20070307
Revision 3 as of 2008-08-06 16:31:36
Clear message
Log
TZ UTC+6
06:00 mrevell Welcome to the first Launchpda users' meeting! 06:01 statik hurrah, users! 06:01 mrevell Here's the agenda: 06:01 mrevell * 06:01 mrevell Welcome 06:01 mrevell * 06:01 mrevell Agenda 06:01 mrevell * 06:01 mrevell Introduction to the Launchpad developers are who present 06:01 mrevell * 06:01 mrevell Invitation to beta team 06:01 mrevell * 06:02 mrevell Annoyance of the week 06:02 mrevell * 06:02 mrevell User questions 06:02 mrevell * 06:02 mrevell Next meeting 06:02 mrevell * 06:02 mrevell Welcome 06:02 mrevell * 06:02 mrevell Agenda 06:02 mrevell * 06:02 mrevell Introduction to the Launchpad developers are who present 06:02 mrevell * 06:02 mrevell Invitation to beta team 06:02 mrevell * 06:02 mrevell Annoyance of the week 06:02 mrevell * 06:02 mrevell User questions 06:02 mrevell * 06:02 mrevell Next meeting 06:02 mrevell This is a chance for Launchpad's users to ask questions of and make suggestions to the Launchpad team. 06:02 mrevell So, if you're a Launchpad developer, please say "hi" to let us know you're here 06:03 flacoste hi 06:03 ddaa hi 06:03 bac hello 06:03 thumper hi 06:03 mrevell Thanks guys. 06:03 salgado hello 06:03 mrevell Iniviation to beta team 06:04 ddaa Aren't we supposed to be introduced to? 06:04 mrevell The Launchpad beta is well under way and is open to anyone who wants to have a role in shaping the future of Launchpad. 06:04 theCore ok, gotta go, see ya all 06:04 mrevell If you'd like to join the Launchpad beta, you can sign up at: 06:05 mrevell https://launchpad.net/~launchpad-beta-testers/+members 06:05 mrevell ddaa: Not sure what you mean 06:05 statik hi 06:05 ddaa The agenda reads "Introduction to the Launchpad developers are who present" 06:05 ddaa ignore me if I'm just causing trouble 06:06 mrevell ddaa: :) You said hello. 06:06 mrevell ddaa: If that's not enough let's discuss it later :) 06:06 ddaa ok 06:06 mrevell Moving onto "Annoyance of the week" 06:06 mrevell Each week, I'd like to gather input on what has frustrated you recently about Launchpad. 06:07 mrevell I'll then take it to the Launchpad developers meeting, which happens each Thursday. 06:07 heno the beta is veeeery slow ... 06:07 mrevell So, if you'd like to report something that has annoyed or frustrated you about Launchpad, please tell us now. 06:07 heno (my annoyance of the week) 06:07 mrevell heno: Thanks. 06:08 mthaddon heno: slower than the non-beta? 06:08 heno mthaddon: yes, very 06:08 heno it can take 10 seconds for a random bug page to load 06:08 mthaddon I haven't played much with the non-beta recently, so can't comment, but that's interesting... 06:08 heno non-beta might be 1-2 secs 06:08 mrevell I haven't noticed beta be that slow. === mwh_ is now known as mwh 06:09 Kuhrscher the current status of the upstream translations import for Feisty annoyed me 06:09 heno I'm actually surprised that the LP devs are not feeling this pain too 06:09 heno it's a common topic in the distro meetings 06:09 ddaa it actually seems to me that it got less slow recently... but maybe I just got used to it. 06:10 heno (see the logs from last week :) ) 06:10 mrevell Are other people here noticing extreme slowness from beta? 06:10 salgado today we seem to be experiencing some problems with our database server, which means the slowness won't be limited to beta 06:10 heno I'll check with SteveA tomorrow at the meeting how his profiling is going 06:10 mrevell Kuhrscher: Thanks, we've got a question about that coming up later. 06:10 mrevell heno: Thanks. 06:11 thumper I do, but I just thought it may have been that I'm in NZ :) 06:11 mrevell Okay, any other points of annoyance that you'd like to raise? 06:11 heno if we set up some instructions for how to do those, I'm sure others would help 06:11 salgado (and I have to admit we're having quite a few timeouts today, probably because of that issue) 06:11 mrevell heno: Set up profiling? 06:11 heno the ooopses today are a separate issue I think === jenda waves 06:12 heno mrevell: instructions on how to use firebug sensibly and collect logs, or whatever it does 06:12 jenda guys - I know it's slightly OT, but any idea what's up with the CC elections? 06:12 salgado heno, yes, it is. that's what I said === jenda doesn't know where to ask 06:12 mrevell heno: Right, thanks, noted. 