IRC Meeting on 2012-10-06




(21.04.46) gforcada: hi all, should we start?

(21.09.51) Tranzistors: just got here

(21.12.41) Tranzistors: gforcada, I guess we should

(21.14.10) gforcada: ok, so first point is the GNOME 3.6 release, any word about it?

(21.14.43) PiotrDrag1: gforcada: best release ever?

(21.15.02) gforcada: :)

(21.15.40) Tranzistors: agree

(21.15.57) gforcada: and string freeze breaks? Compared to 3.0 was like non-existing :)

(21.16.25) PiotrDrag1: not many breaks, but I've found some serious bugs quite late in the cycle

(21.16.50) gforcada: PiotrDrag1: translation related you mean?

(21.16.56) Tranzistors: the weird thing happened with glade, after translation it turned out that the work had to be done on different branch

(21.17.16) PiotrDrag1: gforcada: yes, like g-c-c showing English dialogs instead oflocalized ones

(21.17.19) PiotrDrag1: localized*

(21.18.02) gforcada: oh god...

(21.18.04) PiotrDrag1: I think it's because testing was somehow limited (Fedora is usually stable at this point, but this time it's hell)

(21.18.26) Tranzistors: this is why we need gnome-os

(21.18.33) gforcada: Tranzistors: for that maybe we should send a mail to desktop-devel-list at the beginning of string freeze?

(21.18.39) gforcada: Tranzistors: agree :)

(21.18.53) PiotrDrag1: Tranzistors: there were also something wrong with totem, like it got branched from early 3.5.x stage

(21.19.39) pecisk: there are problems with testing translations soon enough, because at the time of GNOME release Fedora isn't really usable enough

(21.19.57) pecisk: and other distros don't even plan to look at packaging stuff

(21.20.16) PiotrDrag1: pecisk: well, usually it is, just this time the installer re-write broke it

(21.20.28) PiotrDrag1: but true about other distros

(21.21.30) Tranzistors: It seems that already .0 releases are not really shipped anyway and final ones will be from .1

(21.21.31) gforcada: PiotrDrag1: as Tranzistors said, let's hope gnome-os effort can help us on that :)

(21.21.34) pecisk: also I think we need some QA guidelines, just give users to test it is clearly not enough

(21.22.15) gforcada: pecisk: that would be nice, would you want to volunteer on writing them down? :)

(21.23.03) pecisk: gforcada, I could try, no promises though

(21.23.26) pecisk: there are several ideas swirling in my mind...I will try to just to write them down

(21.23.52) gforcada: pecisk: no pressure :) just start them on and keep giving updates on gnome-i18n mailing list

(21.28.17) gforcada: any more to add?

(21.28.35) Tranzistors: no, all other things were ok

(21.28.45) pecisk: btw, we recently tried to do spoaristic survey how people grade our open source translations

(21.29.22) gforcada: pecisk: interesting, how was that?

(21.30.16) pecisk: gforcada, well, most people said that it takes time to used to it

(21.30.18) pecisk: but

(21.30.26) pecisk: when it does it doesn't hurt

(21.30.33) gforcada: :)

(21.31.55) pecisk: altough to get scientistic answer you would have to do full blown survey, because most of people see computer as a tool - as long as they can understand it (trough language of interface or intuition - they don't care)

(21.32.05) pecisk: power users has a problem usually though

(21.34.50) gforcada: any more to add then? :) let's move to next one: new design (let's leave the GUADEC BoF points for last one)

(21.35.04) Tranzistors: oki

(21.35.45) pecisk: :)

(21.36.47) gforcada: comments, critics... ? :) The RTL situation I think is not quite right ... anyone is able to actually test it? I'm not able to see if a UI is right or not for an RTL language

(21.37.09) PeaceKnight: gforcada, it's not :-(

(21.37.33) pecisk: new design of what?

(21.37.46) gforcada: pecisk: of :)

(21.38.00) gforcada: PeaceKnight: are you HTML/CSS literate? I can help you setup the dev environment

(21.38.20) PeaceKnight: hi all

(21.38.20) PeaceKnight ezentúl abderrahim néven szerepel

(21.38.22) gforcada: PeaceKnight: or if you make screenshots of what's wrong and how should it look like I can try fixing it

(21.39.10) gforcada: abderrahim: actually that can be easier, just provide side-by-side screenshots, is that something you can do?

