From Fedora Project Wiki
< QA
12:06 -!- bpepple [n=bpepple@adsl-71-144-15-152.dsl.wotnoh.sbcglobal.net] has joined #fedora-testing 12:06 < Lovechild> wwoods: can you also a) set up a ical thing and b) warn us by mail at least a few days in advance? 12:06 * dmalcolm is in UTC-5 12:06 < n0dalus> but I realize I don't exactly represent a huge number of Fedora developers 12:06 < Lovechild> all hail Brian 12:06 <@wwoods> n0dalus: yeah, but we don't want to totally ignore you 12:06 * bpepple saw there was a QA meeting and figured I would drop by. 12:06 <@wwoods> Okay - next meeting will be at 0100UTC (8pm US Eastern) 12:07 <@wwoods> I'll try to set up a shared iCal 12:07 < n0dalus> you can do that kind of thing with Google Calendar I think 12:07 <@wwoods> and there will be mail to fedora-test-list at least.. 3 days in advance (Monday) - is that OK? 12:07 < Lovechild> going to bed at 2am on thursday.. ah what the hell, I have no life anyways 12:08 <@wwoods> okay! meeting officially starting... now 12:08 < Lovechild> wait.. I can't do math either.. UTC+1 plus a 60min meeting makes it 3am 12:08 <@wwoods> the agenda is at http://fedoraproject.org/wiki/QA/Meetings/20061214 12:08 <@wwoods> we're already on item 2 so let's keep going for a bit 12:09 < Lovechild> I would be pleased if we could alternate the meeting time at least once in a while..just in case we euro trash type people want to you know.. sleep and stuff 12:10 <@wwoods> heh! I understand... is there a good time that accomodates .eu and .au? 12:10 <@wwoods> I've been assuming most folks here are somewhere in the US, so I want one time that's good for .us and .eu, and one that's good for .us and .au 12:10 < Lovechild> I really don't think we can hit one golden time for all, if we aim at hitting the US once a month and EU once a month at least we give people the chance to partake 12:11 <@wwoods> if you all have to stay up late for me, I should at least be willing to do the same for you. heh. 12:12 < Lovechild> well.. I have no life anyways, I can be up most nights if need be 12:12 <@wwoods> so what's a good time for .eu? is this OK? 12:13 <@wwoods> I figured 1700 would make .eu and .us happy, and n0dalus voted for 0100 for .au/.us 12:13 < Lovechild> UTC 1600 would be decent, if not UTC 1900 or 2000 12:13 <@wwoods> oh yeah, 1600 would be better so we don't stomp FESCo 12:13 < Lovechild> otherwise you are hitting smack in dinner time for most of the EU 12:13 < bpepple> A different day of the week would word better for me, since I have FESCo on Thursday. 12:14 <@wwoods> friday or wednesday? 12:14 < bpepple> Either is fine by me. 12:14 < Lovechild> I would like wedensdays better, people tend to have plans for friday 12:14 <@wwoods> true 12:14 < EvilBob> thursday is the hardest day of the week for me 12:14 < bpepple> Lovechild: agreed. 12:14 < Lovechild> and I have a feeling of all crews, the QA team does not need to hold meetings on IRC drunk as all hell 12:15 <@wwoods> next meeting: 0100UTC, Wed. Dec. 20 12:15 < bpepple> wwoods: cool. 12:15 <@wwoods> after that: 1600UTC, Wed. Dec 27 12:15 < n0dalus> sounds good 12:15 < Lovechild> the plan is to meet once a week or there abouts? 12:15 <@wwoods> I may not be able to run the Dec. 20 meeting - I'm flying home that day 12:15 <@wwoods> Lovechild: yeah 12:15 < Lovechild> I can handle the dec 20 meeting if you want me to 12:16 <@wwoods> We've got 8 weeks 'til FC7t1 12:16 < n0dalus> that'll go fast I bet 12:16 <@wwoods> also, the more often the meetings are, the quicker they should be 12:16 <@wwoods> I hope. 12:16 <@wwoods> yeah, I figure we have about 6 weeks of planning / tool building 12:17 <@wwoods> and xmas/new years gobbles a week or two 12:17 <@wwoods> so, yeah. 12:18 <@wwoods> (where "xmas" is a non-denominational end-of-year holiday, ahem) 12:18 <@wwoods> anyway, reviewing stuff from last week 12:19 < Lovechild> I doubt we'll get much work commited from next meeting till after news years to be honest 12:19 <@wwoods> true, but that means we have a couple weeks for just planning stuff 12:19 <@wwoods> which is important 12:20 < bpepple> wwoods: agreed. 