18:44:02 #startmeeting 18:44:02 Meeting started Wed Nov 5 18:44:02 2014 UTC. The chair is colonelqubit. Plugin info at http://wiki.debian.org/MeetBot. 18:44:02 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:44:40 Busy week; I've a couple of things to talk about, but wanted to ask you guys what you've been up to first 18:44:59 beluga: you mentioned you had some stuff to talk about? 18:45:30 Joel is going through reopeneds currently and that is raising the unconfirmed count 18:45:44 * colonelqubit nods 18:46:01 other sources of "sleeper" uncos are needinfo bugs that the reporter has forgotten to set unco 18:46:55 UNCONFIRMED count is 701 (it was down to 650ish) 18:47:03 as well as my next target: bugs that the reporter themselves has set NEW without any replies.. Joel didn't want to mass-change them to unco :) There are about 280 of those, 80 of which are enhancements 18:47:39 * colonelqubit is a little more forgiving of enhancements being self-confirmed 18:48:07 Joel gave me a task: yesterday and today I went through bugs that had the keyword regression or had regression in the summary, but no bibisect* in the whiteboard. There were about 140 of them and I completed the task. 18:48:16 great 18:48:20 but several of them are function already there that the user is not aware of 18:48:54 sophi: oh, certainly. That's why I said I was a *little* more forgiving :) 18:49:05 colonelqubit: ok :) 18:49:51 I think in some projects, enhancements are basically auto-confirmed, so I can imagine new bug reporters thinking we operated in the same fashion 18:50:17 then there's the general idea for future discussion: as we go through more and more of the older uncos, the concentration of difficulty of repro increases. It was said that it would help to add keywords etc., but I skimmed through the old ones and they are very unique, so needs* doesn't really make sense 18:51:01 beluga: agreed 18:51:42 for bugs that require some Linux distro or DE that I don't have, I join the IRC channel of said thing.. Xfce was very helpful. I also left a request on reddit to help with Win 8 Dpi scaling bug repro, but didn't get any replies.. got upvotes though 18:52:07 beluga: yep, identifying the more esoteric bugs is helpful 18:52:32 discussed this with Joel.. he said, if bug concerns some enterprise environment, we might suggest getting paid support 18:53:17 if you sort the uncos by date and skim the oldest ones you will see they are quite a colorful bunch :) 18:53:31 we do have a couple of whiteboard tags such as needAdvice, but nothing (AFAIK) for just marking the difficult-to-repro ones 18:54:27 I will say that we got a few people at the Seattle LibreFest to help repro using different environments 18:54:34 beluga: Joel is right, we can only handle generic environments, not all professional ones 18:54:51 I've been going through them and I'm at July 2014 currently.. but I will divert my attention to the "NEW with 1 comment" ones next.. today I wanted to attack the uncos to make our count look healthier after Joel's work :) 18:55:04 News from fdonew: [Bug 85927] bubble chart wizard is not working as expected  || [Bug 85928] any operators in standardfilter not displayed 18:55:08 one woman said that she might be able to help out with a couple of harder-to-confirm bugs, including one w/Windows Phone 18:55:32 yeah distros and DEs are from the possible-spectrum, we can always fire up a virtual machine with a live distro 18:56:34 Limitations about virtualizing and providing instances of Windows and OS X used to be a barrier to QA, but from what I'm hearing, we're actually doing okay on the more common OSes front 18:56:43 colonelqubit: would you like to create an action item for yourself to ping shauna on #openhatch about participation to their events? 18:57:18 beluga: yep, I have that on my list, but I'll add it :-) 18:57:29 let's make it official :) 18:58:04 #action ping Shauna re: having LibreOffice be more involved in OpenHatch events (e.g. Open Source Comes to Campus) 18:59:19 colonelqubit: Thorsten and Bjoern have already that on their radar from what I know 18:59:31 sophi: good to hear 18:59:52 sophi: do you have any topics to bring up today? 19:00:21 sophi: yes I chatted with Björn, but he said he has been busy with other stuff.. he had an issue about customized Bugzilla queries plugged to the OpenHatch web system 19:00:41 I urged him to create an issue on the OH Github 19:01:02 colonelqubit: I've an opportunity to run a hackaton to l10n MozTrap 19:01:26 cool. Where's that going to be located? 