(21.39.14) abderrahim: Well, I don't quite have the time now, I just came because of the meeting

(21.40.06) abderrahim: but I can try giving some screenshots

(21.40.19) gforcada: abderrahim: no hurries, just whenever you have time, please report them on bugzilla, so it doesn't get lost

(21.40.32) abderrahim: The good point is that it doesn't seem to load fonts like the old one (and that's better because it makes it usable in epiphany in arabic)

(21.40.33) gforcada: abderrahim: I will try to fix them as soon as I can :)

(21.40.46) Tranzistors: why is translation statistics in D-L in <span> tags, and not in table format?

(21.41.05) gforcada: abderrahim: oh, cool, actually it should though (load fonts, it should load Cantarell)

(21.41.16) nitalynx: Yeah. How about some actual tables for the stats? Please?

(21.41.25) gforcada: Tranzistors: where is that?

(21.41.31) gforcada: nitalynx: where?

(21.41.38) Tranzistors: gforcada,

(21.41.40) nitalynx: gforcada:

(21.42.04) abderrahim: How can I see how it looks in English? (without changing my settings, that is)

(21.42.17) Tranzistors: gforcada, <b> 89%</b> <span class="num1">   180</span> <span class="num2">   11</span> <span class="num3">   11</span>

(21.42.50) gforcada: abderrahim: unfortunately you can not.. I used two browsers to try to fix the front page on arabic :S you have to change your accept-language header

(21.43.40) gforcada: Tranzistors: that's for coloring... but what's wrong on using spans?

(21.44.29) nitalynx: gforcada: You can't easily paste them into a spreadsheet, for instance.

(21.45.03) gforcada: nitalynx: oh, I see, but we do have some export formats, don't we? Instead of doing a dirty-hack just file a bug for proper export formats :)

(21.46.01) nitalynx: Also, using tables for tabular data is The Right Thing To Do :P

(21.46.11) Tranzistors: :)

(21.46.25) gforcada: nitalynx: yeah, but not if you want to color them, sort and so on...

(21.46.46) gforcada: nitalynx Tranzistors: or put it in another way ... patches are welcome :P

(21.47.31) Tranzistors: gforcada, good point, just wanted to know, if there are good reasons why things are the way they are

(21.48.09) gforcada: Tranzistors: well, I actually didn't do the styling I just pushed it :) so we should ask Vinicius maybe :)

(21.48.30) abderrahim: looks like my internet connection is too slow this evening

(21.48.55) Tranzistors: gforcada, btw, this was also the case for previous version

(21.49.21) gforcada: Tranzistors: ok, so then Claude :)

(21.49.35) abderrahim: Am I still connected?

(21.49.35) abderrahim elhagyta a szobát (quit: Ex-Chat).

(21.50.13) gforcada: abderrahim: seems not ...

(21.50.15) pecisk: not anymore I suppose

(21.51.26) gforcada: so, let's switch to the "big thing"? (GUADEC BoF points)

(21.51.40) gforcada:

(21.51.51) abderrahim: I'm back

(21.51.56) gforcada: abderrahim: welcome back!

(21.51.58) abderrahim: at what point I got disconnected?

(21.52.21) Tranzistors: abderrahim, right after “Am I still connected?”

(21.52.23) gforcada: you haven't lost anything, just being picky about tables was it :)

(21.53.18) abderrahim: Ok

(21.53.27) gforcada: anyone want to discuss any specific point first?

(21.53.33) claude: about tables vs spans, i don't remember right now, it may be related to the header, but test with a patch and see what's the outcome

(21.53.58) gforcada: claude: oh, you are around :) hi!

(21.54.37) abderrahim: but the last thing I read was gforcada about side-by-side screenshot, did I miss something

(21.55.27) gforcada: abderrahim: there were some questions about why tables are not used properly on tables like

(21.56.10) abderrahim: anyway, I'll try to provide screenshots for what looks weird. Where should I post them?

(21.56.27) gforcada: abderrahim: to bugzilla please, so that we do not lost them!

(21.57.12) abderrahim: ok

(21.57.33) gforcada: so as anyone is bringing any topic ... words! What do you think about switching completely (minus the translation detail) to words?

(21.58.13) PiotrDrag1: gforcada: I would rather dislike such a change

(21.58.25) Tranzistors: not a great fan either

(21.59.30) abderrahim: I wouldn't mind, but I'm curious to how the stat^w lies would look like

(21.59.31) gforcada: ok ... why not? It gives you a more clear view of the effort remaining...

