=== Starting logfile irclogs/launchpad-meeting.log === ubuntulog [i=ubuntulo@ubuntu/bot/ubuntulog] has joined #launchpad-meeting === ddaa [n=ddaa@nor75-18-82-241-238-155.fbx.proxad.net] has joined #launchpad-meeting === spiv [n=andrew@218-214-66-203.people.net.au] has joined #launchpad-meeting [11:57] Good. [11:57] spiv: welcome back [11:57] ddaa: thanks. [11:57] how was the vacation? [11:57] Fun. [11:57] Also, I have stitches in the back of my head. [11:58] You finally had this neural plug implanted, eh? [11:58] I proved it is possible to hit yourself in the back of your head with a snowboard that's still attached to your feet. [11:59] oh, yeah, it's quite feasible, those things are quite long [11:59] 159cm in my case :) === ddaa prefers alpine skiing [12:00] if only because I really cannot use a snowboard, and it's frustrating because it can do rather well with a pair of skis [12:00] lifeless: SteveA: jamesh: ping [12:00] MEETING STARTS [12:00] knee boards aren't too bad [12:00] but they work better on water than snow [12:01] This is the "things with obscure names" meeting, and you all know what we are here for. [12:01] hi [12:01] Next meeting 2006-09-18, 10:00 UTC. [12:01] mpool will be on leave. [12:01] * roll call [12:01] * production status [12:01] * advertising [12:01] * importd batch progress [12:01] * release finder [12:01] * Python import [12:01] * strategic plan [12:01] * bzr-lp features [12:01] * interesting bzr list threads [12:01] * 1.0 targets [12:01] * critical bugs [12:01] * pending sysadmin tasks [12:01] * any other business [12:01] If you wish to change the time of the meeting or add/remove agenda items, say "bzzzt!". [12:02] If we are short on time, the "any other business" item will be automatically, dropped. So if you ''want'' to discuss something more, speak up now. [12:02] hi === SteveA reads agenda [12:02] == Roll call == [12:02] mpool is on leave until September 19th. [12:02] please add "meeting in vilnius" [12:02] SteveA: ok [12:03] Everybody else has said hello already, so let's move on. [12:03] == Prodution status == [12:03] Nothing to report. [12:03] That was easy. [12:04] spiv: it looks like things do not tend to crash and burn while _you_ are away [12:04] Except me ;) [12:04] == Advertising == [12:04] Last meeting actions: [12:04] * spiv: blog about similarities between SVN and bzr checkouts, in relation to Launchpad. [12:04] * ddaa: when rolled out, to blog about branch UI improvements. [12:04] * unassigned: when rolled out, blog about --create-prefix not being needed anymore. [12:04] * ddaa: link to existing launchpad-bazaar blog entries from help.launchpad.net, and link from the front page. [12:04] I did the latter: Blog entries linked from https://help.launchpad.net/BazaarLinks. [12:05] I did an entry about the third [12:05] I'm not sure it's worth blogging about about branch UI improvement since jamesh already mentioned them. [12:05] spiv: still looking forward to reading your stuff. [12:06] this week is possibly not the week for that for spiv [12:06] release freeze is in 1 week [12:06] No hurry. [12:06] I can nag again next week. [12:06] oh, there was something from last week. ddaa, please add to agenda "things to do before you go on leave" [12:07] SteveA: ack [12:07] Yeah, nag me again next week please. [12:07] So I'll drop all actions except for spiv's one. [12:08] == Importd batch progress == [12:08] Last meeting actions: [12:08] * ddaa: sort out what to do with BatchProgress with lifeless and mpool [12:08] I started that discussion. Lifeless suggests leaving the code in Launchpad, I still do not know how to ensure the provided API stays in sync with what bzrlib actually use. [12:09] lifeless: can you help me on that this week? [12:09] jamesh: what's the URL of your blog entry? [12:09] ddaa: probably not. Next week sure. [12:09] SteveA: http://blogs.gnome.org/view/jamesh/2006/09/06/0 [12:09] SteveA: it's on https://help.launchpad.net/BazaarLinks [12:09] it was just a short one [12:09] ddaa: a test perhaps? [12:09] a test in launchpad that checks that the API works [12:09] SteveA: the problem is keeping those tests in sync with bzrlib, while it's not part of bzr [12:10] I don't see that as a problem [12:10] lifeless: okay, I can live with that delta in production for a while, if SteveA is happy with me doing that. [12:10] the tests will fail if they get out of sync, surely? [12:11] and so the test says "hello. API not in sync. do the work to make it in sync" [12:11] can we move on - no actions will be taken on that until next week regardless [12:11] SteveA: talk about that later, I think it more tricky than you appear to think. [12:11] == Release finder == [12:11] * jamesh: report on PRF