== Log == UTC -6 {{{ Jun 04 12:00:43 --> You are now talking on #ubuntu-testing Jun 04 12:00:43 --- Topic for #ubuntu-testing is Testing of Ubuntu | Hardy Final testing, download script: https://wiki.ubuntu.com/Testing/Isoscript | Wiki: https://wiki.ubuntu.com/Testing | Reports: http://qa.ubuntu.com | Your ideas are welcome: http://brainstorm.ubuntu.com Jun 04 12:00:43 --- Topic for #ubuntu-testing set by stgraber at Tue Apr 22 12:25:42 2008 Jun 04 12:00:48 --> persia (n=persia@ubuntu/member/persia) has joined #ubuntu-testing Jun 04 12:00:57 * ogasawara waves Jun 04 12:01:06 * cgregan hello! Jun 04 12:01:10 ahoj! Jun 04 12:01:15 --> ecubuntu_bb (n=ecubuntu@32.158.223.204) has joined #ubuntu-testing Jun 04 12:01:18 hello Jun 04 12:01:23 hi Jun 04 12:01:37 hey Jun 04 12:02:03 hello all! Jun 04 12:02:17 What type of rain in Portland today? :-) Jun 04 12:02:36 * bdmurray cries Jun 04 12:02:43 I also have rain here :-(( Jun 04 12:02:49 here too Jun 04 12:02:55 bdmurray: it is rose festival, it's supposed to rain now. Jun 04 12:03:00 greetings from sunny Oxford :) Jun 04 12:03:02 I had rain an hour ago Jun 04 12:03:06 Its been raining all year though Jun 04 12:03:10 --> effie_jayx (n=effie_ja@ubuntu/member/effie-jayx) has joined #ubuntu-testing Jun 04 12:03:12 now it's sunny again (maybe a bit too much ? :)) Jun 04 12:03:12 thunder and lightning as a bonus here Jun 04 12:03:25 hello from sunny, sunny wolverhampton :) Jun 04 12:03:49 my home town of Bergen, Norway had 100+ days of consecutive rain days last year Jun 04 12:04:07 are we having the meeting here today then as there already seems to be one in the #ubuntu-meeting Jun 04 12:04:18 davmor2: yep Jun 04 12:04:20 ok, shall we start? Jun 04 12:04:25 heno: wow..the most I saw was 90 straight in Portland Jun 04 12:04:41 gosh and i'm complaining for having a week of it Jun 04 12:04:48 we fittingly have an agenda item about meeting times at the end Jun 04 12:04:58 #startmeeting Jun 04 12:05:21 TOPIC: Intrepid QA specifications - Decide which specifications we actually target for Intrepid Jun 04 12:05:58 let's work our way down https://wiki.ubuntu.com/QATeam/Specs Jun 04 12:06:34 * kernel-bug-migration is well under way Jun 04 12:06:38 sorry for not writing mines (except from the subject) but it's a busy week here :( Jun 04 12:06:47 I'm from my cellphone . I'll be in my desktop in a couple of minutes Jun 04 12:07:24 * fixed-bug-verification - this needs some updating for intrepid Jun 04 12:07:56 but basically it's sbeattie taking over this task from pedro_ this cycle Jun 04 12:08:19 we should also move the tracking from the wiki to something better Jun 04 12:09:08 * useful-bug-metrics - spec looks good, though I'd like to see a few more metric examples Jun 04 12:09:28 bdmurray: do you agree with my latest updates there? Jun 04 12:10:36 heno: yes Jun 04 12:11:18 bdmurray: ok, can you flesh that out with some more examples? Jun 04 12:11:34 heno: suer Jun 04 12:12:17 * self-testing-desktop - this is an optional item for Ara after desktop test automation, Low priority target Jun 04 12:13:09 * mobile-automated-tests - needs a bit more drafting IMO, use cases and such Jun 04 12:13:35 heno: Yes...I plan to put some more time into it this week Jun 04 12:13:54 cgregan: do you need some guidance with that? we could do a phone call about it. let me know Jun 04 12:13:54 put gobby notes in from UDS, but need to flesh out Jun 04 12:14:02 ok Jun 04 12:14:09 --> LaserJock (n=LaserJoc@ubuntu/member/laserjock) has joined #ubuntu-testing Jun 04 12:14:15 heno: I will look at others for inspiration Jun 04 12:14:27 If I have questions I will ping you Jun 04 12:14:41 right. I know mobile is in a crunch ATM Jun 04 12:14:58 with your wonky release schedule :) Jun 04 12:15:18 heno: hehe...true...but I should still have a couple hours to put toward the spec Jun 04 12:16:11 * desktop-automated-tests - needs updating post UDS Jun 04 12:16:23 I don't think Ara