12:20 < Lovechild> agreed 12:20 <@wwoods> okay, reviewing from last week: I don't think there's much to talk about. poelcat was going to work on a flow diagram for the updates tool and such, but he's unfortunately leaving town to go to a funeral 12:20 <@wwoods> but we'll discuss how we think that stuff should work 12:21 <@wwoods> and he can make a nice diagram of it for next week 12:21 <@wwoods> (and maybe I'll convince him to run the meeting) 12:21 < Lovechild> sounds good 12:21 <@wwoods> okay, #3: FC6 common bugs 12:21 <@wwoods> I assume you've all seen http://fedoraproject.org/wiki/Bugs/FC6Common 12:22 <@wwoods> 1) are there other commonly-reported bugs that should be on there? 12:22 <@wwoods> and 2) what can we do to prevent them from recurring? 12:22 < Lovechild> should the dreaded yum update causes X to quit bug not be on that list or is that a non-issue now? 12:22 < n0dalus> a couple of those bugs were hit early in the testing, I'm not sure why they weren't fixed? 12:23 <@wwoods> Lovechild: oh, I think that's still happening 12:23 < EvilBob> it is still happening 12:23 <@wwoods> Lovechild: do you know of a bug # and/or workaround/fix for that? 12:23 < dmalcolm> Lovechild: some of the desktop guys figured out what was happening 12:23 < Lovechild> just a sec 12:23 <@wwoods> is it a -testing package, or a live one? 12:23 * dmalcolm looks up the bug 12:23 < n0dalus> yeah I hit that yum/X thing a week ago or so 12:23 < dmalcolm> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=218207 ? 12:24 <@wwoods> dmalcolm: cool. wow, that last comment was 25 minutes ago 12:24 <@wwoods> heh 12:24 < Lovechild> yep that is the one.. damn you beagle for being slower than dmalcolm 12:25 * dmalcolm has been installing/upgrading/watching things go boom, and gathering info on a clone of that bug 12:25 < Lovechild> I'll be back in 5 mins, I have to feed people 12:25 <@wwoods> dmalcolm: if Ray confirms that this fixes the bug, I guess we don't need to add it 12:25 < dmalcolm> halfline + soeren + ajax figured it out using systemtap, IIRC 12:26 <@wwoods> is anyone already on the cc list for that bug? 12:27 < Lovechild> back 12:27 * dmalcolm adds himself 12:27 <@wwoods> I'd like someone to watch it for changes and update the FC6Common bug if it turns out that it will remain in FC6 12:27 <@wwoods> if it's going to magically go away, let us know (on fedora-test-list or something) 12:28 < bpepple> wwoods: I'll watch it. 12:28 <@wwoods> bpepple: awesome, many thanks 12:28 < dmalcolm> (hopefully patching DBus will fix it) 12:28 <@wwoods> There's also a yum-segfault bug that I've seen a bajillion dupes against 12:29 <@wwoods> but I can't remember the bug number offhand 12:29 <@wwoods> I'll own that one 12:29 < Lovechild> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=213963 12:30 < Lovechild> ? 12:30 <@wwoods> Lovechild: yes! good catch 12:30 * Lovechild hugs bugzilla search 12:31 <@wwoods> I'll add it to the FC6Common page after I get some info from jeremy about the ETA on a fix or workaround 12:31 <@wwoods> as for preventing these from recurring.. Ideally we'd just write RHTS tests that trigger the bug 12:31 <@wwoods> e.g. install the bad package, check to see if X restarted 12:32 <@wwoods> the yum segfault is trickier 12:32 < n0dalus> we could make some sort of stress testing tool for yum 12:32 < Lovechild> a crashme type thing? 12:32 <@wwoods> n0dalus: hmm! that would be an excellent test for yum 12:33 <@wwoods> yum is one of the most important parts of any release - pretty much anything else that's broken can be fixed if yum works 12:33 < n0dalus> yeah 12:34 < Lovechild> it's also sadly one of the hardest bits to stress test automatically, it tends to go boom mostly when the package is bad which is not really wrong. 12:34 <@wwoods> is there anything like this already? does anyone want to volunteer to hack at it (or ask skvidal about how to do it)? 12:35 < dmalcolm> I started hacking on one, but am totally doomed timewise 12:35 * dmalcolm finds a link to the source code 12:35 < n0dalus> it should probably use the cli interface and not any bindings, since the bindings might hide bugs in the cli stuff 12:35 <@wwoods> n0dalus: right 12:35 <@wwoods> so probably we want to set up a fake local yum repo and try installing a gazillion packages in weird ways 12:36 <@wwoods> or something 12:36 <@wwoods> using.. a shell script that calls yum -y? was that what you were thinking, n0dalus? 