 14:19:06 * Sage misunderstood the bug 14:19:24 <wagi> Stskeeps: sure, there are some test scripts in the ConnMan repository 14:19:28 <Stskeeps> :nod: 14:19:33 <andre__> but there is no NEEDINFO state? 14:19:34 <Stskeeps> connman-tests 14:19:38 <Stskeeps> andre__: there's no NEEDINFO state? 14:19:39 <wagi> They use work on the ConnMan D-Bus API 14:19:50 <andre__> not by default. except for that you (or somebody) added it. 14:19:53 <wagi> So you can use them to find out what's the current configuration 14:20:05 <Stskeeps> andre__: oh crap :) 14:20:15 <lbt> Sage: doesn't this mean that the GUI isn't telling conman what DNS server to use? 14:20:42 <wagi> You can of course add static DNS configuration if you like 14:20:49 <Sage> lbt: I'm not sure how the user is setting the ip there. 14:20:59 <wagi> Stskeeps: sure, propably I need to create an account first :) 14:21:08 <Sage> lbt: if user sets the IP without using the connman dbus api it is user problem IMO. 14:21:13 <Stskeeps> andre__: hmm, we have a field for it - let's take this in #mer after the meeting 14:21:21 <Stskeeps> yes, it's unclear if user sets IP without using connman APIs 14:21:25 <lbt> Sage: that's what I meant - is this a Mer or Nemo bug 14:21:31 <Sage> lbt: Mer 14:21:56 <lbt> What mechanism is the user using to set the IP config on wifi ? 14:22:01 <Stskeeps> lbt: connman d-bus apis 14:22:08 <andre__> Stskeeps: https://bugs.merproject.org/editworkflow.cgi for getting NEEDINFO 14:22:14 <andre__> (I guess) 14:22:15 <lbt> I don't see that in the bug so OK 14:22:46 <Sage> lbt: user isn't telling that what he is using, but he should be using connman dbus api's. thus NEEDINFO 14:23:03 <Stskeeps> #info User needs to explain what he uses to configure DNS, NEEDINFO 14:23:10 <Stskeeps> #info should be done with connman d-bus apis 14:23:27 <wagi> yep 14:23:31 <lbt> wagi: are you editing that bug then? 14:24:02 <Stskeeps> we'll edit as we go? 14:24:07 <wagi> lbt: Okay, I'll edit the bug 14:24:12 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=9 - Mer_Core_armv6l/maliit-framework compilation error 14:24:17 * wagi creates account 14:24:20 <lbt> Stskeeps: yes - but we should say who or we'll collide 14:24:24 <Stskeeps> yes, ok 14:24:37 <Sage> Stskeeps: just marked that resolved fixed as it is fixed in next prerelease 14:24:38 <Stskeeps> i believe this framework compilation error was fixed with the new toolchain 14:24:46 <Stskeeps> #info RESOLVED FIXED, fixed with prerelease 14:25:10 <wagi> lbt: if you have an account you can update it? It seems to take a while until I have the account. whitelisting :) 14:25:17 <lbt> will do 14:25:22 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=13 - ca-certificates packages is missing from Mer Core 14:25:23 <wagi> lbt: thanks 14:25:30 <veskuh> would be nice to have ALREADYFIXED for the previous kind of cases 14:25:48 <Stskeeps> #info maybe we need ALREADYFIXED? 14:26:03 <Stskeeps> this should be severity=task and priority=high 14:26:08 <andre__> well, ALREADYFIXED is just FIXED with an older TM ;-) 14:26:13 <lbt> we need a bug lifecycle 14:26:18 <lbt> andre__: ^^^ ? 14:26:34 <Sage> Stskeeps: I can take this one as well if you create ca-certificates to gitweb 14:26:44 <Stskeeps> ok 14:26:44 <Sage> I have the packaging ready already for this 14:26:57 <Stskeeps> #info assign to sage, stskeeps makes git repository 14:27:01 <andre__> lbt, well, there is one by default. but maybe it does not (yet) totally fit. :P 14:27:02 <Stskeeps> #info severity=task, priority=high 14:27:39 <Stskeeps> Sage: ca-certificates added 14:27:50 <Sage> ok 14:27:58 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=14 - Bug 14 - Organizer 14:28:03 <andre__> I assume I can close #14 as INVALID, with a comment explaining how to file good bug reports? 