(22.00.16) PiotrDrag1: maybe I'm just used to think in the categories of strings, but I find words unclear

(22.00.36) Tranzistors: reason — if a project needs to get basic UI done and leave the error messages behind, then stats will look disproportionally bad

(22.01.04) Tranzistors: this is especially important, when the 80% barrier is considered

(22.01.16) Tranzistors: which is a whole new topic

(22.01.28) gforcada: Tranzistors: but that's not a reason for not using words, that means that the whole thing (sure a new topic) has to be addressed first

(22.01.48) friedel: we might be interested in two things: coverage, and work

(22.01.49) gforcada: we do have the reduced po files but we are actually not using it on the global listings (AFAIK)

(22.02.05) friedel: strings give a better indication of coverage, and words give a better indication of work/effort (in my opinion)

(22.02.24) kelemeng: gforcada: hi, I think there is one more reason to keep words: marketing. to be able to say "56 languages are 80%+ translated, thus supported"

(22.02.27) Tranzistors: friedel, agree

(22.02.47) PiotrDrag1: how about keeping both words and strings? is it that cluttering?

(22.03.23) kelemeng: friedel: exactly. IMHO, users care more about coverage, i.e. five untranslated short strings bother them more than one long

(22.03.42) gforcada: friedel: good point on the work/effort ...

(22.03.43) friedel: of course, it is still all lies

(22.04.14) gforcada: PiotrDrag1: I guess so (cluttering) at least from time to time I get a bit confused of seeing one of them but not the other and so on

(22.04.33) gforcada: friedel: :)

(22.04.37) kelemeng: PiotrDrag1: I think so. maybe generate both, and make it possible to switch between them?

(22.04.44) PiotrDrag1: gforcada: yeah, sure, but maybe some little changes in layout would be enough?

(22.04.52) PiotrDrag1: kelemeng: +1

(22.05.11) kelemeng: so that translators who are thinking in word could see words, and marketing people see coverage

(22.05.21) gforcada: PiotrDrag1: layout changes are also fine with me, I'm not a UI expert though, patches welcome :)

(22.05.57) kelemeng: but I can not promise to offer any help with this :(

(22.06.25) afranke: gforcada, didn't we say 9pm UTC?

(22.06.43) afranke: (which is in 1 hour from now)

(22.06.43) PiotrDrag1: afranke: 21UTC+2

(22.06.58) afranke: *sigh*

(22.07.24) gforcada: afranke: sorry that was my fault!

(22.08.15) ***afranke just checked, no mail says UTC+2.

(22.08.30) ***gforcada hides somewhere and edits the wiki ...

(22.08.56) afranke: So now I missed the meeting (and I could have been there).

(22.09.00) afranke: I'll read the log.

(22.09.13) nitalynx elhagyta a szobát (quit: Leaving.).

(22.09.41) gforcada: afranke: yeah sorry ... I was really swamped this week and I rushed to create everything yesterday :S

(22.10.24) claude: French people are always late

(22.10.27) ***claude runs

(22.10.46) gforcada: afranke: which brings another point to the table, next meeting in 3 months is ok for everyone? On January, maybe 2nd week of?

(22.11.17) Tranzistors: cool with me

(22.11.45) afranke: gforcada, yep, 3 months should do it.

(22.11.48) kelemeng: gforcada: I think that would be too infrequent. I'd like to propose monthly meetings.

(22.11.59) kelemeng: it's not like there is nothing to do...

(22.12.19) afranke: Yes but will we actually have progressed in one month?

(22.12.21) afranke: I doubt so.

(22.12.22) gforcada: kelemeng: I will not be in Berlin by next month ... but if people like to do it monthly .

(22.12.28) afranke: Make it at least 2 months.

(22.13.17) Tranzistors: monthly would be cool

(22.13.22) gforcada: I'm open to any option, just so that we have fixed times to be able to blog/buzz about it  :)

(22.13.41) pecisk: do monthly, if nothing to talk about at least people will hang out at this channel and tell funny stories

(22.14.01) PiotrDrag1: +1 for monthly

(22.14.02) gforcada: :)

(22.14.09) PiotrDrag1: we can always change it some time later...

(22.14.15) pecisk: yep

(22.14.19) gforcada: ok cool, let's do it monthly :)

(22.15.22) afranke: I have a topic I'd like to discuss (didn't have time to add it to the agenda and didn't check whether you discussed it already).

(22.15.32) gforcada: afranke: go ahead :)

(22.15.33) afranke: Accelerators in app menus.

(22.15.39) gforcada: yep

(22.15.40) afranke: What's up with them?

(22.15.47) afranke: Define yep.

(22.15.48) afranke: :)

(22.15.55) afranke: Already discussed?

(22.16.04) PiotrDrag1: afranke: did you file a bug against gtk?

(22.16.09) gforcada: no no, just like, "yes good topic" :)

(22.16.19) afranke: No, first I want to know the situation, PiotrDrag1

(22.16.22) gforcada: not me, someone said that would do it, no idea if it's done

(22.16.23) PiotrDrag1: I'd be interested in CCing

(22.16.33) afranke: Ok. Let me introduce the problem/questions.

