From Fedora Project Wiki

< I18N‎ | Meetings

Fedora I18N Meeting 2009-02-17



New items:

  • fontconfig issues in rawhide and F10
  • Unicode 5.1 support for indic - RahulBhalerao
  • firefox indic locales smoke tests -RahulBhalerao
  • Open floor
  • Next meeting


juhp hello Feb 17 15:00
dychen_ hi Feb 17 15:01
juhp welcome to the fedora-i18n meeting today Feb 17 15:01
pravins hi Feb 17 15:01
rahulb hi everyone Feb 17 15:01
tagoh3 hi Feb 17 15:01
juhp I18N/Meetings Feb 17 15:02
*juhp added an item on fontconfig Feb 17 15:02
-->asgeirf (n=asgeirf@nat/redhat/x-a78c2dd4c55184e5) has joined #fedora-meeting Feb 17 15:02
-->sflanigan (n=sflaniga@nat/redhat/x-6c583e17b81d22e0) has joined #fedora-meeting Feb 17 15:03
-->jni_ (n=jni@ has joined #fedora-meeting Feb 17 15:04
juhp ok let's get started Feb 17 15:04
-->K410 (n=cchance@fedora/Kaio) has joined #fedora-meeting Feb 17 15:04
phuang_ hi Feb 17 15:05
juhp so just a few ongoing items first Feb 17 15:05
K410 hi Feb 17 15:05
-->mintos (n=mvaliyav@nat/redhat-in/x-9f1ceb802f4a4bac) has joined #fedora-meeting Feb 17 15:05
juhp * ibus Feb 17 15:05
juhp ibus seems to be working ok now in rawhide Feb 17 15:05
phuang_ yeah. Feb 17 15:06
juhp though still some things to improve Feb 17 15:06
juhp I filed a couple of bugs this morning Feb 17 15:06
phuang_ I saw them. and I am working on them Feb 17 15:06
phuang_ I am still consider how to decide which IM should be load by default Feb 17 15:07
juhp hmm great time for my bugzilla password to expire... sigh Feb 17 15:07
juhp phuang_: ok - the algorithm or implementation? Feb 17 15:08
juhp or the list of IMEs? Feb 17 15:08
dychen_ juhp: I think phuang_ means the default engine for each locale Feb 17 15:08
phuang_ juhp, I don't think hardcode IMEs list is good Feb 17 15:08
juhp phuang_: how else to handle langs with multiple choices? Feb 17 15:09
phuang_ multiple choices? Feb 17 15:09
juhp priorities? Feb 17 15:09
juhp s/choices/engines or maps/ Feb 17 15:10
phuang_ juhp, btw, if user set some imes by self, and change the locale later, we how to handle this situation Feb 17 15:10
juhp like Chinese or Indic Feb 17 15:10
juhp phuang_: I would say do nothing Feb 17 15:10
juhp though there could be a default mode (ie with no setting) use the default for the locale Feb 17 15:11
juhp but once user has set then I think ibus should just follow that choice Feb 17 15:11
tagoh3 sounds reasonable Feb 17 15:12
phuang_ juhp, another situation: at first a user use ja locale, and then changed to cn locale, what we should do? Feb 17 15:12
juhp choosing default engine by priority seems reasonable too - though it requires all the IME to maintain that though Feb 17 15:12
phuang_ juhp, and the user not do any setting for ibus Feb 17 15:13
juhp phuang_: well I think I answered that - I suggest either keep initial IME or maybe better(?) which to cn default? Feb 17 15:13
juhp sorry /which/switch/ Feb 17 15:14
juhp I don't think that case is as important though as having a default Feb 17 15:14
juhp enabled Feb 17 15:14
phuang_ ok Feb 17 15:15
juhp typically users would not switch desktop between zh and ja (maybe to/from en might be more common) Feb 17 15:15
juhp probably for en, etc don't need a default engine Feb 17 15:16
phuang_ yeah, maybe en to cn, or ... Feb 17 15:16
juhp right Feb 17 15:16
juhp dychen_, K410: how are you guys going with porting the chewing and table engines? Feb 17 15:16
phuang_ or cn to en. I think user will want to keep default settings on cn locale Feb 17 15:17
-->sdziallas ( has joined #fedora-meeting Feb 17 15:17
juhp K410: acevery's code is in git? Feb 17 15:17
juhp phuang_: right Feb 17 15:17
dychen_ juhp: ibus-chewing is working, just need a few polish. Feb 17 15:17
K410 just merged acevery's master Feb 17 15:17