14:28:11 <Stskeeps> nah, the bug is correct 14:28:15 <Stskeeps> #info should be QtOrganizer missing 14:28:18 <andre__> ah 14:28:37 <Stskeeps> #info http://doc.qt.nokia.com/qtmobility/qtorganizer.html 14:28:44 <andre__> I always have problems if people keep it too short. 14:29:00 <Sage> I have fix for this as well but someone needs to test it 14:29:02 <Stskeeps> Sage: can you look at this together with pkgconfig(qmf)? 14:29:02 <Stskeeps> ok 14:29:08 * andre__ edits the bug summary 14:29:17 <Stskeeps> #info assign to sage as he has fix for it but needs QA 14:29:22 <Stskeeps> medium priority? 14:29:35 <Sage> I guess so if nobody is ugently needing this and willing to test :) 14:30:09 <vgrade> same as https://bugs.merproject.org/show_bug.cgi?id=3 14:30:17 <Sage> Stskeeps: are you handling the assignments later after the meeging btw? 14:30:34 <Stskeeps> yes, after triage 14:30:39 <lbt> when doing NEEDINFO do we assign to the OP ? 14:30:54 <Sage> Stskeeps: ok 14:31:12 <Sage> vgrade: not same, similar. Qt messaging vs qt organizer. 14:31:23 <Stskeeps> #info medium 14:32:02 <lbt> Stskeeps: there is no "medium" (AFAICT) 14:32:09 <Stskeeps> ok, "normal" 14:32:25 <Stskeeps> #info normal 14:32:31 <Stskeeps> consistency++ 14:32:31 <Stskeeps> :P 14:32:36 <lbt> yeah :D 14:32:50 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=21 - qt-mobility has timestamps in debugsource 14:33:25 <Stskeeps> so, why are timestamps bad in binaries or source? :) because it means that if qtmobility recompiles and it compares differently to previously built rpms of same source package, it will rebuild everything utilizing qtmobility 14:34:04 <Stskeeps> so i'd say 'normal' 14:34:21 <Sage> sounds good 14:34:39 <Stskeeps> w00t: you're up for this one? 14:34:59 <w00t> i've not tried mobility before, and have a bit of a backlog, but sure 14:35:01 <w00t> i can take it 14:35:02 <Stskeeps> ok 14:35:08 <Stskeeps> #info normal, assign to w00t 14:35:32 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=23 - openssl contains timestamps 14:35:46 <Stskeeps> same problem as before, except openssl causes a lot more rebuilds 14:36:05 <Stskeeps> it should be a __DATE__ __TIME__ removal patch somewhere, so this is a perfect starter job :) 14:36:33 <vgrade> \o 14:36:40 <Stskeeps> assign to vgrade? 14:36:55 <vgrade> yup 14:36:55 <Stskeeps> #info turn into task, high priority, assign to vgrade 14:37:39 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=24 - Drop libGL dependancy from cairo 14:37:45 <Stskeeps> this is related to Bug 19 - Drop GL support from Mer Core 14:37:56 <Stskeeps> as we want to make a consistent api across platforms, and glesv2/egl is one that's shared 14:38:00 <Stskeeps> should probably be a 'task' instead 14:38:56 <Stskeeps> if you have any objections to the nature of some tasks, those are welcome too 14:39:24 <Stskeeps> #info low priority, task 14:39:34 <Stskeeps> also a nice easy packaging job 14:40:29 <Sage> I can take it if nobody else wants it. 14:40:47 <Stskeeps> #info assign to sage, other takers welcome 14:41:11 <lbt> do we need a backlog assignee ? 