is here today Jun 04 12:16:48 <-- effie_jayx has quit (Remote closed the connection) Jun 04 12:16:54 pedro_: could you help her merge in some notes from UDS? Jun 04 12:17:06 --> yuriy (n=yuriy@ubuntu/member/yuriy) has joined #ubuntu-testing Jun 04 12:17:16 We will focus on LDTP AFAIK this time, etc. Jun 04 12:17:18 heno: sure Jun 04 12:17:25 pedro_: thanks Jun 04 12:17:27 <-- ecubuntu_bb has quit ("used jmIrc") Jun 04 12:18:11 * stgraber has to go afk for a minute Jun 04 12:19:23 * fix-validation-tracking - should we not merge this in with 'fixed-bug-verification'? There is a slight difference, but it will cause endless confusion as it is Jun 04 12:19:41 the infrastructure we use to track it might be the same Jun 04 12:20:01 heno: if I may, I've been working on a script for MOTU SRU to better keep track of the SRUs Jun 04 12:20:13 * stgraber is back Jun 04 12:20:30 LaserJock: ok great, it that viewable somewhere? Jun 04 12:20:32 I would be happy to contribute it or work with somebody on a plan Jun 04 12:20:53 heno: current results are at http://laserjock.us/files/ubuntu/bug_list.html Jun 04 12:20:54 LaserJock: I'm interested Jun 04 12:21:01 cool, perhaps it should be merged with http://people.ubuntu.com/~ubuntu-archive/pending-sru.html Jun 04 12:21:27 heno: yeah, I was planning on having a couple pages for different SRU issues Jun 04 12:21:36 LaserJock's one looks better :) Jun 04 12:21:44 heno: I think those two pages serve different purposes. Jun 04 12:21:49 is it on UbuntuWire? Jun 04 12:22:02 if bdmurray's -proposed bugs tag goes through I'd like to add support for that as well Jun 04 12:22:03 One helps track aging of current -proposed, and the other helps track work-in-progress pre-proposed Jun 04 12:22:20 --> effie_jayx (n=evalles@ubuntu/member/effie-jayx) has joined #ubuntu-testing Jun 04 12:22:23 heno: not yet, I've just been recently hacking on it. I wanted it a little further along before making it "live" Jun 04 12:22:30 LaserJock: one thing we do is tag bugs if they are hardware specific does your script look at tags at all? Are many *verse bugs hardware specific? Jun 04 12:22:38 it would be nice to track the issues before they go into -proposed where possible and they should go into an archive after release Jun 04 12:23:07 bdmurray: I do look at tags (verification-done for instance), and no, we don't often have hardware specific bugs Jun 04 12:23:17 my script uses python-launchpad-bugs Jun 04 12:23:26 so kinda whatever you can get out of that we can do Jun 04 12:23:29 persia: ok, it would be great to cover that in one place Jun 04 12:24:09 heno: Quite possibly, or at least one portal with a few views. Might want to confirm with the archive-admins just in case they have some other use for pending-sru.html Jun 04 12:24:32 indeed Jun 04 12:24:43 for me as a MOTU SRU person there's sort of 3 things I want to track. 1) stuff that's not on my radar and maybe should be. SRUs that perhaps having had our team subscribed or bugs in -proposed packages Jun 04 12:24:58 2) The current list of packages needing approval Jun 04 12:25:16 3) list of packages needing testing so we can move that along Jun 04 12:25:31 and of course knowing when their done so we make sure pitti gets to them Jun 04 12:26:42 that's what I'm working on and would be happy to contribute what I've got and a bit of time to work on it with people if they'd like Jun 04 12:26:56 none of these pages actually track test submissions as the qa.u.c site currently does Jun 04 12:27:21 yes, that's the key that I'd love to see Jun 04 12:27:29 should we simply make a template for tracking that on the bug itself Jun 04 12:27:32 currently testing can get a bit wild Jun 04 12:27:47 right now the SRU policy has people write a Test Case Jun 04 12:27:49 it could contain key words that we could grep for Jun 04 12:28:04 Or we could use tags in some way Jun 04 12:28:11 and then display the result in summary for on the page Jun 