progress. [12:12] My fixes went in last week. Last I checked, PRF was still running today and hadn't crashed [12:12] sweet [12:13] product release finder? === SteveA checks the "P" [12:13] 2 years after inception and it will be deployable [12:13] since it is an initial run, it has to do a fair bit of work and downloads [12:13] product release finder [12:13] thanks [12:13] ddaa: PLEASE stop abbreviating ithis [12:13] okay, then "release finder" [12:14] product release finder [12:14] we've had long threads about this [12:14] while it hasn't been run in production, it will make the staging nightly.sh quite slow since it will essentially be doing a first run every night [12:14] we still disagree [12:14] it's fine to abbreviate provided the first time you introduce it in a meeting or a text you say: ... the Product Release Finder (PRF) [12:14] then all is clear [12:14] ddaa: currently you have been outvoted, its not 'me vs you' - can you at least go with the majority and have the disagreement as a separate discussion [12:15] SteveA: okay [12:15] lifeless: the first person to abbreviate it in IRC was not me. [12:15] So if we are happy with the results, I think we should do at least one run in production [12:15] so that we'll be seeing the incremental runtime of the program [12:15] jamesh: if it passes successfully, lets get a report of the productname : version : tarballs it found [12:16] jamesh: in tabular form. that will be easy to sanity check [12:16] okay [12:16] in fact, productname : series : version : tarball [12:16] sounds good, though I think something that people will ask very soon after will be ability to delete releases [12:16] ddaa: and the answer will be "no" :) [12:17] I think that's a bad answere to give users. [12:17] they might ask if they can fix the release dates though [12:18] bah... we could have a discussion about that, either users will make my point soon enough, or it's not worth the time, so I'll shut ip. [12:18] jamesh: good news, thank you [12:18] == Python import == [12:18] https://launchpad.net/products/launchpad-bazaar/+bug/56360 [12:18] * ddaa: report on Python import progress [12:18] The resolver errors were fixed. The Python import is now blocked on what appears to be a bzr bug in Testament.as_text_lines when file names contain non-ascii chars. [12:19] More details: At the revision of the failure, the svn repo adds files like "/python/trunk/Mac/Contrib/PyIDE-src/Scripts/Hack/Remove .pyc files". In Testament.as_text_lines, the expression [line.encode('utf-8') for line in r] raises UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 74: ordinal not in range(128). [12:19] I think that's from ancient history [12:19] ddaa: is there a bug files for that ? [12:19] IOW, it's currently blocked on getting a more recent bzr in rocketfuel [12:19] I remember we saw something like that when we tried importing python with tla or baz [12:19] ddaa: I'm not sure that this bug is fixed in bzr.dev. [12:19] SteveA: not the same issue, please dont be distracted by it [12:20] ok [12:20] ddaa: can you please verify it is fixed by 0.10 ? [12:20] lifeless: I can run a smoke test to check. [12:20] from looking at the code, looks like "line" is accidentally ascii, then gets implicitly decoded to unicode in order to run .encode() on it [12:20] ACTION: ddaa to try reproducing this bug with bzr 0.10 [12:20] and bzr.dev too [12:20] ddaa: thank you [12:21] moving on [12:21] == strategic plan == [12:21] * SteveA, ddaa, lifeless, jamesh: asked to review 32/Bazaar. [12:22] I think I did (though I do not remember for sure) [12:22] I haven't sent him comments yet [12:22] spiv: I think the same applies to you. [12:22] ddaa: I already reviewed it as it was written [12:22] ok [12:23] I read it, but I don't have any comments to make. [12:23] SteveA: ? [12:24] not done yet [12:24] fun https://features.launchpad.net/products/bzr/+spec/bundle-piping [12:24] okay, no big rush, but jamesh and SteveA please make mpool happy before he comes back. [12:24] something like 9 days from now [12:25] == bzr-lp features == [12:25] * mpool: report on progress for bzr-lp features [12:25] mpool is on leave [12:25] does somebody want to speak for him? [12:25] nope [12:25] hurh hurh hurh [12:25] is that the lp:// URL stuff, or something else? [12:25] jamesh: both [12:25] theres a collection of stuff [12:25] which is all moving along piecemeal [12:26] as is this meeting [12:26] == Interesting bzr list threads == [12:26] After discussion with SteveA about the difficulty of keeping up with the bzr mailing list to stay updated about interesting developments in the bzr community, I would like to ask the fine folks here to point me to "interesting" threads that happened in the past