14:02:21 #startmeeting Mer bug triage 9/1/2011 14:02:21 Meeting started Mon Jan 9 14:02:21 2012 UTC. The chair is Stskeeps. Information about MeetBot at http://wiki.merproject.org/wiki/Meetings. 14:02:21 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:02:32 #info From next week, we have this meeting at 12:00 UTC instead 14:02:49 we'll be going through this bugzilla search: https://bugs.merproject.org/buglist.cgi?emailassigned_to1=1&query_format=advanced&bug_severity=critical&bug_severity=major&bug_severity=normal&bug_severity=trivial&bug_severity=enhancement&bug_status=NEEDINFO&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=TRIAGEDUPSTREAM&bug_status=REOPENED&email1=need-triage&emailtype1=substring 14:03:15 #topic https://bugs.merproject.org/show_bug.cgi?id=76 - Bug 76 - Update zypper to latest upstream version 14:03:34 lbt, will you do the honors of setting the bug flags etc as we go? 14:03:46 hehe .... was just typing that 14:04:00 i think this one should be a task, with high priority 14:05:08 any objections? 14:05:17 nope ... any takers yet? 14:05:31 "not-taken" then 14:05:54 as per http://wiki.merproject.org/wiki/Bug_Lifecycle#Triage ? 14:05:57 right 14:06:09 done 14:06:31 #topic https://bugs.merproject.org/show_bug.cgi?id=91 - Bug 91 - Mer build/versioning policy needs clarification 14:06:43 medium, we'll discuss a bit direction of it in release management meeting tomorrow? 14:06:46 and should be a task 14:07:14 I'll take, cc you 14:07:18 Sage: cc you too? 14:07:40 (yes) 14:08:14 err... done 14:08:27 #topic https://bugs.merproject.org/show_bug.cgi?id=95 - Bug 95 - Mer package signing is not implemented 14:08:50 same, task, medium, -> "Implement Mer package signing in release process"? 14:09:04 discuss tomorrow 14:09:10 OK, that works 14:09:15 medium or high 14:09:37 lbt: sure 14:10:07 Stskeeps: :nod: 14:10:22 medium, i guess, since it needs to be done correctly and thought about, not a high prio rush job :) 14:10:56 mmm ... high means "start soon", not "finish soon" 14:11:11 mm, true 14:11:12 and "stop procrastinating about it" 14:11:16 ok, high then 14:11:37 *totally* agree on thinking it through... it's why I've never set one up :) 14:11:48 ok, moving on to not-taken bugs 14:11:50 not-taken for a bit 14:11:52 #topic not-taken bugs: 14:12:05 https://bugs.merproject.org/buglist.cgi?emailassigned_to1=1&query_format=advanced&bug_status=NEEDINFO&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&email1=not-taken&emailtype1=substring 14:12:47 mm 14:12:55 maybe we should do this first next time :) 14:13:04 #topic https://bugs.merproject.org/show_bug.cgi?id=34 - Bug 34 - python randomly fails to build 14:13:14 nah, some of them are pretty important to get taken :P 14:13:36 this one, we need someone to investigate what graminit.c is initialized from - it looks autogenerated or it wouldn't crash each time 14:13:57 there are multiple potential reasons why it randomly breaks: make -j too high, or sort ordering in file system somehow being off at times 14:14:09 any takers to look into what causes it? 14:14:49 \o 14:14:55 alright 14:15:00 #info assign to vgrade 14:15:12 vgrade: if you have any questions with it, feel free to prod me naturally 14:15:16 it is difficult to chase 14:15:39 thanks 14:15:46 assigned 14:16:10 #topic https://bugs.merproject.org/show_bug.cgi?id=73 - Bug 73 - ofonod segfaults at boot on N900 14:16:34 we have backtraces now 14:16:57 #1 0x00074364 in ofono_dbus_signal_property_changed (conn=0xea1220, path=, interface=0xccd88 "org.ofono.NetworkOperator", name=0x0, type=115, value=0xbe8525f4) at src/dbus.c:193 seems wrong 14:17:12 (see the 0x0?) 