06:12 mrevell jenda: We're having a Launchpad users' meeting atm, you're welcome to join in. 06:13 jenda ah 06:13 jenda sorry to barge in like that ;) 06:13 mrevell Okay, if there aren't any other annoyances that people would like to raise, we'll move on :) 06:13 mrevell jenda: No worries :) 06:13 mrevell User questions 06:14 mrevell Throughout the week, Launchpad users have added their questions to the meeting agenda. 06:14 mrevell Here's the first. 06:14 mrevell It will be GREAT to have upstream GNOME distribution translation ability!Will we benefit this in a near future? Vladimer Sichinava GNOME/Ubuntu translator. 06:14 mrevell danilos: Is that something you can answer? 06:15 mrevell And is Vladimer here today? 06:15 mrevell Okay, I'll seek an answer and give it in the next user meeting. 06:15 danilos mrevell: I'd like to provide that as well since I am also a GTP spokesperson and Serbian GNOME team coordinator, but it depends on other priorities (it's definitely something we want to have, but we can't set a date on it) 06:16 mrevell danilos: Ah, thanks. 06:16 mrevell danilos: It appears Vladimer isn't in the meeting, so hopefully he'll see the post-meeting report. 06:16 mrevell Moving onto the next question. 06:16 mrevell I'm not happy about the fact that there has never been a fixed date for Ubuntu translation opening. Although I respect Rosetta developers, I feel bad about us having less than a month to prepare the Feisty translations. We (the Czech team) have already found a number of regressions from Edgy and it will be difficult to fix all those bugs and make Ubuntu better translated in time. I often have a feeling that although Ubu 06:16 mrevell ntu prides itself with our translations, we don't get much credit and respect within the Launchpad paid developers/Canonical. Can this issue be better resolved in the future? -- MartinBhm 06:17 mrevell mhb: That's your question, I believe. === mhb is Martin Bhm 06:17 mhb yes, it is 06:18 mrevell mhb: The Launchpad team and Canonical do greatly value the Ubuntu translators. I believe that the process for opening translations for Feisty+1 will be smoother. === carlos is also a launchpad developer and is around (sorry, I was distracted...) 06:19 heno Perhaps we should do an informal impact study on this kind of thing: what is more disruptive having Launchpad down for a day or delaying translation import until it's almost too late? This issue will likely be fixes properly at some point, but similar dilemmas could come up later 06:19 mhb mrevell: I think there was the same belief for feisty as the edgy opening was also relatively late 06:19 mrevell mhb: You say don't feel the Launchpad team gives translators much credit or respect. How could we improve that impression? 06:19 carlos mhb: the idea is to have a fixed date for open it, but Edgy and Feisty had some technical problems involved 06:19 carlos that delayed the process much more than expected 06:19 Kuhrscher I hope it would also be possible to lock the translation until all imports are done. 06:20 carlos Kuhrscher: there is already a bug to do that, yes 06:20 Kuhrscher carlos: nice to hear 06:20 mhb mrevell: we can't make any changes to the timetable, so creating one for the opening &really make it happen at the fixed date would be great === sinzui is now known as sinzui_lunch 06:20 mhb mrevell: we can't say "there has been so little time, give us more" 06:21 mhb mrevell: and I think other teams should not have that privilege either 06:21 mrevell mhb: I'm sorry that the Feisty translations opening date has been held up, giving you little time. 06:22 Kuhrscher Could we introduce the next release some time to check the imports for bugs _before_ the translation is opened? 06:22 carlos Kuhrscher: it was not a 'bug' problem 06:22 carlos but a performance problem 06:23 mrevell mhb: We do want to keep to the dates for opening translations and I'm sure that, having found out more about what can go wrong, Feisty+1 will be much improved. 06:23 Kuhrscher carlos: The last two releases there had been a lot of failed imports... 06:23 carlos mhb: Well, that's a problem, yes. But we keep releasing translations updates even after final release. It's not the solution to the problem, of course, but It mitigates a bit the problem of being late or the lack of time before final release 06:23 Kuhrscher or wrong templates... 