(22.17.01) afranke: A *lot* of apps removed their accelerators as the transitionned to GMenu.

(22.17.20) afranke: GNOME Shell indeed doesn't use accelerators in app menu, so in this case it makes sense.

(22.17.42) afranke: But in other environments, the app menu is concatenated with the rest of the menus.

(22.18.04) afranke: So I guess the accelerators should still be used.

(22.18.06) afranke: BUT

(22.18.20) afranke: I don't know if GNOME Shell actually has a problem with accelerators.

(22.18.45) afranke: Maybe it just won't use them, maybe it behaves badly if they are present.

(22.18.52) afranke: Does anyone know anything about it?

(22.19.03) gforcada: should we bring it to d-d-l?

(22.19.24) PiotrDrag1: (some GMenu apps have accelerators, so g-shell probably have no problem with them)

(22.19.24) PiotrDrag1: I wonder how a11y team would look at this, accelerators are afaik important for them

(22.20.09) gforcada: the most important thing would be to have clear guidelines, because there's lots of apps with accelerators and lots without...

(22.21.51) afranke: gforcada, I think d-d-l is indeed a good place to discuss that.

(22.22.10) afranke: I can start the thread if you want.

(22.22.26) afranke: Not before monday though. :)

(22.22.35) gforcada: please, whenever you have time

(22.22.43) afranke: Ok, I'll do that.

(22.22.50) PiotrDrag1: afranke: great!

(22.23.08) afranke: BTW nice work on the new design.

(22.23.21) gforcada: the desgin pages on the wiki or the GNOME Goal does mention anything about accelerators?

(22.23.40) afranke: I don't know. Can you check?

(22.24.39) PiotrDrag1: afranke: nothing from what I can see

(22.24.56) PiotrDrag1: ops, that was an answer for gforcada question

(22.25.24) PiotrDrag1: but example commit (nautilus) does have accelerators

(22.26.05) PiotrDrag1: so maybe filing bugs against specific components asking for adding _ would resolve the issue?

(22.26.54) gforcada: PiotrDrag1: I would guess so, but maybe a question on d-d-l asking about it would bring the attention to it and the designers can say something ...

(22.26.56) PiotrDrag1: or maybe even asking d-d-l for global approval and then add them by ourselves?

(22.27.09) gforcada: PiotrDrag1: that would work too

(22.27.14) PiotrDrag1: I could do that, providing we have developers approval

(22.27.33) PiotrDrag1: so I'm waiting for afranke's mail :)

(22.27.45) gforcada: then you will definitely need afranke's mail :)

(22.30.10) gforcada: should I bring another topic? On GUADEC's BoF we said to do another survey, anyone volunteering to do it? Starting collecting questions on gnome-i18n mailing list maybe ...

(22.30.32) ***afranke finished reading the log.

(22.31.34) afranke: I think I volunteered at the BoF for the survey, but I'm not sure I remember the topics that should be covered.

(22.32.02) gforcada: afranke:

(22.32.11) gforcada: seems I was volunteering too :)

(22.32.42) gforcada: we could do a quick round (say a week) of collecting questions from everyone and then decide if it's actually worth or not...

(22.33.42) gforcada: but I think that one point that we wanted to make was to have a broader view of coordinator's responsiveness, so a really small survey would be ideal (something that you can do in 5 minutes time)

(22.39.59) gforcada: kelemeng: what about the dropping docbook? Have you contacted the docs team? :)

(22.40.10) gforcada: I mean

(22.40.27) kelemeng: gforcada: um, not, sorry.

(22.41.25) gforcada: no problem ;) will try to do that for the next meeting? :P

(22.41.29) kelemeng: I think I can write a mail or two during the next weeks

(22.41.55) kelemeng: absolutely! without deadlines, nothing would ever be finished :P

(22.41.55) gforcada: cool!

(22.42.14) kelemeng: also, there is one BoF item I'd like to discuss - splitting modules:

(22.42.14) kelemeng:

(22.42.14) kelemeng: I didn't had time to work on this yet, but I think it would be awesome to have this done by 3.7.1 time. (oct. 22, or whenever the 3.8 category will appear - and yes I volunteer to move things forward :))

(22.42.25) kelemeng: There is one thing I don't really understand about the current proposal: what's the difference between core and extra apps? can we define that exactly?