14:18:08 any takers? 14:19:48 alright, since no takers for the bug we'll leave it at not-taken@ 14:20:08 prio change? 14:20:22 seems correct to me 14:20:32 maybe critical since it's a crash 14:20:51 critical prevents device from booting iirc 14:21:05 mm 14:21:09 "major major loss of function" 14:21:12 I guess 14:21:15 ok, major then 14:21:31 I assume the rest of the device stays up :) 14:21:55 (just making sure we use the values as per definition - or change the definition) 14:22:01 right, that's the end of the list for today 14:22:18 i think we should possibly have another meeting to help prioritize unprioritized tasks 14:22:23 https://bugs.merproject.org/buglist.cgi?emailassigned_to1=1&query_format=advanced&bug_severity=task&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&email1=need-triage&emailtype1=substring 14:22:40 that should be a one-off shouldn't it 14:23:00 hmm? 14:23:19 all tasks should be given a prio in triage 14:23:34 so sure, we need to do this... but only once (hopefully) 14:23:53 think it's a continous evaluation of tasks 14:23:56 (btw.... most of them need re-assigning to not-taken) 14:23:59 yes 14:24:29 all I'm saying is that that search will be empty most of the time 14:24:40 true 14:24:51 but I was going to add that we should be looking to review prio of not-taken regularly 14:25:00 yes 14:25:01 which may be the same goal :) 14:25:31 is now good? 14:26:20 i vote to switch to not-taken@ all of them and then review them, i guess 14:26:25 OK 14:26:43 i'll do that.. 14:26:48 nah 14:26:50 wait 14:27:04 (there's a change multiple bugs functionality) 14:27:08 it's processing 14:27:12 yes, using that 14:27:24 done 14:27:26 k 14:27:38 *g* shows how slow our bz is 14:28:14 ok, https://bugs.merproject.org/buglist.cgi?priority=High&priority=Normal&priority=Low&priority=Undecided&emailassigned_to1=1&query_format=advanced&bug_severity=task&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&email1=not-taken&emailtype1=substring then 14:29:07 #topic Task priority triage 14:29:27 #topic https://bugs.merproject.org/show_bug.cgi?id=15 - Bug 15 - Switch from libjpeg to libjpeg-turbo 14:29:38 i'd say high priority, as there is significant performance benefits in this 14:29:41 especially on ARM systems 14:29:44 and on SSSE3 systems 14:29:49 sec... that's only 'tasks' 14:29:50 https://bugs.merproject.org/buglist.cgi?emailassigned_to1=1&query_format=advanced&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&email1=not-taken&emailtype1=substring 14:31:02 mmkay 14:31:15 any objections to high? 14:31:35 or takers (anyone who wants to do it), for that matter? 14:32:20 (FYI I'm not going to take until my list is smaller) 14:32:29 ok, if anyone wants to take a task, just raise a hand as we go :) 14:32:32 #info high 14:33:05 #topic https://bugs.merproject.org/show_bug.cgi?id=19 - Drop GL support from Mer Core 14:33:06 I'm adding a note to the bug - feedback to OP that there is activity 14:33:34 low priority, it's not really harming anyone at the moment, but it's a nice thing to do to make packaging story simpler 14:33:39 and APIs in use 14:34:34 done 14:34:35 #info low 14:34:47 #topic https://bugs.merproject.org/show_bug.cgi?id=26 - Evaluate dash as /bin/sh vs /bin/bash, disk/memory footprint/boot time/impact 14:35:08 medium, as it's a bit of a cost vs benefit thing, would be useful to see what we would actually gain on it 14:35:15 architectual backlog, FWIW 14:35:36 btw, that Drop GL is need info. Solution exists if decision is made. 14:35:40 i'm also pondering if we can support dual stack Mer, one with gnu utilities, one with busybox 14:35:43 ok 14:36:00 how do you want to mark architectural backlog ? 