week. [12:26] My definition of interesting here is: [12:26] * Design discussions [12:26] * Discussion of new/future bzr features [12:26] * Release management decisions [12:26] * Discussions providing interesting insights into the needs of users [12:26] * Anything related to Launchpad [12:26] That explicitly excludes "noise" like: [12:26] * Simple code review [12:26] * Discussion of internal implementation choices [12:26] * Trivial user support [12:26] uhm [12:26] I've been pushing the working tree revision properties stuff through, which is a prerequisite for the AutomaticBugBranchLinks spec [12:27] wow, I have no idea how to generate that list trivially [12:27] jamesh: seen that, looks like good idea IMO [12:27] RM I've nothing to report on [12:27] nothing major has changed in the design philosophy [12:27] It's a bit of an experiment, but maybe guys here could take some notes with pertinent keywords (e.g. words in the subject of the emails) [12:27] new features have been creeping into 0.11 at quite a reasonable pace [12:28] Anyway, since it's the first time you read of that, let's just see if that gives something useful for next week's meeting. [12:29] == Things to do before you go on leave == [12:29] SteveA: the stage is yours [12:29] thanks [12:29] so, iirc, there were some loose ends left around that we found out about in this meeting after spiv went on vacation [12:30] can anyone remember the details? [12:30] yes [12:30] push all branches [12:30] something about spiv branches not being pushed [12:30] send an email saying where its at [12:30] I'm going to put up a standard policy for "things to do before you go on vacation" -- a checklist [12:30] for the launchpad team [12:30] and I want to make sure it will cover the situation we had a short while ago [12:30] ok [12:30] so [12:31] - push all branches [12:31] A checklist would be really handy. [12:31] - send an email about ongoing / outstanding issues [12:31] including details of branches in progress [12:31] right, the situation was : Martin and I wanted to move the smart server forward, but could not because about 2 days work was not pushed [12:31] the idea being that others can continue where you left off === ddaa has been pondering using bound branches to avoid forgetting to push [12:32] in this particular case, there is a bunch of uncommitted reorganisation going on which is fine - but it would have been good to have an rsync'd copy of that available [12:32] spiv: in this case, did it occur to you to do this kind of thing as you were preparing to leave on vacation? I'm wondering how best to remind people [12:32] even if we have a checklist, it needs to be used [12:34] I find before a holiday I tend to go "oh oops, I need to make sure I have StaffCalendar updated, make sure I have nothing still assigned in PendingReviews, oh I better make sure I have a holiday notice on my review queue, while I'm on the wiki fill in apologies for meetings...." [12:34] ah, so more things for the checklist [12:34] particularly for reviewers [12:35] SteveA: if the checklist includes some sort of report, it should be pretty easy to check for people who just completely forget. [12:35] With the pushing in this case, I usually push work regularly, but as lifeless points out there was a lot of uncommitted state due to the nature of the work at that point. [12:35] reviewers need to put their leave in the reviews page [12:35] having things assigned in the reviews page is ok [12:35] ok [12:35] I will reassign if you put a 'I'm on leave' marker there - but an email to me would help [12:35] thanks for talking through this [12:36] I'll formulate a launchpad policy for this [12:36] ddaa: thanks, done [12:36] ACTION: SteveA to increase the bureaucracy of the Lauchpad team [12:36] ;) [12:36] == Meeting in Vilnius == [12:36] SteveA: ? [12:36] tim penhey is joining us soon [12:36] cool [12:36] he'll report to mpool, work from NZ [12:37] That's cool! [12:37] ? [12:37] and he may stop over in sydney on his way to dunedin [12:37] Report to mpool? [12:37] or however you spell dunedin [12:37] he was a couple o years ahead of me at uni, small world syndrome [12:37] mpool will be taking a greater interest in how launchpad is used with bzr, once 1.0 is out [12:37] and, considering timezones too, this is a natural thing [12:38] SteveA: does that mean I'll be reporting to mpool too? [12:38] no [12:38] you'll stay reporting to me [12:38] and then you and tim will work together a lot [12:38] we'll see how it goes [12:38] Weird to have a pair with two different line managers. [12:38] not so weird [12:38] ddaa: I think locality of time zone is quite important [12:38] I sometimes coordinate with mdz a lot [12:38] but also, mpool