12:36 < dmalcolm> n0dalus, wwoods: https://testing.108.redhat.com/source/browse/testing/trunk/rhts/tests/sandbox/yum/smoketest/ 12:36 < n0dalus> wwoods: something like that 12:36 < dmalcolm> https://testing.108.redhat.com/source/browse/testing/trunk/rhts/tests/sandbox/yum/smoketest/run-yum-smoketest?rev=229&view=markup 12:37 < dmalcolm> I got as far as code that generates packages and yum repos (from scratch) with known properties 12:37 < dmalcolm> e.g. def make_local_repo_depchain_test(chainLength, isTerminated, addUnrequired): 12:37 < dmalcolm> """ 12:37 < dmalcolm> Generate a yum repo consisting of: 12:37 < dmalcolm> test-package-base which has a requirement on test-requirement-chain-001, 12:37 < dmalcolm> which starts a chain up to test-requirement-chain-<count> 12:37 < dmalcolm> plus test-unrequired-package outside the chain 12:37 < dmalcolm> """ 12:37 < dmalcolm> which works 12:38 <@wwoods> cool! anyone wanna take this and run with it? 12:38 < dmalcolm> but never got it to actually run yum upon these test repos 12:39 < dmalcolm> I can help anyone who wants to pick up this code and run with it, but I haven't the time to hack on it myself ATM 12:39 <@wwoods> I can try to poke at it in my Copious Spare Time (cough) 12:39 < Lovechild> maybe we could advertise for some help here? 12:39 <@wwoods> I'll see if skvidal has any suggestions, too 12:40 <@wwoods> yeah, I think this is something we could bring up with -devel and on -test-list and see if anyone can help 12:40 < dmalcolm> please CC me when you do, I'm hopelessly behnd on those lists 12:40 <@wwoods> sure, I'll do that 12:41 <@wwoods> okay, moving on (so we don't run really long) 12:41 < n0dalus> would we generate a set of mostly empty rpms with semi-complex dependencies, or would it be easier to just randomly remove and add non-vital rpms from the fedora packages on a running system 12:42 <@wwoods> n0dalus: Eventually, both, but whichever is easier should be first 12:42 <@wwoods> sounds like dmalcolm was working towards the first approach 12:42 <@wwoods> so that might be the easier one 12:42 < n0dalus> ok 12:42 < dmalcolm> yes, my approach as the first of n0dalus' suggestions 12:42 <@wwoods> Anyway, so, I mentioned the proposed F*7 schedule earlier 12:43 < dmalcolm> I wrote the rpmfluff makes it easy to create small broken RPMs 12:43 <@wwoods> it was outlined by Max here: https://www.redhat.com/archives/fedora-advisory-board/2006-December/msg00057.html 12:43 < dmalcolm> s/rpmfluff/rpmfluff tool, which/ 12:43 < Lovechild> how does that schedule stack up against major projects like GNOME? 12:44 <@wwoods> The subsequent discussion (https://www.redhat.com/archives/fedora-advisory-board/2006-December/msg00069.html is a really good link) outlined the need for better release criteria 12:44 < dmalcolm> re the schedule: I weighed in on that thread already, but here goes: where in the schedule (if anywhere) do we try to define what high-level goals we have for it? 12:44 <@wwoods> Lovechild: I'm not sure, actually. I think that, normally, we try to align with GNOME releases pretty closely 12:44 <@wwoods> dmalcolm: yeah, that's a big question-mark right now 12:45 < Lovechild> wwoods: March 14th GNOME 2.18.0 Final Release! 12:45 <@wwoods> at some point, there needs to be a meeting to decide what features are going into F*7, and what priority each feature has 12:45 < Lovechild> I would really like to have test3 include the final 2.18.0 release, that will encourage people to jump on test3 and help us out 12:46 < Lovechild> and the current plan allows for that, test3 is set for MArch 27 which means freeze if on the 20th.. plenty of time to get the packages in 12:46 <@wwoods> Lovechild: I'd like that too - we should definitely bring that up when planning happens 12:46 <@wwoods> I think they've accounted for that already but I'd like that to be made explicit 12:46 <@wwoods> normally they'll say so when they put out the final schedule 12:47 < Lovechild> I brought it up to Jesse Keating last time around and he tended to agree 12:47 <@wwoods> right, and that's what we did with FC6 (if memory serves) 12:47 <@wwoods> we used the late beta packages until we had the final ones 12:47 < Lovechild> onlyu because of slips 12:47 < Lovechild> you should do so intentionally this time 12:48 < EvilBob> I would just like a fedora release where "hey everything works" is the big hype item not the inclusion of some half working 'new tech' 12:48 <@wwoods> I think the GNOME releases are usually stable enough that we can have it be a required feature for F*7 12:48 <@wwoods> EvilBob: yeah, totally. This is why I want to have a feature planning meeting 12:48 <@wwoods> where we decide which features we can just drop if they don't work by test3 12:49 < Lovechild> might I make a radical suggestion, some time ago I was talking to Jesse about getting good bug reports. It turns out that we don't even enable the debuginfo repos for development series.. I think we should take it a st ep further, hack bug-buddy to point at our bugzilla and install debuginfo packages by default for development 12:49 <@wwoods> and which ones are drop-dead absolutely necessary (or very, very likely to work just fine) and we can slip the release until they work 12:49 <@wwoods> Lovechild: GNOME has told us that they don't want us to hack bug-buddy like that 12:49 <@wwoods> they want the reports to go to them 12:49 <@wwoods> I agree about installing debuginfo by default in devel 12:49 < n0dalus> Lovechild: I don't know what kind of net connection you have, but I find it hard enough to download all the updates and test releases as it is 12:50 < Lovechild> I have 4Mbit.. besides, it's part of the price you pay for running devel. 12:50 <@wwoods> but I think we might need a fedora-specific bug tool (or plugins/modifications to bug-buddy) 12:50 <@wwoods> so when something crashes you can click the "download debuginfo and restart" button 12:50 <@wwoods> and then you can attempt to reproduce the problem 12:50 <@wwoods> and then it will send reports to GNOME *and* Fedora 12:50 < Lovechild> the problem is recreating a crash in most cases 12:51 < dmalcolm> you don't need to reproduce: detach gdb, download debuginfo, then reattach 12:51 < n0dalus> you shouldn't need to restart after getting debuginfo 12:51 < dmalcolm> heh 12:51 < n0dalus> :) 12:51 <@wwoods> dmalcolm: oh, so it can be a "download debuginfo and get backtrace" button? 12:51 < Lovechild> at the very least if it's hard to time but happens often if we have debug symbols by default, then we get a decent backtrace 12:51 < dmalcolm> wwoods: I believe it's implementable, yes 12:51 <@wwoods> that's excellent. and that's a really good case for a fedora-specific bug tool 12:52 < dmalcolm> and you could parse the backtrace, figure out what binaries are involved, and then figure out which debuginfo RPMs are needed 12:52 < n0dalus> is bug buddy enabled in standard Fedora releases? 12:52 <@wwoods> does bug-buddy have any kind of plugin arch or would we need to add that (or roll our own tool)? 12:52 < Lovechild> I can live with that, I just worry about the amount of un-debugable reports any project gets 12:52 < Lovechild> n0dalus: no.. but it should be 12:52 < Lovechild> wwoods: I don't think it does, but we might be able to poke the maintainer in a polite manner 12:52 <@wwoods> It would be really cool to work with them to add something like that 12:53 < dmalcolm> I think that the way to do this is to implement it in a modular way, so it can be of use to every distribution, with a Fedora-specific backend 12:53 <@wwoods> dmalcolm, Lovechild: agreed 12:53 < dmalcolm> that way, upstream bugbuggy is more likely to buy-in 12:53 < dmalcolm> s/bugbuggybugbuddy/ 12:53 <@wwoods> does anyone know the upstream maintainer? 12:54 < Lovechild> only by name, but he seems like a nice fella.. I'm no big coder though so I couldn't hack up a prototype to show off 12:55 < n0dalus> Fernando Herrera <fherrera@onirica.com> 12:55 < n0dalus> according to the changelog 12:55 < Lovechild> yep 12:55 <@wwoods> Well, I'll see if I can talk to him about the idea - could be something other folks have requested already 12:56 <@wwoods> maybe he's already started work on it. who knows. 12:56 < n0dalus> hmm time is almost over isn't it? 12:56 <@wwoods> yeah, close to it 12:56 -!- mspevack is now known as mspevack_out 12:56 < Lovechild> bug buddy is getting a lot of attention lately, we can even attach scripts to fetch extra information and stuff now 12:56 < Lovechild> so it wouldn't surprise me if we could convince him to help out 12:56 <@wwoods> Lovechild: awesome, sounds really good 12:57 < dmalcolm> (big privacy debate, and what happens in the script crashes) 12:57 < dmalcolm> s/in/if/ 12:57 < n0dalus> if we can attach scripts even that would be enough, but it would be nice of course to have an interface more suited to what we want 12:58 < Lovechild> I'm mostly worries about what happens if someone edits the script to like rm -rf ~ 12:58 <@wwoods> As for the features / release criteria stuff, I'll mail f-a-b and emphasize that we want to see a feature-planning meeting 12:58 < n0dalus> couldn't stick a rm -rf ~ into any script though? 