14:36:09 project-core severity=task is architectual backlog 14:36:24 just letting you know it's in that category 14:36:41 * lbt reaches for the wiki 14:37:04 #info medium - cost/benefit thing, useful to check what we can gain on it 14:37:13 #info drop GL is in NEEDINFO, solutions exist but need info to proceed 14:37:37 #topic https://bugs.merproject.org/show_bug.cgi?id=27 - Evaluate benefits eglibc vs glibc, in terms of packaging, cross compilation and footprint/impact 14:37:58 i propose medium priority - it's best for all of us if we stay close to linaro gcc and their setup, which is eglibc 14:38:05 as we can get help for compiler errors 14:38:48 the fedora glibc packaging is a tad awkward, so 14:40:08 done? 14:40:28 #info medium priority - it's best for all of us if we stay close to linaro gcc and their setup, which includes eglibc 14:40:48 #topic Bug 28 - Drop gamin dependancy - https://bugs.merproject.org/show_bug.cgi?id=28 14:41:11 #info this is in glib2 and a oneliner - we can drop the gcc atomics patch for armv6 now as well while someone's at it anyway 14:41:35 low prio - sage, perhaps you can take it as you touched it last? 14:42:59 I can take this sure 14:43:04 #info low, sage takes 14:43:18 done 14:43:24 #topic https://bugs.merproject.org/show_bug.cgi?id=29 - Drop gamin from core 14:43:34 low, this can be assigned to me when gamin dep is removed 14:43:39 #info low, this can be assigned to me when gamin dep is removed 14:44:09 done 14:44:17 #topic https://bugs.merproject.org/show_bug.cgi?id=30 - bash should link against system readline, not against it's own copy 14:44:21 nice junior job 14:44:37 low priority, not terribly important, but nice to clean out memory a little bit 14:44:51 I'll take it - I need to run through some process checks 14:44:58 k 14:45:12 #info low, not terribly important, but nice to clean out memory a little bit, lbt assignee 14:45:58 #topic Move ntpdc and ntpq to ntp-utils package to save footprint - Bug 31 - https://bugs.merproject.org/show_bug.cgi?id=31 14:46:14 this is also a junior job, the reasoning is that we then can drop libedit from image 14:46:21 connman uses ntp directly, not through command line tools 14:46:31 low priority? 14:46:39 yes 14:46:53 #info low, junior job, the reasoning is that we then can drop libedit from image 14:47:00 #info connman uses ntp directly, not through command line tools 14:47:35 #topic pixman should be upgraded to 0.24 - https://bugs.merproject.org/show_bug.cgi?id=35 14:47:57 #info medium, is probably related to overall performance improvements 14:48:26 #link http://gitweb.merproject.org/gitweb?p=mer-core/pixman.git;a=tree 14:48:40 aside: do we need an upstream-new-release-tracking process ? 14:48:47 it would be useful to have something 14:49:42 it's almost like we need a package database... 14:49:52 packages.xml ;) 14:49:59 i'm going to skip 38-51 as they are the bashism ones and already low priority 14:50:39 #topic https://bugs.merproject.org/show_bug.cgi?id=52 - Switch from DWARF-3 to DWARF-4 in order to save debuginfo space 14:50:45 #info assign to stskeeps, medium 14:50:56 i already have a patch waiting for this, so 14:52:30 #topic Switch db4 to libdb, 5.2.36 and switch RPM to it - https://bugs.merproject.org/show_bug.cgi?id=53 14:52:49 this is a bit more complex, integrating a new package and switching RPM to be using this instead of db4 14:53:05 i'd say medium prio as it can be benefits in package performance/solidity 14:53:48 any info for potential takers? 