06:23 mhb carlos: AFAIK, the updates were delayed a lot, too 06:24 mrevell mhb: I'm not sure what you mean about teams having priveleges to ask for extra time. 06:24 danilos Kuhrscher: it might also be due to incompatible change in GNU gettext msgfmt (a lot of failed imports I see is due to this) 06:24 mhb mrevell: I meant that the Rosetta folks can say that, because they don't have a fixed date for the opening 06:24 carlos Kuhrscher: most of the time is a bug in the package not in the import process itself 06:24 danilos Kuhrscher: (msgfmt was changed to require '%d' in all plural form messages with 0.15, and earlier versions allowed eg. singular form not to have it) 06:25 mhb mrevell: can you or someone else make sure that the "Rosetta opening" date will be set in the Feisty+1 release schedule? 06:25 mhb mrevell: and that the Rosetta folks start testing early enough so that they can open the translation up at that date? 06:25 Kuhrscher Sure, but it would be best to start the translation process only if all those problems are identified and fixed. 06:26 carlos mhb: the updates should be more or less done monthly, that's why we have daily snapshots. I think there are some delays from time to time, but I know Martin tries to do one update each month first Monday of each month 06:27 mhb carlos: as far as I remember there were close to zero Edgy translations updates before January (that's what some of the translator said, I am not sure about this) 06:28 mhb carlos: having a fixed date early enough and really open translations at that date matters more for me, though 06:29 carlos mhb: I don't think it was so late.. let me check... 06:30 mrevell mhb: As I understand, the improvements we've made to Launchpad will mean that we can open Feisty+1 translations as soon as Feisty+1 has created in Launchpad. That should mean we can give a date for translations opening and that it will give you much more time. 06:30 carlos mhb: For the Spanish one, there were updates almost every month since release... 06:30 carlos https://beta.launchpad.net/ubuntu/+source/language-pack-es 06:30 carlos https://launchpad.net/ubuntu/+source/language-pack-es 06:31 mhb https://beta.launchpad.net/ubuntu/+source/language-pack-cs/ 06:31 carlos mrevell: well, we still need to implement some changes to be able to do that without turning launchpad off 06:31 Kuhrscher Please start this translating period after the UpstreamVersionFreeze. 06:31 carlos mrevell: but the idea is that one 06:32 mrevell carlos: Thanks for the clarification. 06:32 Kuhrscher Just to avoid interference with the upstream translators... 06:32 mhb carlos: can you do it fast enough then? 06:32 mhb carlos: and really set the date in the Feisty+1ReleaseSchedule and open it then? 06:32 mhb with more than a month for us? 06:33 ddaa Shutting down launchpad for a day is not really an option, since we want to attract upstreams that may not be interested in translations 06:33 carlos mhb: not really fast enough, but without turning launchpad down 06:33 ddaa but may be interested in bzr, or the bug tracking system 06:33 carlos which is the main problem until now, how to schedule that 06:33 carlos to avoid problems like what ddaa is pointing 06:33 mhb carlos: that's not much of a assurance for us 06:34 mhb carlos: what do you mean by fast enough? 06:34 mhb fixed dates is what would help a lot 06:34 carlos mhb: what's your ideal date to do the opening? 06:34 heno ddaa: right, but not delivering translation imports several times in a row will also leave a bad track record, which will also reach the ears of those upstreams 06:34 danilos mhb: with 'not fast enough' carlos was thinking that the process will take a while (like a week when you'll be unable to do any translations), but we won't have to shut launchpad down completely 06:35 heno It will stay on mailing list and IRC logs and can be googled forever :) 06:35 carlos anyway, as Kuhrscher points, I don't think we should open translations 6 months before release... 06:35 carlos so it should not be a problem 06:35 heno one day of downtime may not be a bad alternative 06:35 carlos heno: believe me, it is 06:35 mhb carlos: perhaps after the DebianImportFreeze 06:36 danilos also, you guys should consider one more problem with opening translations too early: you'll have to repeat work you do in feisty and feisty+1 unless we do the copying later on as well (we've designed it so we can do it, but I don't think we've done it so far) 06:36 ddaa but presumably, disabling just translations for one week, would be less of a problem 06:36 ddaa the issue is separation of concerns, people who do not care about the translation should not see their version control system server go offline because of them. 06:36 carlos mhb: UBuntu's wiki page is being slow here 06:37 carlos mhb: so I cannot check the schedule... 