(22.43.25) gforcada: kelemeng: that's one of my top priorities also, the thing is that we have to somehow show that on

(22.43.26) kelemeng: other than that, I think the proposal is solid enough to give it a go (probably after asking d-d-l's opinion)

(22.44.56) kelemeng: gforcada: you proposed the categories, could you give a scientific-ish definition to the categories? :)

(22.45.11) gforcada: the other day I was thinking about it ... as a translator you don't care at all if a module is desktop, platform, or whatever ... we do care if it's a library (thus hardly noticeable strings) or a UI module (nautilus, gnome-shell ...) so I would say to move away from release set categories from release team

(22.45.36) gforcada: I mean, using the modules defined them but rearranged to suit our needs

(22.46.11) afranke: Well desktop and platform have different audiences.

(22.46.39) kelemeng: afranke: +1. developers can wait :)

(22.47.29) PiotrDrag1: that would mean gtk+ in one category, and gtk+-properties in another? I'm all for that

(22.48.00) PiotrDrag1: there is some fuzzy ground however (some glib strings are shown in nautilus etc.)

(22.48.09) kelemeng: PiotrDrag1: and gtk+-properties at the very bottom of the list :P

(22.48.16) PiotrDrag1: kelemeng: :D

(22.48.29) kelemeng: PiotrDrag1: really? like when?

(22.48.29) afranke: kelemeng, for Dropping docbook, there's but it looks unmaintained.

(22.49.02) PiotrDrag1: kelemeng: maybe I'm wrong, but I thought some GIO strings are shown when mounting

(22.49.31) PiotrDrag1: anyway, +1 for the proposal

(22.50.13) gforcada: PiotrDrag1: yeah the grey area will be really hard to sort, but I think that it's more important to have first wins that having to go through gtk+-properties and clutter (and cogl!) pain...

(22.50.34) PiotrDrag1: gforcada: yup

(22.51.29) afranke: I'd say for those in the grey area, put them in the most prioritized set.

(22.51.36) afranke: This way you don't risk too much.

(22.51.40) gforcada: the thing that worries me is how to show that on the UI, using the categories on is quite limited because we do not show them anywhere else

(22.52.20) gforcada: afranke: or create a second set for that grey area? Like you have your first win (the first set) but now go to the pain to really finish it the first set

(22.53.38) abderrahim elhagyta a szobát (quit: Ping timeout: 600 seconds).

(22.54.10) afranke: Anyway we can see that later when the first things are sorted out. :)

(22.55.40) PiotrDrag1: what about the extras-stable set? it's mostly outdated, there are even some 3-2 branches

(22.55.45) gforcada: yeah, but it will be quite moot to do all the slicing but at the end still count everything as 3.8 for the 80%-ness

(22.56.17) kelemeng: gforcada: I think we were only speaking about new category names, and still keep everything on one page (for simplicity, and for the 80% mark)

(22.56.20) gforcada: PiotrDrag1: needs someone to update them, anyone able to go through them? File a bug maybe so we remember it...

(22.57.02) kelemeng: PiotrDrag1: I think that's another item, namely "dead modules"

(22.57.19) kelemeng: please don't start on that just yet :)

(22.57.22) gforcada: kelemeng: that's an intermediate step then, but "my" final goal is to really split the "one page" for the 80% and so on

(22.57.29) PiotrDrag1: kelemeng: oh, sorry

(22.58.12) kelemeng: gforcada: so that there would be a new set for "gnome-not-so-visible"?

(22.58.17) gforcada: I mean, I really would like to see at some point a translation team starting completely from 0% and be able to have some kind of decent support in 2 releases (so one year)

(22.59.09) gforcada: kelemeng: yeah, naming is always hard, more than not-so-visible, I would name it as "something not so urgent"...

(23.00.42) kelemeng: okay, I think we could cut the list under accessibility. maybe in the next iteration :)

(23.01.35) gforcada: we can discuss that on next meetings maybe when things are a bit more solid maybe, but keep thinking on it :)

(23.01.44) kelemeng: but first things first: what's the difference between core apps and extra apps?

(23.01.49) afranke: gforcada, I think it's important to still have somewhere a way to see the stats for "the real GNOME 3.x".

(23.02.13) gforcada: afranke: agreed 100% but maybe just like (just a second ...)

(23.02.24) afranke: kelemeng, I'd say that's a question for design and release teams.

(23.02.25) kelemeng: I mean empathy is replaceable in the same way as evolution...

(23.02.43) afranke: Is evolution in extra?

(23.02.58) kelemeng: afranke: no. these categories here were invented by gforcada :)

(23.03.05) kelemeng: afranke:

(23.03.08) kelemeng: here

(23.03.15) afranke: kelemeng, see for one possible definition of Core.