19:01:39 colonelqubit: I'm currently waiting for an answer about the budget attribution 19:02:06 okay, sounds good 19:02:07 colonelqubit: it will be organized by a company I know in Montreal together with the local python community 19:02:31 Nice. Are you coordinating with Marc or Immanuel there? 19:02:43 colonelqubit: this community is pretty large there, the company organize the hackaton and we participate in beer an pizza 19:02:51 * colonelqubit tries to remember who else is up in Montreal :-) 19:03:06 colonelqubit: not directly, Marc is aware but not in Montreal 19:05:59 * colonelqubit nods 19:06:45 Okay, let's see... Unifex -- we haven't heard from you yet 19:06:53 How goes the fuzzing/testing? 19:07:07 colonelqubit: about the BH session, it has been announced, relayed on the social media and will be done again next week 19:07:41 Probably 2 weeks from releasing chkodf.py, and 2 more for initial fuzzodf.py. 19:07:46 sophi: thanks; I'll make a note of it in my follow-up to the LibreFest people 19:08:39 colonelqubit: ok 19:08:53 Unifex: Do you have a home for those tools yet? or perhaps you'll mention it when they're released? 19:09:40 Still localhosted on laptop, no github acccount for these tools yet, I'll announce them on the LO and AOO QA lists. 19:10:00 Unifex: could you give a brief description of what they do? 19:10:25 Excellent. It's great to see more testing being done against ODF and our codebases; it really helps them to become more mature and solid tools 19:11:02 chkodf is security tool,like OfficeMalScanner or PDF-Tools, except for ODFs, not PDF/DOC/XLS/PPT. 19:11:42 Unifex: ok 19:11:46 fuzzodf takes the ODF seedfiles from Rob's ApacheCon talk on fuzzing ODF and continues it, starting with smart fuzzing of ODF markup, starting with Flat ODFs. 19:12:57 Unifex: ok, didn't see those seedfiles, but thanks for sharing your work 19:13:54 Robinson has a copy of Rob's 250MB zip of files. Rob said it was ok for public to use them. I need to get some similar files from LO bug database, Robinson gave me a script, worried to run it. :-) 19:14:17 * colonelqubit laughs 19:14:20 Unifex: ah ok :) 19:14:40 So we just need to be careful when pulling a lot of data from a number of bug trakcers 19:14:44 *trackers 19:14:50 Rob has 2.6GB of files, mostly DOC/XLS/PDF, but i'm focusing on ODF. 19:15:06 Hmm...moggi isn't here; I was going to ask him how often we re-download all of those test files 19:16:30 Unifex: yes, if you could ask moggi tomorrow on the dev channel I'm sure you'll have what you search for :) 19:17:04 one random thought: has there been discussion with the design guys that with the self-hosted Bugzilla, the look can be customized à la Mozilla & KDE? I noticed mirek was ranting about the Bugzilla look/UX on July 2014 19:17:48 No rush on the additional LO files, i have chkodf.py coding to work on, and then Rob's initial files to deal with. Once done with Rob's files, then will be looking to get the LO files. 19:17:56 Unifex: while it might make some data-sense to download all of the files in one place, it's probably easiest from a testing/data-rights perspective for each consumer to run the script 19:18:16 Unifex: ok, no problem 19:18:43 beluga: I haven't spoken with the design folks about Bugzilla much 19:18:51 * colonelqubit pokes jphilipz 19:18:54 No downloading of LO involved in chkodf.py, that is OOB download by me. No chkodf/fuzzodf users will be causing additional downloads of LO bug DB files. 19:19:34 Unifex: oh, sure; I meant more that when we talk about files outside of our own infrastructure, we often don't have as clear a picture about redistribution rights, etc.. 19:19:39 colonelqubit: from the rant I read, it seemed they were not aware that it is possible to create our own theme for Bugzilla 19:19:39 beluga: we usually add our own design to all our own instances, so I think that will happen one day or another :) 19:20:04 first we will add green ;) 19:20:08 * colonelqubit laughs 19:20:17 baby steps.. 19:20:19 I'll dig up the email from Rob giving permission to use those AOO seedfiles.zip, and send to Robinson. 19:20:49 * colonelqubit nods 19:21:03 So speaking about the Bugzilla migration, let me give you a quick update 19:21:32 Right now we've got a test install of Bugzilla running up on a TDF VM. If any of you'd like to play with that, just drop me an email. 