06:37 danilos mhb: I am at least not that well into Ubuntu schedule and timeline, so it'd help if you gave exact dates, relative dates inside 6-month cycle or point me to a page where I can read on them :) 06:37 carlos for the case of Feisty, when is that? 06:37 mhb December 21st 06:38 carlos that's 4 months before release, isn't it? 06:38 mhb carlos: well, yes 06:38 mhb carlos: the second important date for Ubuntu release cycle is 2 months after that 06:38 carlos it would interfere with GNOME upstream translations, but that's doable 06:39 carlos that gives us two months after opening a distro release so the process even being slow will be done at that time 06:40 mhb carlos: it should happen some time before the FeatureFreeze, because that's relatively late in the process 06:40 carlos as we will be 'hidding' the imports until it's done, we could start the opening at any time we want once Launchpad knows about that new distro release 06:40 carlos mhb: I'm talking about DebianImportFreeze 06:41 carlos the fact that we delay it until FeatureFreeze or not it up to you (the translators) 06:41 carlos s/it up/is up/ 06:41 mhb carlos: I'm for opening it up as soon as possible 06:41 carlos is that compromise enough? 06:42 danilos carlos: do we plan to copy distrorelease to distrorelease+1 contributions even after distrorelease+1 is opened? if we don't, translators will have to do a lot of duplicate work if we open it too soon 06:42 mhb carlos: sure, 3-4 months is enough time, I believe ... but please write that date in the Feisty+1ReleaseSchedule so we can count on it 06:42 danilos mhb: ^^^ (iow, we want to schedule such copying as well) 06:42 carlos danilos: if we are able to do the opening without turning launchpad offline, that's doable too 06:43 danilos carlos: I know it's doable, but if we've got a date for one thing, we need a date for another as well 06:44 carlos danilos: well, I would test how much time it would take and maybe, schedule it to be done once per month or something like that 06:44 carlos but nothing we could schedule right now 06:44 carlos the opening is something we could put in the schedule as a way to get the compromise of getting it done then 06:45 mhb I guess that's answered, then. 06:45 mrevell mhb: Thanks for your question. 06:45 mrevell mhb: If you'd like to discuss it further, don't forget we have the launchpad-users mailing list. 06:45 mhb and thank you for the answers 06:45 mrevell Next questions also relates to Rosetta and is from mhb. We're running out of time, though, so please lets make the answers as quick as possible. 06:46 mrevell One of the other things that make me believe Ubuntu translators and translations are being rather neglected is the fact that Launchpad/Rosetta has failed to implement a simple search tool for the translations. Without it Rosetta seems to be rather clumsy for fixing bugs and typos. As far as I know, the Launchpad/Rosetta developers are aware of this but almost since day 0 but still they haven't managed to implement it. T 06:46 mrevell hey often claim it is very hard to implement but I think the lack of pressure from the community is another factor. I hope we can raise our voices louder so that even the paid developers know we really need it sooner than for a feisty+2 or even later. A fixed date for it would be nice, as it seems there are little fixed dates in the development process. -- MartinBhm 06:46 mrevell So, the question is: When can we have searchable translations in Launchpad? 06:46 carlos danilos: that one is for you ;-) 06:46 kiko mrevell, you can search using google, perhaps. 06:47 kiko we now allow google to index our translation pages 06:47 carlos mhb: just as a brief answer, google has been indexing us since a January 06:47 carlos so you should be able to use that as a workaround (that's why we did that change) 06:47 mrevell mhb: Do you want to try with Google and report to launchpad-users how you get on? 06:47 danilos another point is that we've got a huge database (which is what makes text search hard), and we've been working on improving it, so we should have something in the near future (though, more DB changes will be required) 06:48 mrevell carlos, kiko, danilos: Thanks for the suggestions and for giving an insight into why we don't yet have searchable translations. 