12:58 < Lovechild> well what do I know.. I'm a retard 12:59 <@wwoods> the script should be root-owned, so nobody should be messing with it except the bug-buddy-fedora maintainer 12:59 < dmalcolm> it could also run inside its own SELinux domain, for that matter 12:59 <@wwoods> dmalcolm: indeed 12:59 < dmalcolm> so that it can only ever read data, never modify anything 13:00 < Lovechild> sounds like a perfect solution.. 13:00 < Lovechild> so who wants to bother poor fer? 13:00 < dmalcolm> (there are privacy issues as well) 13:00 <@wwoods> a quick poll: do we need more categories for features (beyond "must-have, will slip release dates for" and "non-essential, will be dropped if broken")? 13:00 < Lovechild> dmalcolm: no more than we have on regular backtraces in terms of picking up urls, passwords, usernames.. 13:01 < n0dalus> there are potential privacy issues with anything like that though -- even backtraces can be parsed for things like IP addresses and possibly other things 13:01 <@wwoods> I guess maybe that depends on the definition of "broken", doesn't it? 13:01 <@wwoods> there's some things we can include as-is, even if it's not working right, and fix with an update after release 13:01 < Lovechild> wwoods: by any standard of broken Evolution is a showstopper there 13:01 < dmalcolm> Lovechild: yeah - I believe bug-buddy already tries to clean things up: http://cvs.gnome.org/viewcvs/bug-buddy/src/forbidden-words.h?rev=1.2&view=markup 13:02 <@wwoods> Lovechild: hah! Yeah, I dunno, it hasn't crashed for me, personally, in weeks 13:02 < Lovechild> dmalcolm: you are right I just tried the version in Development and it does ask me to review for personal information 13:02 <@wwoods> but you make a good point 13:02 * dmalcolm hears "evolution" and runs and hides 13:03 < Lovechild> wwoods: we need to define broken a bit better and also a set of core packages where core functionality must exist (like Evolution which on FC6 doesn't send mail on many setup currently) 13:03 * bpepple actually had many problems with Evolution, though he seems to be in the minority. 13:03 < EvilBob> Evolution Un-Broken? Never! 13:03 <@wwoods> how do we decide between "buggy but still good enough to stay in the distro" and "so horribly broken that it must be removed until it works" 13:03 < n0dalus> wwoods: things like evolution are used by too many people to remove 13:03 <@wwoods> Lovechild: http://fedoraproject.org/wiki/QA/ReleaseCriteria was my attempt to define Core Functionality 13:04 < EvilBob> bpepple: one rule here when it comes to Evilution as we call it NEVER again 13:04 <@wwoods> but yeah, I agree with n0dalus - it should include core apps like Firefox and Evolution 13:04 <@wwoods> things we depend on and cannot ship without 13:04 < Lovechild> I would like to include something relating to the primary purpose of an app being functional.. such as "Evolution sends/recieves mail without blowing up" 13:05 < Lovechild> not just starting up.. starting up is a bad test 13:05 <@wwoods> Lovechild: right, but it's a very basic smoke test (and easily tested) 13:05 <@wwoods> I didn't have time during the FC6 test cycle to write up a list of core apps and short functional test plans for each one 13:05 < Lovechild> yet it did slip for Evolution the last time 13:06 < EvilBob> Heck having Evolution actually delete mail from an imap server would have been nice not that long ago 13:06 < Lovechild> I can help with that test matrix for FC7 13:06 <@wwoods> Did it? I don't remember slipping for Evolution itself 13:06 <@wwoods> Lovechild: did you see the test template I wrote? (http://fedoraproject.org/wiki/QA/HowToTestTemplate) 13:06 < Lovechild> wwoods: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=217567 13:06 <@wwoods> I was thinking we could use that to write functional tests for packages 13:06 -!- rdieter [n=rdieter@sting.unl.edu] has left #fedora-testing ["Konversation terminated!"] 