14:41:32 <lbt> for triaged but non-allocated tasks 14:41:38 <Sage> fixed ;) 14:41:45 <lbt> :P 14:41:51 <Stskeeps> we try to set a priority and then that becomes a task available for taking 14:42:27 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=25 - Bug 25 - Drop libGL dependancy from Xephyr / drop Xephyr 14:42:39 <Stskeeps> i'm a little unsure why xephyr needs libGL in dependancies, may be bad packaging 14:42:39 <Sage> that might be a bit harder one 14:42:42 <Stskeeps> yeah 14:42:48 <Stskeeps> low priority? 14:43:04 <Sage> I can check if it is easy or if it needs more work. 14:43:09 <Stskeeps> ok 14:43:21 <Stskeeps> #info sage will look into the nature of the task to elaborate on it 14:43:26 <Stskeeps> #info low priority 14:44:07 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=34 - python randomly fails to build 14:44:16 <Stskeeps> this one is really annoying as it fails our build checks very often 14:44:19 <Sage> btw, in which package xephyr is? 14:44:23 <Stskeeps> xorg-x11-server 14:44:43 <Stskeeps> so i'd say normal priority, it may be a race condition 14:45:22 <Stskeeps> and fixable with -j1 14:46:01 <Stskeeps> we run with -j8 which occasionally stumps builds 14:46:22 <Stskeeps> #info normal , may be fixable wih -j1, we use -j8 which occasionally breaks matters 14:47:11 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=36 - Bug 36 - Information missing from project-core gitweb summary page 14:47:17 <Stskeeps> this should be over in IT and gitweb instead? 14:47:38 <lbt> seems reasonable 14:47:49 <Stskeeps> medium priority, it annoys developers 14:48:15 <Stskeeps> #info should be in IT, gitweb - medium priority 14:48:38 <lbt> not sure if it's a feature request 14:48:47 <Stskeeps> enhancement, i guess 14:48:52 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=37 - cross-*-gcc Requires: libgcc/cpp/libgomp/libmudflap but contains it itself 14:49:09 <Stskeeps> this one is fixed in prerelease, http://review.merproject.org/270 14:49:17 <Stskeeps> #info RESOLVED FIXED, fixed in prerelease, http://review.merproject.org/270 14:50:31 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=38 - Bug 38 - binutils, binutils-devel has a possible bashism in RPM scriptlets 14:50:37 <Stskeeps> okay, so, now we're at the bashism bugs 14:51:15 <Stskeeps> #info http://www.mail-archive.com/mer-general@lists.merproject.org/msg00054.html for a description of these 14:51:35 <Stskeeps> we can convert all those into tasks, ie, "remove bashism from RPM scriptlets" type of summary 14:51:38 <Stskeeps> and low priority 14:51:56 <Stskeeps> the idea is that we want to make it possible to perhaps switch to dash as /bin/sh over time 14:52:26 <Stskeeps> ie, https://bugs.merproject.org/show_bug.cgi?id=26 14:53:20 <Stskeeps> any objections to this assesment? 14:53:40 <Stskeeps> some of them are good junior jobs to take, easy to fix, others need to fix RPM macros 14:53:51 <lbt> will we stop shipping bash? 14:54:13 <Stskeeps> that remains to be seen, it's very useful as a user shell still ;) 14:54:23 <lbt> I'm thinking that as long as it says #!/bin/bash then bash-isms are OK 14:54:26 <Stskeeps> yeah 14:54:32 <Stskeeps> we just have to limit them 14:54:36 <lbt> but policy should say "use dash unless we have to" 14:55:00 <Stskeeps> #info Bashisms: convert to task, summary "Remove bashism from RPM scriptlets" type of summary, low priority 14:55:15 <Stskeeps> #info idea is that we want to make it possible to switch to dash as /bin/sh over time, see https://bugs.merproject.org/show_bug.cgi?id=26 14:56:36 <Stskeeps> #info do this for bugs 39,41,42,43,44,45,46,47,48,49,50,51 too 14:56:51 <Stskeeps> i'm skipping those bugs in the triage, hope that's OK :) 14:57:26 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=62 - gcc on x86 should only bootstrap when needed 14:57:54 <Stskeeps> right now gcc runs a full bootstrap, ie, first builds a gcc, then builds itself with that gcc each and every time on x86 14:58:18 <Stskeeps> this isn't terribly needed and some logic to prevent it from doing so unless the source has modified in any way 14:58:31 <Stskeeps> i'd say low priority as it's a build time slowness? 