juhp phuang_: so yeah probably best to keep current choice Feb 17 15:17
juhp dychen_: cool Feb 17 15:17
K410 now test and dev based on his progress Feb 17 15:17
juhp K410: does it work? Feb 17 15:18
K410 user operation is working, confirming quality now Feb 17 15:18
juhp anyway encourage all you who use inputmethods to file bugs for anything you spot and also rfe for features that are missing Feb 17 15:18
juhp K410: ok good Feb 17 15:19
pravins sure Feb 17 15:19
juhp there are only two left for feature freeze Feb 17 15:19
juhp * fonts repackaging Feb 17 15:20
juhp well not too much to say here but still tracking this Feb 17 15:20
juhp I think tagoh3 updated vlgothic and rahulb lohit so that is good Feb 17 15:21
rahulb juhp, yep Feb 17 15:21
juhp I went through all the open bugs and labelled the summaries with the font names (fun;) Feb 17 15:21
rahulb I did culmus and lohit so far Feb 17 15:21
juhp the numbers are going down Feb 17 15:21
juhp rahulb: yes thanks Feb 17 15:21
juhp I guess there are still a few more to do... Feb 17 15:22
juhp maybe we should have a F11I18n tracker or something Feb 17 15:22
rahulb yep, most others are single font packages, would be easier Feb 17 15:22
juhp un-core-fonts has not been done yet either Feb 17 15:23
juhp rahulb: cool Feb 17 15:23
juhp I may have to do that one... Feb 17 15:23
<--abadger1999 has quit (Read error: 60 (Operation timed out)) Feb 17 15:23
juhp I will just comment here on another smaller item that I dropped from earlier - about i18n-team cvs mails bouncing - so actually it is due to the committer's fp.o email not being on the mailing list Feb 17 15:25
juhp finally worked that out... so just thought I would mention that hear Feb 17 15:26
tagoh3 no bouncing anymore you mean? Feb 17 15:26
juhp so subscribing to fedora-i18n-bugs with one's fp.o address would allow the commit mails fwiw Feb 17 15:27
juhp no :) Feb 17 15:27
tagoh3 aha. cool Feb 17 15:27
<--sdziallas has quit ("Ex-Chat") Feb 17 15:27
juhp but yeah we could try drop cvscommit mails for i18n-team, hmm Feb 17 15:27
juhp K410: forgot how about subpackaging for ibus-table tables btw? Feb 17 15:28
K410 cangjie? Feb 17 15:29
juhp K410: I thought the tables were going to be separated upstream Feb 17 15:29
juhp K410: that too Feb 17 15:29
juhp erm sorry not subpackaging Feb 17 15:30
juhp I meant separating to a new package Feb 17 15:30
K410 upstream has i-t-cangjie and i-t-erbi Feb 17 15:30
juhp hope that will still happen in time for f11? Feb 17 15:30
juhp K410: aha just those two? Feb 17 15:30
juhp I think it would be better to decouple the data completely if possible Feb 17 15:31
K410 on git, yes Feb 17 15:31
juhp ok Feb 17 15:31
juhp let's move to the next topic Feb 17 15:31
juhp * fontconfig issues in rawhide and F10 Feb 17 15:31
K410 sorry, wubi, extraphrase, zhengma too Feb 17 15:31
juhp this is an interesting one Feb 17 15:31
K410 (github collapsed that up) Feb 17 15:32
juhp K410: ok cool - all separate projects now? Feb 17 15:32
K410 yes Feb 17 15:32
juhp ok that is a lot packages then... Feb 17 15:32
K410 will do it right after api fix Feb 17 15:32
juhp K410: maybe time to starting packaging them too when you have time Feb 17 15:32
juhp cool Feb 17 15:32
juhp so we thought we have fixed the fontconfig issues for CJK in rawhide and now they turn up for f10 too it seems Feb 17 15:33
juhp tagoh3: particularly since vlgothic was pushed to f10 updates? Feb 17 15:34
juhp hmm probably I should have focus the item text a little more carefully Feb 17 15:35
tagoh3 juhp: yes. I should push updates synchronously since fontconfig conf in other font packages is relevant to fix this Feb 17 15:35
juhp okay Feb 17 15:35
juhp tagoh3: so we need to update all the main cjk fonts? Feb 17 15:35