(23.03.17) gforcada: so we could see the "core" "core apps" and the sum of all as the last column

(23.04.19) PiotrDrag1: kelemeng: can I edit this wiki page? PolicyKit-gnome is a fallback module...

(23.04.21) Tranzistors: aren't the extra modules those, which have their own release schedule?

(23.04.53) kelemeng: Tranzistors: those are gnome-extra, these are core-extra

(23.05.09) kelemeng: PiotrDrag1: sure

(23.05.14) Tranzistors: kelemeng, oh ok

(23.05.56) gforcada: does seem sane my approach on the stats to kind of replace/enhance ?

(23.08.31) kelemeng: gforcada: I can't really imagine it right now at all, sorry.

(23.08.54) kelemeng: could you answer my question?

(23.09.31) gforcada: kelemeng: about core extra and that? don't bother about them we will recreate them at some point, was just a wait to split things...

(23.10.03) kelemeng: otherwise, I'll propose to not differentiate between these

(23.11.24) kelemeng: okay then, thanks

(23.13.35) kelemeng: let's say I'll update the proposal (also dig into jhbuild modulesets :)), and mail d-d-l to ask their opinion

(23.14.29) gforcada: cool, mail also gnome-i18n with updates on your progress!

(23.15.32) kelemeng: sure

(23.16.19) kelemeng: so, next topic? :)

(23.16.27) gforcada: someone said dead modules? :)

(23.16.43) Tranzistors: more like zombies

(23.16.57) Tranzistors: only kept alive by l10n teams

(23.18.03) kelemeng: also back on the regular IRC topic: I think I'll send a mail to g-i18n@ asking team coordinators about the exact days (like first monday/... of the month)

(23.18.19) gforcada: there was a proposal to try to find modules with only l10n commits, anyone able to do that git sorcery?

(23.18.30) gforcada: kelemeng: cool thanks!

(23.18.32) kelemeng: Tranzistors: that's a way to put it, yes

(23.18.43) PiotrDrag1: gforcada: I can do it manually quite easily

(23.19.08) PiotrDrag1: but if someone have some git-fu, then I'm shut up :)

(23.19.11) kelemeng: gforcada: I think Andre showed the necessary git magic on the list...

(23.19.22) afranke: gforcada, andre.

(23.19.38) pecisk: well, there's frequently that app which is very useful but not supported anyway

(23.19.42) gforcada: PiotrDrag1: send (on bugzilla) that script then! we can implement that on D-L and then send emails to maintainers pestering them to either shutdown that module or keep getting annoying mails :)

(23.19.43) kelemeng: PiotrDrag1: I used to do it manually too, when I was young and had a lot of time... :)

(23.19.47) pecisk: does that constitute as dead module?

(23.20.11) afranke: gforcada, for git sorcery, I'd recommend andre and fredp.

(23.20.40) gforcada: afranke: I'm quite good at it too, but would be cool to ask the masters andre and fredp :)

(23.20.47) kelemeng: pecisk: yes. if there is no release, users won't see our translations, so there is no point to translate.

(23.21.17) pecisk: kelemeng, someone could just pull them

(23.21.19) pecisk: ok

(23.21.25) pecisk: that's just my thoughts :)

(23.22.31) kelemeng: even if there is no code improvement, a maintainer can do a .x release, saying "7 translations updated". if they don't have time even for that...

(23.23.06) pecisk: ok

(23.23.08) pecisk: good point

(23.23.22) afranke: Some modules might even not compile anymore.

(23.23.48) pecisk: propably most of them

(23.23.51) kelemeng: maybe, but that's another problem

(23.24.01) pecisk: that's not our problem :)

(23.24.01) afranke: Not really.

(23.24.46) afranke: If the program doesn't have releases because it's stable and entered maintainance mode, then we should push for releases with new translations.

(23.24.47) gforcada: kelemeng: that's how envision the tool: see how many translations since the last code commit/release and then if the threshold is enough, mail gnome-i18n CC'ing the maintainer :)

(23.25.10) gforcada: afranke: completely agree

(23.25.34) pecisk: if there is a maintaner he will be reminded that he must do release with translations, if there's not...yep, sounds nice

(23.25.35) afranke: If it's not released because it has become useless, we should make everything we can to avoid new translations so that noone wastes their time.

(23.25.48) pecisk: agreed

(23.25.55) gforcada: we should have some kind of way to show that on (like a comment on a module saying that maybe it's not worth translating this module because it doesn't seem maintained or that makes releases often)

(23.26.13) gforcada: afranke: again 100% agree

(23.26.27) gforcada: there's enough to translate in GNOME to spend time in dead modules

(23.26.41) Tranzistors: or like gconf — deprecated

(23.27.00) gforcada: Tranzistors: unfortunately is still no 100% deprecated I think ... isn't it?