06:48 mrevell The next question is also related to searching, but searching of bugs: 06:48 mrevell Are there any plans for improving search facilities in Launchpad? It always strikes me how difficult it is to find something using only keyword search compared to Bugzilla's "Advanced Query" panel. -- DavorCubranic 06:49 mrevell Davor, are you here? === carlos needs to leave now. Thanks for your questions and for coming 06:49 mrevell carlos: Thank you for your input this evening! 06:49 kiko mrevell, well, we do offer advanced searching too. 06:49 flacoste this is not a definitive answer, but we are upgrading to PosgreSQL 8.2 which will allow us to improve greatly our text searches 06:50 flacoste or so says our DBA 06:50 heno We should perhaps look at integrating some of the search features developed in bughelper directly in LP 06:50 Kuhrscher He is absoultely right. I it is very difficult to find bugs even if you know it is existing... 06:50 mrevell flacoste, kiko: Thanks. It looks as though Davor isn't here, so hopefully he'll see your replies in the meeting report. === danilos has to leave as well 06:51 danilos thanks for the questions, and don't forget that you can always email us or catch us here in #launchpad as well 06:51 mrevell Okay, moving onto the next question. 06:52 mrevell "the bug numbers are atomic, so it would be convenient that the search box (or another text box) acts like a quick "go to" when you enter a bug number 06:52 mrevell tabs should be click-able everywhere on it, not just on the text 06:52 mrevell disabled tabs should disappears or have another color" 06:52 mrevell That question is from TheCore, who had to leave early. BjornT is that something you could answer? 06:53 ddaa tabs clickable everywhere has a bug currently assigned to Usman 06:53 mrevell BjornT: As far as I know, the search box does allow you to enter a bug number. 06:53 mrevell ddaa: thanks 06:53 ddaa jump-to-bug-number currently works, and there's a bug open about changing it because it prevents searching for numbers 06:54 mrevell ddaa: Thank you. 06:54 ddaa But it does not appear to work in the global search 06:54 mrevell We had a number of other questions as well, relating to Rosetta. I'll post those to the launchpad-users list, along with the other questions from tonight's meeting. 06:54 mrevell Do any users present have questions that were not raised in the agenda? 06:55 ddaa yeah, it only works in the bug search box 06:55 alex_muntada mrevell: when will next meeting take place? 06:55 mrevell alex_muntada: Aha :) That's the next item on the agenda. 06:55 alex_muntada :) 06:55 mthaddon Is there any date in mind (fixed or simply targeted) for the release of UI 1.0? 06:55 Kuhrscher We don't talk about Rosetta? 06:56 mrevell Kuhrscher: Quite a large part of the meeting has been about Rosetta-related issues. Unfortunately, both main Rosetta developers have had to leave now. I'll raise the remaining questions on the launchpad-users list. 06:57 Kuhrscher That's really annoying... 06:58 mrevell mthaddon: We want Launchpad 1.0 to be perfect before we make it the default choice for Launchpad users. However, you can use the beta right now. 06:58 ddaa would be nice to have it happen sooner than later 06:59 thumper well, it'll never be perfect 06:59 ddaa dealing with the divergence between production and rocketfuel will become increasingly painful 06:59 thumper we just want it to look good and work 06:59 alex_muntada Any news about LP UI translation? 06:59 mthaddon mrevell, I agree, but am wondering if there's a targeted date for that or just "when it's ready"? 06:59 kiko okay, off the phone! 06:59 mrevell Kuhrscher: I'm sorry that we haven't got to your questions. It's the first time we've run the meeting, so please blame me for my lack of experience. 