tagoh3 juhp: if we've changed them to apply conf for locale-specific, yes. Feb 17 15:36
juhp and how about vlgothic for all sans? Feb 17 15:36
juhp ok Feb 17 15:36
juhp Since we didn't hear anything from Behdad still unclear if this is a side-effect of fontconfig changes upstream, or due to the new packaging or new intended fontconfig behaviour? Feb 17 15:37
-->abadger1999 (n=abadger1@nat/redhat/x-7d57d825d3478398) has joined #fedora-meeting Feb 17 15:38
juhp tagoh3: probably I should file a bug for that? Feb 17 15:39
juhp anyway let's go to the next topic Feb 17 15:39
tagoh3 a bug what? Feb 17 15:39
K410 on rawhide, I have done the .conf patch on cjkuni side Feb 17 15:39
tagoh3 for font packages on F-10? Feb 17 15:39
juhp tagoh3: I mean seeing vlgothic for ascii on the desktop Feb 17 15:39
juhp ja desktop Feb 17 15:40
tagoh3 ah Feb 17 15:40
juhp was trying to reproduce on a test instance but didn't get there yet Feb 17 15:40
tagoh3 sure. dunno which packages the bug is really for. Feb 17 15:41
juhp * Unicode 5.1 support for indic - RahulBhalerao Feb 17 15:41
-->linuxguru_ (n=linuxgur@unaffiliated/linuxguru) has joined #fedora-meeting Feb 17 15:41
rahulb I was working on 5.1 support in pango Feb 17 15:41
rahulb figured out pango behavior for all new characters Feb 17 15:41
-->thomasj_ ( has joined #fedora-meeting Feb 17 15:42
rahulb there is bug filed on fedora for the same Feb 17 15:42
juhp aha Feb 17 15:42
buggbot Bug 485050: medium, low, ---,, NEW, Unicode 5.1 support in pango - Indic Lanuages Feb 17 15:42
rahulb also filed it upstream with a patch for indic module Feb 17 15:42
rahulb waiting for behdad to commit it Feb 17 15:42
rahulb if that is done the rendering support will be complete Feb 17 15:43
juhp sounds good Feb 17 15:43
juhp rahulb: you could link the fedora bug to upstream if you like Feb 17 15:43
rahulb still some designing contribution needed on fonts side Feb 17 15:43
juhp ah yes Feb 17 15:43
rahulb juhp, I have just mentioned the upstream bug number there Feb 17 15:43
juhp yep Feb 17 15:44
rahulb anyway that was all about it Feb 17 15:44
juhp I just meant "External Bugs" Feb 17 15:44
juhp thanks Feb 17 15:44
rahulb ok Feb 17 15:44
juhp * firefox indic locales smoke tests Feb 17 15:45
rahulb nothing much on the next topic, just like to mention that Feb 17 15:45
juhp for 3.1? Feb 17 15:45
rahulb yep Feb 17 15:45
rahulb actually I met Set Bindernagel at Gnunify last week Feb 17 15:45
juhp so you're planning so testing of that there? Feb 17 15:45
juhp aha Feb 17 15:45
rahulb and we had some discussion about moving 'beta' of localized ff to live Feb 17 15:46
juhp ah good Feb 17 15:46
rahulb and thus there needs some testing to be completed Feb 17 15:46
juhp rahulb: so it means upstream smoke tests? Feb 17 15:46
rahulb its almost in a final stage Feb 17 15:46
rahulb yep Feb 17 15:46
juhp excellent Feb 17 15:46
rahulb once that is done, very soon many indic loacles will come live for firefox Feb 17 15:47
rahulb thanks juhp Feb 17 15:47
juhp rahulb: does it affect 3.1b3? Feb 17 15:47
juhp or that is probably already frozen? Feb 17 15:47
rahulb depends on the translation status of particular language Feb 17 15:47
juhp okay Feb 17 15:48
juhp that is everything on the agenda Feb 17 15:48
rahulb thats is it, I will further helping ff team with these locales hereon Feb 17 15:48
juhp nice Feb 17 15:48
juhp * open discussion Feb 17 15:48
juhp so any other topics to bring up today before we close the meeting? Feb 17 15:49
juhp if not then thanks everyone for participating Feb 17 15:51
juhp we meet again here in two weeks time (at the time of the feature freeze) Feb 17 15:52
juhp # meeting closed Feb 17 15:52

Generated by 2.7 by Marius Gedminas - find it at!