(23.27.15) PiotrDrag1: I can have a list of dead modules on l10n.g.o in half an hour, I'll post in a new bug in bgo

(23.27.21) afranke: Deprecated is not really a good indication. gconf is still used by many modules and is still shipped by distros.

(23.27.44) afranke: Even though it's officially deprecated and replaced by dconf.

(23.27.47) Tranzistors: gforcada afranke, but it is still very low priority

(23.27.53) afranke: Sure.

(23.28.08) Tranzistors: there should be an indicator on D-L for that as well

(23.28.09) gforcada: sure, that's we should make them sorted in a way that they are at the bottom :)

(23.28.10) kelemeng: priority is another topic, will be solved soon-ish :)

(23.28.22) gforcada: hopefully yes!

(23.29.07) kelemeng: PiotrDrag1: that would be awesome, also please take care to show us how to do that :)

(23.29.27) afranke: To do what? Set priorities?

(23.29.51) PiotrDrag1: kelemeng: click on every suspicious module in extras, click on it, see there is no release in 9 years, add it to the list ;)

(23.29.57) ***afranke hopes those will be defined once for all languages and not by coordinators per language.

(23.30.26) kelemeng: afranke: no, I meant the dead modules list

(23.30.27) afranke: Oh, "do that" == find dead modules! Sorry. :)

(23.30.56) kelemeng: PiotrDrag1: that way, half an hour wont't be enough. been there, done that :P

(23.31.25) PiotrDrag1: kelemeng: I underestimate me! half an hour is probably more than I need ;)

(23.31.30) gforcada: that's why we need to make D-L warn by itself

(23.31.54) kelemeng: also, one more thing - how do we want to define "dead"?

(23.32.10) kelemeng: right now: 2 years without commits (asking the maintainer before)

(23.32.21) kelemeng: this is what the wiki says

(23.32.54) kelemeng: do we want to redefine this? there was no agreement on this at the Bof.

(23.32.58) kelemeng: IIRC...

(23.33.00) afranke: BTW if a module is dead, it should also be moved to archives on so please also file bugs for that.

(23.33.38) PiotrDrag1: afranke: I personally think we should be more aggressive on removing dead modules than r-t

(23.34.04) kelemeng: afranke: I think git maintainers are a bit more liberal on this topic than we should

(23.34.06) gforcada: what I would propose is to send automatic emails from D-L if in six months there has been no release but there has been translation updates

(23.34.41) kelemeng: gforcada: or manual emails, until this feature is implemented :)

(23.35.28) kelemeng: and how about throwing things out after one year of no release, *despite* email warnings?

(23.36.00) gforcada: kelemeng: I hope we do not get to that situation... but we could move to another set...

(23.36.48) afranke: PiotrDrag1, kelemeng, if the module deserves it, a bug should be filed. bkor would like maintainers to be able to do it themselves but in the meantime he'll happily solve such requests.

(23.36.50) kelemeng: gforcada: sure, I was thinking about the obsolete set

(23.37.11) gforcada: kelemeng: or "grey sink" :D

(23.37.22) ***afranke has already made many such reports and many modules were archived this way.

(23.37.39) ***gforcada applauses afranke

(23.38.04) ***afranke was motivated by something else at that time though: his module was pushed to page 2 because of other dead modules.

(23.38.04) afranke: ;)

(23.38.34) afranke: gforcada, +1 for mail after six months.

(23.38.47) gforcada: :D

(23.39.01) kelemeng: okay, so to sum up: can we agree on the new thresholds? 6 months = warning, 12 = obsolete set?

(23.39.41) kelemeng: that would also mean that *someone* would start to check regularly for modules becoming obsolete

(23.39.55) gforcada: 6 months warning and 12 report on bugzilla and obsolete set?

(23.40.08) afranke: gforcada, I propose "Antiquated stuff" for the name of the set collecting dead modules. :)

(23.40.12) gforcada: hopefully that one will be D-L :)

(23.40.17) gforcada: afranke: +1

(23.40.24) kelemeng: and handling the emails... that would be quite some work.

(23.40.58) gforcada: kelemeng: not if D-L does it, we just need to parse the reply from the maintainers

(23.41.12) kelemeng: gforcada: sure, but in the meantime... we already have a lot of zombies, draining translator brains

(23.41.20) gforcada: :D

(23.41.40) gforcada: on gnome extras set you mean?

(23.41.43) kelemeng: so I'm talking about short term...