14:54:33 sec 14:55:27 http://pkgs.fedoraproject.org/gitweb/?p=rpm.git;a=blob;f=rpm.spec;h=323ee6af83ac6e422e0ff07258f899c3b4defb14;hb=HEAD , http://pkgs.fedoraproject.org/gitweb/?p=libdb.git;a=tree 14:55:36 we don't need to care about db1.85 14:56:22 #topic https://bugs.merproject.org/show_bug.cgi?id=54 - Bug 54 - Include package build logs to the releases 14:56:26 triage in tomorrow's meeting? 14:57:10 medium prio, i think 14:57:20 or even low 14:57:28 it is not so important for general usage 14:57:38 more like nice to have if someone has time to do it 14:57:39 it's part of what vendors need 14:58:02 well, that is true. 14:58:03 how about NEEDINFO 14:58:17 and putting some thought into what/when/how ? 14:58:26 mmk 14:58:45 #info handle in RM discussions 14:58:46 This is part of Mer's information systems 14:59:02 #topic https://bugs.merproject.org/show_bug.cgi?id=57 - dbus-glib is now merged to glib upstream so we should remove it 14:59:06 and should be part of our deliverables - nice clear info 14:59:14 OK... moving on then... 14:59:40 this is still in use by upower and ConsoleKit 14:59:46 and gconf 15:00:05 and dbus-python(?) 15:00:17 so low priority for now? 15:00:57 Sounds good to me. 15:01:04 ConsoleKit can be removed but need to check rest 15:01:40 #info low, still in use by upower and ConsoleKit and gconf and dbus-python 15:01:53 #topic https://bugs.merproject.org/show_bug.cgi?id=58 - nss should be upgraded to NSS_3_13_1_RTM 15:02:12 high, as this may include security fixes, and is one of our central security libraries 15:03:22 #info high 15:04:51 done 15:05:05 #topic https://bugs.merproject.org/show_bug.cgi?id=59 - nspr should be upgraded to 4.8.9 15:05:14 probably is a requirement of NSS anyway 15:06:06 blocks? 15:06:14 think so 15:06:19 a 'weak' blocks, at least 15:06:39 so medium prio 15:06:50 blocks a high 15:06:53 yeah, ok 15:06:56 high then? 15:06:59 #topic PAM should use system db4, not it's own - https://bugs.merproject.org/show_bug.cgi?id=69 15:07:37 you'd be surprised how many copies of sqlite and db4 there's around 15:07:47 :) 15:07:50 low priority too though 15:07:52 yes 15:08:02 * Sage thinks that he can take some of those 15:08:26 just assign them to yourself then 15:08:30 #topic e2fsprogs should be upgraded to 1.42 15:08:31 Sage: time to teach people how to fish I think.... 15:08:51 medium, junior job, always good to follow basic system utilities 15:09:23 lbt: ? :) 15:09:42 Stskeeps: You probably will announce list of stuff people can take if interested? 15:09:45 yes, i will 15:09:54 give a man a fish/bug, feed him for a day.... teach a man to fish/debug... feed him for life 15:10:04 so leave some free ;) 15:10:14 Stskeeps: I was thinking we could do a blog/tweet 15:10:18 next one is ofonod bug, we'll skip that 15:10:20 lbt: mailing list atm 15:10:25 Stskeeps: hehe, ok :) 15:10:37 yeah, or email ... 15:10:47 bug 76 we already prioritized 15:11:09 #topic https://bugs.merproject.org/show_bug.cgi?id=78 - Bug 78 - Releases should have delta RPMs 15:11:31 medium, useful to have, discuss tomorrow in RM? 15:12:09 sure, I don't see this very high prio as it doesn't block anything afaik, just increases the speed of update 15:12:20 :nod: 15:12:23 not low? 15:12:30 low then 15:12:43 this is 'today's prio' ... we get to do this ... monthly ? 15:12:49 14 days or so 15:13:13 better 15:13:54 #topic https://bugs.merproject.org/show_bug.cgi?id=84 - Package eglinfo and glestest for easy EGL testing 15:14:02 junior job, low, but very useful for hardware adaptations 15:14:15 imporance high :) 15:14:23 yes 15:14:39 ok, high.. 15:14:39 :P 15:14:49 pain in the *** when doing new adaptation :) 15:14:54 package name? 15:15:18 was it in mesa or something? 