and I do talk a lot === ddaa remembers the section about Hybrid Organizations in the book he read recently [12:39] one of things that we had difficulty with was effective communication [12:39] and we talk with each other a lot too :-) [12:39] due to us being effectivelyt 12 hours out [12:39] Makes sense, after all. [12:39] So, what about Vilnius? [12:40] so, tim and I have talked about him coming to vilnius to get an introduction to bzr in launchpad [12:40] first week of october is one set of dates possible [12:40] ddaa: woiuld you be able to come then too? [12:40] I do not have anything specific planned. [12:40] another possibility is 2nd week of october [12:40] So, provisionally, yes. [12:40] and that might work out better for me, as host [12:40] I'll know after wednesday [12:41] ddaa: is either week okay with you? [12:41] mbp also posted about https://wiki.canonical.com/BazaarSprintSydney2006 last week [12:41] as far as I recall, yes. I'll start preparing Ewa to it. [12:41] is any coordination with that needed? [12:41] should not be [12:42] separate topics, codebases, agendas [12:42] lifeless++ [12:42] jamesh: have you considered going to the sydney sprint? [12:43] SteveA: if it would be useful for me to go, then sure. [12:43] would you like to come ? [12:44] we want to be doing bzr code primarily during that time [12:44] smart server, dirstate, performance [12:44] I see various lp things on the agenda [12:44] if the lp things will be in one half of the week, james could go for that [12:44] its 2 weeks long [12:45] s/week/weeks/ [12:45] the LP integration stuff looks interesting, and I've been doing a bit of work on that. It would be interesting to see what the bzr folks think of our plans [12:45] anyway, doesn't need to be deicided in this meeting [12:45] there is LP stuff there [12:45] jamesh: one week on the lp/bzr stuff in that meeting is fine with me [12:46] sort it out with lifeless and mpool [12:46] because bzr/lp is the strategic overview [12:46] but I fully expect most of the time to be the detail of getting bzr to win vs hg [12:47] Launchpad as a compliment to Bazaar could definitely help there ... [12:47] yup [12:47] ddaa: anything else on the agenda? [12:47] that has been the plan for like... 11 months now? [12:47] == 1.0 targets == [12:47] [skipping] [12:48] == Critical bugs == [12:48] Good news, worth showing off. [12:48] https://launchpad.net/bugs/31308 Cannot set branch associated to a product series. [12:48] * Last meeting jamesh agreed to work on that. Jamesh, progress? [12:48] its in staging [12:48] https://launchpad.net/bugs/37897 renaming project, product or series breaks vcs imports. Fix released. [12:48] * ddaa: report on deployment [12:48] https://launchpad.net/bugs/51130 cannot use +admin on a branch I own. Fix released. [12:48] I tried it out for some products I own, and it works nicely [12:48] importd-publish-source was deployed, squashing https://launchpad.net/bugs/37897. A few imports did not have a cscvs working tree and failed the transition: [12:48] * silo (now on git), console-data (svn branch renamed), xrestop (now on git), libmusicbrainz and libtunepimp (now on svn). All these are obsolete and unfixable and are now STOPPED. [12:48] * exe (https://launchpad.net/products/exe/trunk) also failed: their svn repo appears offline, the download instructions have not been updated, and I had no reply when asking on IRC (#exe on freenode). Marked the import STOPPED by lack of information needed to fix it. [12:49] theres a report on lp-users of a CVS repo that needs the host changed in lp [12:49] In other words, all our criticals will be fixed by the next rollou. [12:49] lifeless: thanks, will do today. [12:50] MEETING CLOSED [12:50] Sorry for the abrupt end, be we are overdue already. [12:51] ddaa: I'd like a voice call with you today. [12:51] sorry for forgetting to report last week [12:51] I was a bad boy. [12:51] What time would you like to voice? [12:53] A bit later in the afternoon would be nice, say between 1400 and 1600 UTC? [12:53] 1600 is way too late [12:53] 1400 is possible. earlier is better for me [12:54] I'm concerned about coordinating lunch with Ewa, she's not well at all at the moment. [12:54] And out of the house right now. [12:55] SteveA: we can have voice in 30 mins if you want. [12:55] Just time for me to have a little break. [12:56] that's a very good time for me. thanks [12:56] 30 mins past the next hour [12:56] That's 1130 UTC [12:56] lifeless: got time for a quick phone call? === ddaa goes to the place [12:57] SteveA: after the review meeting [12:57] which is in 2 mins? [12:57] yes [12:57] ok, thanks === jamesh_ [n=james@203-59-208-48.dyn.iinet.net.au] has joined #launchpad-meeting