13:07 < Lovechild> also how would people feel about the concept of Fedora Love days, get users together once every 2 weeks or so to clean up bugzilla, confirm bugs 13:08 < n0dalus> is a good idea I think 13:08 < Lovechild> I would be happy to cheerlead the effort and give my time 13:08 <@wwoods> The basic idea is that we write a really simple document - it will be all stuff that seems obvious - to test something like Firefox (Start browser. Load slashdot.org. etc) 13:08 <@wwoods> so later we can automate that process 13:08 < n0dalus> it'd be nice if we could give some sort of recognition to people who help out with things like that 13:08 < bpepple> wwoods: +1 13:08 <@wwoods> Lovechild: excellent! we try to do those on fridays 13:08 <@wwoods> historically the BugZappers stuff does that every Friday in #fedora-triage, but that's kind of half-dead 13:09 <@wwoods> we could just have people come here for that sort of thing (or #fedora-qa, if I can ever get the channel from mharris) 13:09 < Lovechild> okay, I feel rather useless so I'll stick my head on the block and try to rally people for that 13:09 <@wwoods> Lovechild: but anyway, yes, it's a fantastic idea and I love it to bits 13:09 <@wwoods> heh 13:09 <@wwoods> we were talking before about having a "Bugzilla RPG" 13:09 < n0dalus> lol 13:10 <@wwoods> like GNOME has, where you get points for closing dupes and triaging bugs and such 13:10 <@wwoods> also points for writing HowToTest docs and things 13:10 < Lovechild> well not being a great coder or anything.. I can only help out that way, but I do have the advantage of having all the time in the world 13:10 < Lovechild> kinda like Launchpad' Karma concept.. 13:11 <@wwoods> It doesn't take mad hax0r skillz to write test docs, mostly just time and patience to write down, you know, "how to make sure that firefox is working right" 13:11 <@wwoods> oh they have something like this too? Interesting 13:11 <@wwoods> what does Karma get you? 13:11 <@wwoods> other than.. yknow.. Karma 13:11 * Lovechild hates firefox with the fury of a million suns 13:12 < Lovechild> it's basically just an ego thing I guess.. it gives you status amongst the Launchpad users, nothing more. 13:12 <@wwoods> aha 13:12 <@wwoods> then yes, exactly like that 13:12 < n0dalus> maybe we can put the top names in a prominent place? 13:12 < Lovechild> but one would be surprised what people will do for an ego boost.. 13:12 <@wwoods> although I personally want to send out stuff to people with high karma now and then 13:12 <@wwoods> n0dalus: oh absolutely 13:12 < n0dalus> or add them to a file in /usr/share/doc/fedora-... 13:13 < n0dalus> big thanks file or something 13:13 <@wwoods> a big ol' leader-board, each user gets a page that shows their karma, any trophies they may have received 13:13 <@wwoods> n0dalus: ooh! that's another good idea 13:13 < n0dalus> I personally would find it cool to have my name somewhere on the filesystem 13:13 <@wwoods> definitely 13:13 < Lovechild> add the top 10 to the "about Fedora" thing, the Fedora project would like to thank the following community heroes for helping us bring you F7? 13:14 <@wwoods> Lovechild: I can't promise that - the developers and extras packagers and so on would all want in - but something like it would be completely awesome 13:14 <@wwoods> but maybe the devs and packers and stuff can run their own karma system. or ours can be Fedora-wide! 13:14 < Lovechild> well I'd be happy if we got a link to the status page from there or something 13:15 <@wwoods> except then things get dicey. how do you decide whether making a new package is more valuable than fixing three bugs? 13:15 <@wwoods> I definitely want to have some accolades for the QA Team though 13:15 <@wwoods> a link from About Fedora would probably be fine 13:16 < Lovechild> sometimes closing 3 bugs is worthless dup removal whereas a new package is always good 13:16 -!- craigt [n=cthomas@static-68-236-219-35.nwrk.east.verizon.net] has joined #fedora-testing 13:16 < Lovechild> other times 3 bugs can be incredibly hard work.. you're right hard job to determine 13:16 <@wwoods> Lovechild: right, but sometimes closing 3 bugs involves weeks of intense debugging, and a new package just involves modifying a pre-existing spec 13:16 < Lovechild> however if we add points for comments to those bugs, that would tend to even out the score right 13:16 <@wwoods> it's really hard to compare the two, so I'd prefer not to cause wars between devel/extras/QA about who's more important 13:17 <@wwoods> although maybe that's what people want 13:17 <@wwoods> competition! whee! 13:17 < Lovechild> if you helped