(23.41.50) kelemeng: yes

(23.42.03) kelemeng: where else? :)

(23.43.24) kelemeng: gforcada: is there an easy way to get the list of module names in a given set?

(23.43.39) ***kelemeng already thinks on a script...

(23.44.05) gforcada: kelemeng: describe that script on a bugzilla report so that it does not get lost and it gets at some point implemented

(23.44.19) afranke: Slighty off-topic (but not so much since we touched the topic of brains being drained): couldn't we mark modules that are "not yet frozen" to avoid new contributors wasting their time?

(23.44.22) gforcada: kelemeng: I think so (getting the module names) but is not enough the listings itself?

(23.44.50) gforcada: afranke: care to elaborate? I can not grasp the full meaning of it

(23.45.47) kelemeng: gforcada: I think afranke is thinking about the problem we have listed on the wiki under "Calendar"

(23.46.00) afranke: Say we're in June. It's too early to start working on those three untranslated strings in eight modules.

(23.46.05) afranke: kelemeng, yes, actually.

(23.46.20) gforcada: oh, sorry, yeah

(23.46.39) ***kelemeng is a mind-reader :D

(23.47.11) gforcada: we could add a warning on the GNOME set ones that shows when they are supposed to be released (September - April) so that the translator knows about it

(23.47.18) gforcada: sounds sane?

(23.47.29) afranke: This cycle, we've had at least two contributors that have worked on *many* modules that had only a couple of strings and then all those modules were marked as translated at the beginning of string freeze even though they were far from being translated.

(23.47.32) gforcada: btw ... we should add bug reports about all ideas, or we will forget all of them :)

(23.48.31) afranke: And coordinator telling them to stop and focus on other things was not enough because they cared about the lies (reaching 100% for a moving target).

(23.49.45) afranke: So I think it should be done at an upper level. Maybe having a yellow background for the comment textarea and a warning stating that really, you shouldn't be translating that yet.

(23.49.47) kelemeng: afranke: I had an idea to solve this, but didn't filed a bug report or added to the wiki... sorry.

(23.50.34) PiotrDrag1: [I've got the list with 53 modules, most of them not even touched since svn>git migration]

(23.51.01) afranke: By upper level I mean Damned lies telling you you're wrong instead of "some random guy" (because to newcomers the coordinator may be seen as such).

(23.51.53) afranke: PiotrDrag1, did you check whether some of them are already moved to Archive/ on

(23.52.10) gforcada: we could show a red banner saying that is useless to translate that module? At the end though, we can't force people to not work on something ...

(23.52.15) PiotrDrag1: afranke: they aren't

(23.52.17) gforcada: PiotrDrag1: amazing! send the list on gnome-i18n

(23.52.37) afranke: PiotrDrag1, cool ;)

(23.52.46) PiotrDrag1: gforcada: I've filed a bug

(23.52.49) afranke: (More dust to remove)

(23.52.53) kelemeng: afranke: I'll document my idea & do the homework (sending e-mails), please give me a little time :)

(23.52.54) gforcada: cool

(23.52.59) PiotrDrag1:

(23.53.02) gforcada: thanks!

(23.53.12) gforcada: kelemeng: great

(23.53.31) afranke: gforcada, that's why I'm in favor of a strong warning rather than a strict enforcement.

(23.53.47) gforcada: afranke: sure, agreed

(23.53.51) afranke: kelemeng, sure, great.

(23.56.19) gforcada: should we leave it for today? (almost tomorrow) :)

(23.56.27) kelemeng: PiotrDrag1: could you add some reasoning to the bug on why should they be moved?

(23.56.34) gforcada: one last thing though, who organizes the next meeting? :)

(23.56.40) kelemeng: like "last release 1+ year ago" ?

(23.56.57) PiotrDrag1: kelemeng: sure, just give me some more time

(23.57.56) kelemeng: gforcada: I guess I'll do... given that I already volunteered to get opinions about when would that be good to people :)

(23.58.02) kelemeng: PiotrDrag1: thanks

(23.58.05) gforcada: kelemeng: amazing!

(23.58.33) gforcada: thanks *a lot* to all of you!!

(23.58.53) pecisk: thanks for making this happen :)

(23.58.59) kelemeng: would someone post meeting minutes to the list, for those interested?

(23.59.11) gforcada: btw ... anyone has the full log? I just noticed that empathy just cuts it ....

(23.59.24) kelemeng: I think I have :)

(23.59.32) gforcada: kelemeng: I could try if someone has the full log and pastes it to

(23.59.33) ***kelemeng is still using pidgin

(23.59.39) gforcada: :D

TranslationProject/Events/IRCMeeting20121006 (last edited 2012-10-25 21:19:36 by GilForcada)