15:15:32 there are links to the source 15:15:34 this is just some non-mesa ones since mesa usually drags in odd windowing system stuff 15:15:37 so I assume a new pkg 15:15:37 works just fine 15:16:03 #topic Bug 85 - Investigate variables for deciding if a %post/RPM scriptlet happens during build, image install or on active device - https://bugs.merproject.org/show_bug.cgi?id=85 15:16:23 this is a bit of a feature request/thing that came up in discussion, in order to differ in %post and %pre between those scenarios 15:16:40 sec 15:16:52 84 : Will need a new standalone package, git repo etc. ? 15:16:59 yes 15:17:02 done 15:17:19 hmm 15:17:28 i'd say medium priority because i'm just sure we'll hear than in actual product programmes too.. discuss fully in RM meetings tomorrow? 15:17:29 I'm looking into this kind of thing 15:17:39 than=that 15:17:44 with the lua/rpm 15:18:24 :nod: 15:18:40 #topic Switch to using %if's for test suites for QEMU issues instead of %ifarch - https://bugs.merproject.org/show_bug.cgi?id=86 15:19:01 this is a bit of a project-wide thing, that we can tell if we're building with a qemu or an actual device 15:19:08 because eventually, we will want to run testcases on actual devices 15:19:23 i'd say medium priority 15:20:13 right now we do %ifarch %{arm} mipsel etc 15:20:16 which isn't cool 15:20:16 :P 15:21:16 I think make this blocked by an "rpm/variable usage policy" 15:21:29 which I'm kinda working on with the kernel pkg 15:21:45 ok 15:22:00 nb... when we do policy stuff ... we should validate it 15:22:02 automatically 15:22:08 yess 15:22:18 #topic Mer should contain a package that contains default kernel settings - https://bugs.merproject.org/show_bug.cgi?id=90 15:22:22 high? 15:22:22 :P 15:22:36 I'm not sure it's a package 15:22:53 i think it should be, else we can't deploy changes easily 15:23:01 a .config? 15:23:02 CONFIG_'s for systemd, that kind of stuff 15:23:08 yes, it's a .config, but we need to spread it somehow 15:23:15 and ideally through automatic rebuilds 15:23:20 either way high, RM meeting? 15:23:26 OK... that's different then 15:23:38 just basically what the userland expects 15:23:47 hmm... 15:23:48 it should be part of the default HA kernel packaging 15:24:04 yes, but that packaging will stall ;) 15:24:05 ie 'validate needed Mer CONFIG_' check 15:24:10 http://wiki.merproject.org/wiki/Adaptation_Guide#Kernel <- ? 15:24:20 Sage: yes .. I like that 15:24:30 but Stskeeps is right... needs to be auto and updated 15:24:42 we had config-generic in meego and it was a nightmare 15:24:47 as everyone had their own copy 15:24:50 yes 15:24:51 so the kernel HA package depends on this Mer core as a BR 15:25:11 #topic Bug 92 - Include --with-linker-hash-type=gnu, may improve app launch performance - https://bugs.merproject.org/show_bug.cgi?id=92 15:25:18 high prio, assign to me (i'm working on it atm) 15:25:23 so mer core provides 'minimal' CONFIG_ list which is used by HA 15:25:35 right 15:26:09 and script that checks that user has them when building? 15:26:32 however, it is uses responsibility to call it 15:26:38 s/uses/users/ 15:26:58 Sage: I'd make it part of the kernel packaging 15:27:31 Warning: Your kernel does not provide all Mer CONFIG_ settings. See.... http://..." 15:27:48 * lbt takes it 15:28:09 bug 95 we already went through 15:28:14 so we're done for today 15:28:32 lbt, we also need 'your kernel provides xxx but must not for Mer' 15:28:45 vgrade: good point 15:29:09 PARANIOD_ANDRIOD etc 15:29:11 vgrade: if you have known valueas could you list them to the wiki page so we dont' forget? 15:29:28 yes 15:29:45 okay, thank you all for coming 15:30:00 #endmeeting