debug a problem, then you are likely to have lots of comments on the relevant bugs 13:17 < n0dalus> and larger comments too 13:17 <@wwoods> yeah, commenting on a bug should get you some small amount of karma/XP/points, but it need to be diminishing - can't have people gaming the system 13:17 < Lovechild> say if you set the provides information bit you get 5 points 13:17 < n0dalus> attachments 13:18 < Lovechild> but lets not spend to much time on that, it's just fun and games - the object is to make Fedora better 13:18 < n0dalus> wwoods: I think it would be fairly obvious if someone was trying to submit lots of junk changes. What do we do with people that do that already? (does anyone ever do that) 13:18 < Lovechild> we only have 8 weeks till test1, we should spend them wisely 13:19 <@wwoods> n0dalus: nobody does that, because there's very little incentive to make/change bug reports at all 13:19 <@wwoods> other than the hope of getting a bug fixed 13:19 <@wwoods> and a desire to make fedora more awesome 13:20 < Lovechild> on sidenote I'll reward whom ever allows me to stab those blog comment spammers in the face via the internet 13:21 <@wwoods> I've been working getting facepunchd into Fedora 13:21 <@wwoods> very little traction so far :/ 13:22 <@wwoods> it seems people don't *want* to be punched in the face, *especially* when they need it 13:22 < n0dalus> there's probably worse things in Fedora already 13:22 < EvilBob> bugzilla is the least fun internet game ever 13:22 <@wwoods> yeah, I'd like bugzilla to be a little more like kingdom of loathing 13:23 <@wwoods> or something 13:23 < EvilBob> as a user it sucks because many times you file a bug and never know if anyone even looked at the dang thing 13:23 <@wwoods> EvilBob: yeah! it's terribly frustrating 13:23 < Lovechild> like callion when you file bugs on his extras packages.. 13:24 <@wwoods> but if there was more of us going through bugzilla and trying to respond to bugs 13:24 <@wwoods> and pinging developers about easy-fix bugs 13:24 <@wwoods> and such 13:24 <@wwoods> it would suck less. 13:24 < n0dalus> the less bugs there are open the faster everything will work 13:24 <@wwoods> and if doing that was kinda fun, or had some kind of social aspect attached, maybe we'd have more people 13:24 < Lovechild> I hope we can get the Love days up and running, that would help greatly on that 13:24 <@wwoods> so, yeah 13:24 <@wwoods> Lovechild: definitely 13:25 <@wwoods> should we have them here, or keep the traditional #fedora-triage location? 13:25 < Lovechild> I'd like it to be here.. this is -testing after all 13:25 < n0dalus> I know people were looking a while ago into adding more features to redhat bugzilla? like the ones on gnome's for example 13:25 <@wwoods> fair enough 13:25 <@wwoods> n0dalus: yeah, the bugzilla maintainer for redhat sits in the next cube over 13:26 <@wwoods> but I haven't bothered him about this stuff yet 13:26 <@wwoods> mostly I want to pull info out of bugzilla and have the Bugzilla RPG live somewhere outside bugzilla 13:26 <@wwoods> so that it can pull data from many sources (the wiki, test tools, etc) 13:26 < n0dalus> interesting 13:26 < Lovechild> It would be cool to find some kind of silly prize for the "bugzilla cowboy of the week" 13:26 <@wwoods> Lovechild: not a bad idea! 13:27 <@wwoods> speeking of bugs-of-the-week: http://bugzilla.redhat.com/bugzilla/buglist.cgi?cmdtype=runnamed&namedcmd=FC6-thisweek 13:27 < n0dalus> does red hat have some cheap hats or t-shirts or something they can send out to 10 people a year or something? 13:27 <@wwoods> n0dalus: oh definitely 13:27 <@wwoods> all sorts of funky stuff 13:28 <@wwoods> and, of course, free FC media and such 13:28 <@wwoods> http://redhat.brandfuelstores.com/ 13:28 <@wwoods> but I wish we had more fedora-specific stuff 13:28 < Lovechild> is there an easy way to create a list of bugs filed against "release" in the past week - it would be helpful for the love days, limiting the amount of reports and all 13:28 < n0dalus> wwoods: that FC6-thisweek link only works for your account I think 13:28 <@wwoods> oh wait, I made it public 13:29 <@wwoods> https://bugzilla.redhat.com/bugzilla/buglist.cgi?cmdtype=runnamed&namedcmd=FC6-thisweek&namedowner=wwoods@redhat.com 13:29 <@wwoods> https://bugzilla.redhat.com/bugzilla/buglist.cgi?cmdtype=runnamed&namedcmd=FC5-thisweek&namedowner=wwoods@redhat.com 13:29 <@wwoods> the second link is for FC5 13:29 < EvilBob> I was shocked at the quality of the shirts and hats from BrandFuel 13:29 <@wwoods> I have to keep reminding myself that we still support that 13:29 < EvilBob> The quality was a lot better than I actually expected 13:30 < n0dalus> speaking of handy bug links, I have some Firefox quick url things you can import from http://fedoraproject.org/wiki/KieranClancy 13:30 <@wwoods> oh nice! 13:31 < EvilBob> Cool 13:31 < n0dalus> I use them all the time 13:31 <@wwoods> does anyone want to be Head Cheerleader for the bugzapper / Fedora Love day stuff? Lovechild? 