14:58:47 <lbt> yes 14:59:04 <Stskeeps> #info right now gcc runs a full bootstrap, ie, first builds a gcc, then builds itself with that gcc each and every time on x86 - low priority 14:59:59 <lbt> I'll take that one 15:00:09 <Stskeeps> ok 15:00:14 <Stskeeps> #info assigned to lbt 15:01:20 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=64 - - kernel-adaptation-n810 compiled using pre-release fails boot before reaching fbcon 15:01:52 <Stskeeps> this is a bug in toolchain, workaround is available in https://lkml.org/lkml/2011/5/31/385 15:02:04 <Stskeeps> high? 15:02:46 <lbt> what's the bug? 15:02:56 <lbt> toolchain or kernel ? 15:03:01 <Stskeeps> -fconserve-stack misaligns variables on the stack 15:03:14 <lbt> sorry, I wasn't clear 15:03:15 <Stskeeps> and it's being used in ARM kernels 15:03:26 <lbt> what's bug 64... 15:03:37 <lbt> fix the toolchain or fix the kernel conf 15:03:42 <Stskeeps> fix the toolchain 15:03:48 <Stskeeps> we need to track upstream for a fix 15:05:01 <Stskeeps> #info waiting for upstream for fix, workaround exists 15:05:06 <Stskeeps> #info high 15:06:02 <Stskeeps> #info rename to "toolchain -fconserve-stack misaligns variables on the stack, for ARM" 15:06:32 <lbt> heh... was just hovering on the "edit" 15:06:53 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=65 - upower needs pm-utils 15:07:00 <Stskeeps> Sage: i can create git repo? 15:07:06 <Sage> I have this ready when git repo appears 15:07:07 <Stskeeps> ok 15:07:16 <Stskeeps> done 15:07:19 <Sage> :) 15:07:25 <Sage> so assign to me and I'll fix it 15:07:29 <Stskeeps> #info medium, assign to sage 15:09:25 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=66 15:09:47 <Stskeeps> high priority, perl rebuilding is really annoying :) nice junior task too 15:10:00 <Stskeeps> happens when a build happens overnight 15:10:03 <Stskeeps> as the timestamps change 15:10:25 <lbt> o/ 15:10:26 <Stskeeps> any takers to find and patch ths? 15:10:30 <Stskeeps> ok 15:10:34 <Stskeeps> #info lbt, high 15:10:39 <Stskeeps> #info assign to lbt, high 15:10:58 <Sage> out of bugs? :) 15:11:19 <Stskeeps> almost 15:11:20 <lbt> 73 ? 15:11:30 <Stskeeps> #topic https://bugs.merproject.org/show_bug.cgi?id=73 - Bug 73 - ofonod segfaults at boot on N900 15:12:01 <Sage> ofono :/ 15:12:07 <Stskeeps> this looks awkward, would be good to see if this changed in prerelease 15:12:28 <Stskeeps> should we ask reporter to see if he sees it in new release when we have nemo on top of that? 15:12:31 <Stskeeps> i'd say medium priority 15:12:39 <Sage> sure 15:12:54 <Stskeeps> and help him collect coredumps 15:14:08 <Stskeeps> #info ask reporter to check if it happens in next mer release + nemo on top, high priority 15:14:14 <Sage> we should do guide how to collect core dumps :) 15:14:18 <Stskeeps> yes 15:14:22 <Sage> corewatcher would be nice btw 15:14:39 <Sage> at least for nemo 15:14:48 <Sage> but that needs patching the gtk stuff out. 15:15:06 <Stskeeps> #info medium priority 15:15:33 <Stskeeps> okay, that was it for the triage list 15:16:16 <Stskeeps> thank you all for coming 15:16:19 <Stskeeps> #endmeeting