04 12:28:32 what I would love would be to have a qa.u.c tracker grab that Test Case and display it and let people submit results there Jun 04 12:28:52 I was thinking the test case could also be an attachment with a standarized name Jun 04 12:29:07 I suspect that would be easier to pull out. Jun 04 12:29:12 python-launchpad-bugs could be modified to to get that attachment Jun 04 12:29:36 bdmurray: The problem with attachments is that they are hard to see except programatically. Jun 04 12:29:59 right. There is benefit to storing everything on the bug and then pulling it out for display Jun 04 12:30:13 <-- nand has quit ("leaving work") Jun 04 12:30:21 hmm, could we extend the greasmonkey scripts to inline testcase attachments? Jun 04 12:30:30 but you can use search in your browser :) Jun 04 12:30:35 I think there must be a technical solution where you can do both Jun 04 12:30:45 sbeattie: That doesn't help lots of people who don't use the greasemonkey scripts :( Jun 04 12:31:05 Whether it be greasemonkey or python-launchpad-bugs putting the attachment in the description Jun 04 12:31:15 Maybe someone could check with the LP team about a clean solution that works both programmatically and visually? Jun 04 12:32:20 ok, this clearly needs more discussion; let's take it to the mailing list or after the meeting Jun 04 12:32:45 sbeattie: can you organise that and get input from everyone? Jun 04 12:32:52 heno: yes. Jun 04 12:32:58 anyway, I'd like to offer what I can for this project Jun 04 12:33:18 though that may not be a whole lot :-) Jun 04 12:33:36 I think we can agree that this will be targeted this cycle, but some details still need sorting out Jun 04 12:33:49 LaserJock: one extension I'd like to see is tracking of released updates, they disappear once they fall off of the tracking webpages. Jun 04 12:34:02 Let's try to have the spec finalised by Monday Jun 04 12:34:16 sbeattie: ah, yes, I think that can be done Jun 04 12:34:44 heno: did you still want to merge this in with fix-validation-tracking Jun 04 12:35:50 sbeattie: or if not we should rename that to hot-bug-watch or something to reduce confusion Jun 04 12:36:10 the tracking required is similar Jun 04 12:36:25 but the audience may be quite different Jun 04 12:37:00 the sru tacking will be used by more people Jun 04 12:37:13 tracking even Jun 04 12:38:33 IMO we should use the same technology for these two cases Jun 04 12:38:45 anyone else have strong views? Jun 04 12:38:55 <-- effie_jayx has quit (Nick collision from services.) Jun 04 12:39:30 --> hggdh (n=hggdh@pool-96-226-96-202.dllstx.fios.verizon.net) has joined #ubuntu-testing Jun 04 12:39:35 The main point of fix-validation-tracking is to track testing of fixes that have gone in after Beta, RC and final freezes Jun 04 12:39:48 to make sure they don't break things Jun 04 12:40:01 so very similar to the SRU case Jun 04 12:40:13 but needs a separate page of course Jun 04 12:40:34 sbeattie: but yeah, I think those could live on the same spec Jun 04 12:40:49 that's fine with me. Jun 04 12:41:48 * package-status-pages - UI Changes is 'TBD', but could perhaps just be removed Jun 04 12:42:12 stgraber: did you read my mail about this and the QA site? Jun 04 12:42:23 yep Jun 04 12:42:51 what do you think about implementing this in python outside of any CMS? Jun 04 12:43:22 we'll have to see how we want the output to be Jun 04 12:43:29 that would allow us to simply build on the work already done by Leann and Brian Jun 04 12:43:37 it may be some generic report, parsable page instead of html as they currently are Jun 04 12:43:55 or eventually replacing the output by DB queries in some cases but I'm fine with keeping the base scripts as standard python Jun 04 12:43:59 stgraber: see the two layout proposals on https://wiki.ubuntu.com/QATeam/Specs/PackageStatusPages :) Jun 04 12:44:30 we also do that with our current LP integration scripts, so that's similar to what we currently do :) Jun 04 12:44:40 we should actually make machine-friendly versions from the start too Jun 04 12:44:50 well, if that's only taking a .png from a place and displaying it, that'll be easy :) Jun 04 12:44:53 ok, cool Jun 04 12:45:10 yeah, I love machine-friendly (cdimage.u.c would also need that :)) Jun 04 12:45:48 so do we need to vote on the layout? :) or take further suggestions? Jun 04 12:46:15 no 2 Jun 04 12:47:10 2 is hot Jun 04 12:47:48 I don't think no 1 will lend itself to the rest of the site. Jun 04 12:47:53 +1 for no2 (of course :)) Jun 04 12:48:07 sorry, I need to leave for 15-20 minutes ... Jun 04 12:48:46 I'll update the spec Jun 04 12:49:02 so, should we do that as static pages via python, or use the drupal infrastructure? Jun 04 12:49:25 the latter would save us doing theme work Jun 04 12:49:29 heno: what ever is easiest to implement and maintain Jun 04 12:49:51 I think we want to be able to make changes quickly as we are developing it Jun 04 12:49:58 we could just pull in the main page part from pre-generated HTML (?) Jun 04 12:50:17 heno: couldn't you use the drupal framework and just insert the image create in python Jun 04 12:50:42 images tend to be a separate http call, which makes that sort of thing fairly easy Jun 04 12:51:25 ok, so the example has only images in the middle, but we would probably want more content there too Jun 04 12:52:14 stgraber: could you make a proof-of-concept page in drupal that pulls in static content to the main page and portlets? Jun 04 12:52:55 the super-plain html version could double as the machine-readable one perhaps Jun 04 12:54:42 --> fbs (n=felipe@124-78-246-201.adsl.terra.cl) has joined #ubuntu-testing Jun 04 12:54:47 re remaining specs are for the qa website. I'm happy to extend the timeline for finalising those Jun 04 12:55:01 since stgraber has exams ATM Jun 04 12:55:37 are there any items we should spec up that are not currently on that page? Jun 04 12:57:44 ok, let's move to the next topic Jun 04 12:57:55 TOPIC: Hardy SRU Verifications for 8.04.1 Jun 04 12:59:08 I guess the most difficult verifications are left towards the end Jun 04 12:59:35 sbeattie: do you have a feel for how we are positioned at this point? Jun 04 13:00:11 I think we're doing okay, if lagging a little on verifications. Jun 04 13:00:37 Some of the more difficult ones to verify involve bugfixes in the installation process, I believe. Jun 04 13:01:18 --> slangasek (n=vorlon@cl-177.ewr-01.us.sixxs.net) has joined #ubuntu-testing Jun 04 13:01:19 --> pitti (n=pitti@ubuntu/member/pitti) has joined #ubuntu-testing Jun 04 13:01:21 hi Jun 04 13:01:22 bweep Jun 04 13:01:35 I've invited pitti and slangasek to give their views also Jun 04 13:01:46 what's the question? Jun 04 13:01:49 --> zul (n=chuck@ubuntu/member/zul) has joined #ubuntu-testing Jun 04 13:02:08 we are talking about how we are doing in terms of getting verifications in place for hardy.1 Jun 04 13:02:39 pitti: I remember that we were often stuck with some stragglers in the past at the end Jun 04 13:02:48 often blocked on available hw Jun 04 13:03:03 I'd say that about 2/3 of SRUs get verified within 7 days Jun 04 13:03:11 we should try to get an overview of such potential blockers early this time Jun 04 13:03:16 but many of the 'less interesting' ones have bad or no feedback Jun 04 13:03:30 http://people.ubuntu.com/~ubuntu-archive/pending-sru.html mentions the #days a package resides in -proposed already Jun 04 13:03:48 some of them are installer related Jun 04 13:03:50 bad as in reported failures or not good enough quality of feedback? Jun 04 13:03:58 --> seb128 (n=seb128@ubuntu/member/seb128) has joined #ubuntu-testing Jun 04 13:04:07 solving that will require to wait until linux -19 is in, build d-i against it, and build CDs Jun 04 13:04:14 everything else can be tested on an installed system Jun 04 13:04:21 heno: no feedback Jun 04 13:04:37 heno: we have a few with "regression!" feedback, they