13:32 <@wwoods> (we're over 90 minutes now, so we should start cleaning up) 13:32 < Lovechild> wwoods: I'll be sure to wear my pom-poms and look really stupid 13:32 <@wwoods> haha 13:32 <@wwoods> hooray! 13:33 <@wwoods> I'm in here from 10-6 EST on Friday (I guess that's.. 1500-2100UTC), so it would be good to have folks in other time zones who will be around for different parts of the day 13:34 <@wwoods> oh, you guys know about #fedorabot right? 13:34 < Lovechild> I figure Fri 5th of January might be a good day to start things off, no holidays to mess it up and people should be sober.. plus it gives me time to set things up and send out mails 13:34 <@wwoods> Lovechild: sounds excellent 13:34 < Lovechild> fedorabot.. do not ring a bell 13:35 <@wwoods> there's a bot in #fedorabot that sends a message whenever someone creates a new bug / closes a bug / etc. 13:35 <@wwoods> it's really good for monitoring current bugzilla goings-on 13:35 < Lovechild> cool.. 13:35 < n0dalus> it's in my auto-join thing 13:35 < BobJensen-Away> #FedoraBot :You can't join that many channels 13:35 < BobJensen-Away> Oh well 13:35 <@wwoods> BobJensen-Away: haha no way! bummer 13:35 <@wwoods> you have officially spread yourself too thin! 13:35 < BobJensen-Away> lol 13:36 <@wwoods> so yeah, testing-type folks will always be in here, but every couple of fridays we will have official Love Days for cleaning up bugzilla 13:36 <@wwoods> or for working on bugs of a specific type/component 13:37 < Lovechild> BobJensen-Away: it's IRCs way to telling you to slow down.. like having a heart attack 13:37 < BobJensen-Away> Lovechild: too many points of interest I guess 13:38 <@wwoods> Lovechild: I'll let you decide when, send out announcements, decide on themes (if you want to have one), etc. 13:38 <@wwoods> Should we have people in #fedora who should be submitting bugs come here for help? 13:38 < Lovechild> wwoods: will do 13:39 <@wwoods> Lovechild: awesome, many thanks 13:39 < Lovechild> well, the more helping hands we can get really 13:39 <@wwoods> oh absolutely 13:39 <@wwoods> I'll be there whenever 13:39 < BobJensen-Away> training users how to file good bugs IMO is important 13:39 <@wwoods> just say the word 13:39 < n0dalus> been nice talking with everyone, but I should probably be going now (5am) 13:39 <@wwoods> n0dalus: yeah no doubt! get some sleep! 13:40 < Lovechild> also I can't stay awake for 24hours, so I'll try to find someone to eventually take the other two - three shifts so there will always be someone to help out newcomers 13:40 <@wwoods> We should probably call this meeting over. We didn't get to discuss QA workflows or QA tools but we can do that next week (when poelstra is around) 13:40 <@wwoods> Lovechild: definitely 13:41 < Lovechild> okay.. thank you all for the great time.. it's time for me to stuff my face with food 13:41 <@wwoods> same here 13:41 < n0dalus> I won't be able to come next time, but hope you guys have a productive discussion 13:41 <@wwoods> yeah, I may or may not be here 13:41 <@wwoods> but I trust poelstra to come up with sane decisions on how the tools should work 13:41 < Lovechild> I'll stay up just to be here.. even if it means playing n0dalus and staying up till 5am 13:42 <@wwoods> dmalcolm too - they're at Red Hat and have been an incredibly important part of building the mighty test infrastructure we have 13:43 <@wwoods> I'm really excited about getting them and the testing.108.redhat.com stuff to be an official part of the Fedora QA process 13:43 <@wwoods> better tools = less work, better Fedora, more fun 13:44 <@wwoods> anyway! I will post the logs ASAP, and some condensed meeting notes later today 13:44 <@wwoods> thanks for staying up so late to talk, guys! 13:44 < n0dalus> no worries