19:21:40 beluga: yes, instead of blue it will be all green 19:22:01 I've got a number of people who've volunteered to help out with Bugzilla administration. If anyone else is interested in helping in that role, email me as well. 19:22:28 Tollef (the friendly sysadmin at FDO) has been very helpful this week with information about their Bugzilla install 19:22:34 colonelqubit: I'll drop you a mail to be a tester 19:22:47 awesome, thanks 19:23:10 colonelqubit: you're welcome, ten years of bugzilla should help me ;) 19:24:14 Right now we have some old bugzilla data we're using for testing purposes, but Tollef will be getting us an updated database dump, so we can test with an environment that closely matches what we're using on FDO 19:25:12 News from fdonew: [Bug 85930] STYLES: Showing unused custom styles in the style drop down 19:25:12 colonelqubit: great, and you'll add our own changes then? 19:25:27 Regarding a migration date, there are multiple events such as the LibreOffice releases that will influence when we can perform the migration 19:26:50 I'll get you updated information about dates as soon as all of the stakeholders can agree :-) 19:27:12 sophi: So we do want to be a bit cautious about adding our own changes before performing the migration 19:28:04 colonelqubit: I mean in the test instance, but ok, lets test a own BZ live first 19:28:57 sophi: Right, so my current plan is to create a separate branch for changes against our bugzilla test instance in which we can put changes we'd like to make after we complete the successful migration 19:29:08 (separate branch in a git repo) 19:29:59 Starting next week I'll be open to chatting with people about changes that can go in that repo 19:30:14 colonelqubit: ok 19:30:33 Again, whatever changes we make in advance will need to be rebased onto whatever codebase we use for the actual migration 19:31:13 but this should allow someone to, say, try out a re-themeing of Bugzilla with TDF colors 19:32:35 colonelqubit: the most important currently is to have a running instance to allow the migration, so lets concentrate on that 19:32:51 To avoid confusion between FDO and our test instance, we may put some minor speedbump in front of it (e.g. use .htaccess) 19:33:07 But I'll provide more information about that in the coming days 19:34:23 colonelqubit: mail sent :) 19:34:34 Okay, any other q's about Bugzilla Migration? 19:35:48 Hi 19:36:04 A quick note about Seattle: We had a great LibreFest, with about 30 people showing up. Coffee/bagels/etc.. were all very appreciated, as was the O.J. provided by Unifex :-) 19:36:08 * colonelqubit waves to Liongold 19:36:18 Liongold: just giving a quick recap of our Seattle LibreFest 19:36:46 Liongold: hello :) 19:37:17 sophi: colonelqubit Hi. Hope it went well. 19:37:22 Big thanks to everyone who helped out the newcomers in the IRC channels. 19:37:42 Liongold: yes, thanks :) 19:37:55 Liongold: yes, it went very well :-) 19:38:10 time for dinner for me, see you! 19:38:20 sophi: Bon Apetit 19:38:27 Liongold: merci :) 19:38:31 (something like that) 19:38:43 If anyone is planning or interested in planning a future 'LibreFest' event, I'd be happy to give you more specific information about logistics/planning. 19:38:59 * colonelqubit waves to sophi 19:39:12 Liongold: Do you have any QA-related topics to bring up? 19:39:16 colonelqubit: see you! 19:39:23 did LibreFest have any impact on bug count, was it productive? I was off in the corner writing python code, not part of main group. 19:40:06 colonelqubit: Nothing. 19:41:07 Unifex: yes, I believe that the LibreFest had an impact on the bug count. I don't think it was very huge, but I think that it got more people interested in the bug hunting process 19:41:27 What I thought was great about the event was that I got a number of "Sorry I couldn't attend!" replies 19:42:12 So definitely there are a lot of people interested, and some new faces that would be interested in a follow-up event 19:42:51 It had a noticeable impact on unconfirmed count 19:43:13 :-) 19:43:35 Anyone have any other topics to bring up? 19:44:59 Okay, thanks all for attending! We'll have our next meeting in 2 weeks (same time/place) 19:45:06 Nice to see LO people in Seattle at LibreFest. Glad to hear it had some impact. Hope there's another one next time, at SeaGL/LFNW/etc. I need to leave IRC. Bye! 19:45:16 * colonelqubit waves 19:46:01 #endmeeting