are marked as v-failed Jun 04 13:04:49 right Jun 04 13:04:54 re Jun 04 13:04:57 or that the fix was incomplete. Jun 04 13:05:18 sbeattie: right, that happens sometimes Jun 04 13:05:31 but an incomplete fix is not a reason to reject the entire SRU if it fixes other things Jun 04 13:05:35 we just keep the bug open Jun 04 13:05:41 pedro_: suggested we appeal more widely for help with SRU testing Jun 04 13:05:51 a regression needs a followup upload, of course Jun 04 13:06:00 yeah, that would be very good Jun 04 13:06:06 stgraber: could you arrange QA blog access for sbeattie so he can use that? Jun 04 13:06:15 sure Jun 04 13:06:41 ok so so far so good, we just need more hamsters :) Jun 04 13:06:54 we'll revisit this next week as well Jun 04 13:07:08 if a bug even just says "I have used that new version for three days without a problem", that's already very useful Jun 04 13:07:26 <-- LaserJock has quit ("Leaving") Jun 04 13:07:40 Yeah, there's two aspects of increasing interest: Jun 04 13:07:52 1) validating specific fixes Jun 04 13:08:05 pitti: have you seen https://bugs.edge.launchpad.net/proposed-tracking/+bugs ? Jun 04 13:08:07 2) general usage to catch regressions. Jun 04 13:08:43 heno: no, I didn't; what is it? Jun 04 13:09:03 counting people using hardy-proposed Jun 04 13:09:19 (in a very improvised maner) Jun 04 13:09:21 <-- melat0nin (n=laurence@80-192-82-243.cable.ubr02.edin.blueyonder.co.uk) has left #ubuntu-testing Jun 04 13:09:40 oh, nice! Jun 04 13:10:29 bdmurray: could you set up a p-lp-b query to count the comments in those bugs? Jun 04 13:10:52 heno: sure Jun 04 13:11:13 sbeattie: right. perhaps you should blog about a few examples from #1 Jun 04 13:11:29 heno: sure, can do. Jun 04 13:11:41 I've already blogged about #2, but we can revisit that on mailing lists also Jun 04 13:12:03 ok, final topic: Weekly Meeting Scheduling Jun 04 13:12:27 sbeattie: done you can now access http://qa.ubuntu.com/node/add/story Jun 04 13:12:35 I wonder if we should bring the meeting time back to a fixed time each week? Jun 04 13:12:49 that would reduce confusion Jun 04 13:13:06 I would suggest 17.00 UTC Jun 04 13:13:13 ane make sure it's posted on the fridge :) Jun 04 13:13:23 indeed Jun 04 13:13:41 +1 for 17:00 UTC Jun 04 13:13:41 +1 Jun 04 13:13:51 +1 Jun 04 13:13:54 +1 Jun 04 13:14:06 ok for me +1 Jun 04 13:14:20 +1 as long as no other meeting conflicts. Jun 04 13:14:31 stgraber has a change of schedule now and liw is not required for these meetings anymore (and those were there original reasons for the split) Jun 04 13:14:53 that was easy :) Jun 04 13:14:58 yeah Jun 04 13:15:02 17.00 UTC it is Jun 04 13:15:15 * persia likes rotation. This has been an informative meeting, but I'd never make 17:00 unless there were exceptional circumstances Jun 04 13:15:15 (on the other hand, you all do all the work, so my vote oughtn't count for much) Jun 04 13:16:18 17.00 works for me. Jun 04 13:16:22 persia: what are the boundaries of sane hours for you? Jun 04 13:16:41 (and why are you up now? :) ) Jun 04 13:17:10 heno: My preferred hours of sleep are about 15:00 - 21:00, although I don't always reach them. Jun 04 13:17:41 (14:00 meeting overran, and interesting 16:00 meetings to watch rather than going straight to bed) Jun 04 13:17:58 ok, we have at least 3 people for whom those are core work hours :-/ Jun 04 13:18:03 Plus I wanted to make sure that Jordan's code got shared during the spec review Jun 04 13:18:47 heno: No worries. I'll only visit occasionally. I'm more interested in integration, and with well publicised meetings and minutes, I oughtn't miss much that can't be fed by proxy :) Jun 04 13:19:11 pedro_: can you update the wiki and chase the Fridge? Jun 04 13:19:41 heno: yep, np Jun 04 13:19:59 persia: ok cool. I'm sure we'll meet regularly now anyway Jun 04 13:20:09 any other topics? Jun 04 13:20:14 (briefly) Jun 04 13:20:40 thanks everyone! Jun 04 13:20:46 #endmeeting}}}