07:00 flacoste alex_muntada: the localization of the Answer Tracker is targeted for after the 1.0 release 07:00 mrevell mthaddon: Much like a pint of Guinness, when it's ready :) 07:00 kiko about a date for 1.0; right now the issue is polish and warts. there are a lot of warts and quite a bit of polish but there is no heavy lifting to be done 07:00 alex_muntada flacoste: too bad :( 07:00 Kuhrscher Sorry, but my impression is that upstream i18n issues are always treated like that :( 07:00 mthaddon mrevell: gotcha - well much like a pint of Guinness it's looking pretty good right now and I'm sure it'll taste great when it's done 07:00 kiko so it's more a matter of weeks than months. 07:01 mrevell mthaddon: :) 07:01 Kuhrscher And that is not only my impression... 07:01 kiko that's a great way of not giving a deadline isn't it? 07:01 kiko Kuhrscher, we give upstream i18n a lot of priority, but the truth is that rosetta is not a trivial application, and that work there takes time. 07:02 kiko we've had a lot of serious performance issues over the past six months, and fixing those issues takes effort; not bean-counting effort, but debugging, refactoring, redesign, etc. 07:02 kiko so I make no apologies for our lack of progress on that front, though I do promise that we are working hard on it and that there is a horizon of improvement. 07:02 ddaa in other words, it's a freaking three headed cerberus, and one developer has to hold it down while the other developers tries to beats it down silly while jumping around to dodge the bites 07:02 Kuhrscher kiko: To be honestly I don't see any improvements regarding the most annoying issues for us. 07:02 heno kiko: that has a Pentagon ring to it ;p 07:03 kiko Kuhrscher, I don't think that's a very fair comment. 07:03 heno 'at a time of our choosing' 07:04 kiko we've in the past three months: improved general performance significantly, improved the way the translation for is structured for reviewers, allowed public indexing and browsing of translations, improved the way both imports and exports were handled, and along the way, fixed dozens of bugs that affected "you" === flacoste is starving to death 07:04 kiko we've also done major work towards implementing native support for firefox and openoffice translations 07:04 kiko planned the next generation database schema === flacoste is now known as flacoste_lunch 07:05 kiko ... the list goes on. so I'm sure that you've seen some improvements; perhaps not the ones that annoy you the most, but we have many users, and it's hard to keep everybody happy. 07:05 mrevell kiko: Thanks very much. 07:05 Ubugtu New bug: #90429 in malone "Badly worded 'release manager' text when nominating a bug for a release" [Undecided,Unconfirmed] https://launchpad.net/bugs/90429 07:06 kiko mrevell, about the bug searches; I'm not sure if the issue is that the advanced search is invisible, or if it's because the search isn't useful enough. it might be both! 07:06 kiko when davor replies via launchpad-users we'll know. :) 07:06 mrevell Kuhrscher: Again, I'm sorry that we didn't fully cover your questions. I'll get answers to them for you. 07:06 Kuhrscher Kiko: Sorry, if I annoyed you. 07:06 mrevell Thank you everyone for attending, for your questions and for your time. 07:06 mrevell I propose the next meeting at 09:00 on Wednesday 14 March 07. Any objections? 07:07 Kuhrscher Kiko: The only import point for me is, to have all upstremas translations in Rosetta and in the distribution. 07:07 kiko Kuhrscher, sure. I'll relay that to carlos and danilos who have spent days and nights working to make it a better service. :) 07:07 thumper mrevell: 9:00UTC? 07:07 bac mrevell: that's pretty tough for the Americas. 07:07 mrevell bac: Yeah, but 17:00 UTC is bad for Australia. 07:07 thumper bac, but ok for ozzies 07:07 kiko 9:00UTC? wow! 07:08 mrevell thumper: Yeah 07:08 Kuhrscher Kiko: And the last two releases there had been so many problems. If it is better now, no one will be happier than me :) 07:08 bac mrevell: if it's alternating that's fine 07:08 mrevell kiko: I can change it if it doesn't work, but I want to make sure we cater for .au and .nz 07:08 mrevell bac: Yeah, I'll alternate. 07:08 kiko Kuhrscher, I kind of want to slit my wrists at the performance issues to be honest. they are proving to be much harder to fix than we expected 07:09 kiko we've made changes, waited to see them go live, and been dismayed at how bad things went 07:09 kiko right now danilos has just landed a change which might improve things again 07:09 mrevell Okay, well, thanks again everyone. Obviously you're free to continue your conversations, but I'll draw the official meeting to a close for this week.