05:03:56 <tagoh> #startmeeting i18n
05:03:56 <zodbot> Meeting started Thu Dec  9 05:03:56 2010 UTC.  The chair is tagoh. Information about MeetBot at http://wiki.debian.org/MeetBot.
05:03:56 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
05:04:01 <tagoh> #meetingname i18n
05:04:01 <zodbot> The meeting name has been set to 'i18n'
05:04:05 <tagoh> #topic agenda
05:04:07 <juhp> hi
05:04:11 <tagoh> https://fedoraproject.org/wiki/I18N/Meetings/2010-12-09
05:04:23 <epico_laptop> hi
05:04:30 <tagoh> hi
05:04:36 <dueno> hi
05:04:46 <paragan> hi
05:05:24 <pravins> hi
05:05:41 <tagoh> f12 things was got out of the agenda finally :)
05:05:49 <tagoh> anyway, let's get started
05:06:14 <tagoh> #topic F15
05:06:52 <juhp> cool
05:08:19 <tagoh> as you may or may not got a mail about FTBFS issues:
05:08:37 <tagoh> http://lists.fedoraproject.org/pipermail/devel/2010-December/146682.html
05:08:39 <tagoh> http://lists.fedoraproject.org/pipermail/devel/2010-December/146683.html
05:08:55 <juhp> hm no target bugs
05:09:36 <tagoh> please check if your packages is there and encourage you to fix them.
05:10:10 <juhp> right
05:11:26 <tagoh> only one target bug?
05:11:40 <tagoh> erm, blocker
05:11:45 <juhp> rawhide seems to be installable again - though I was getting some bad hangs yesterday (X?)
05:11:55 <tagoh> aha
05:12:10 <juhp> even live spins
05:12:17 <juhp> (verb)
05:13:42 <juhp> tagoh: how about the console locale bug?
05:13:47 <juhp> might be a good target at least
05:13:49 <dychen> hi
05:13:56 <tagoh> no responses
05:14:00 <tagoh> yep
05:14:17 <juhp> is it systemd?
05:14:26 <tagoh> grubby
05:14:33 <juhp> ah yeah
05:15:16 <tagoh> maybe a good idea to have a look at rawhide bugs once and evaluate if it's worth escalating to target or blocker?
05:15:19 <juhp> added
05:15:25 <juhp> yep
05:16:21 <juhp> added my gettext bug too ;)
05:17:50 <tagoh> .bug 246325
05:17:52 <zodbot> tagoh: Bug 246325 need LANGUAGE settings for translation fallbacks - https://bugzilla.redhat.com/show_bug.cgi?id=246325
05:18:02 <tagoh> how is it? any progress on that?
05:18:15 * juhp just "renamed" it :)
05:18:37 <juhp> I think it is stalled
05:18:46 <juhp> maybe we should poke notting
05:19:22 <tagoh> ok
05:19:34 <tagoh> .bug 444888
05:19:36 <zodbot> tagoh: Bug 444888 RFE: switch to PackageKit - https://bugzilla.redhat.com/show_bug.cgi?id=444888
05:20:01 <tagoh> any chance to get this fixed?
05:21:18 <juhp> wonder if 246325 is a dup of 246325
05:21:27 <juhp> tagoh: oops
05:21:48 <juhp> tagoh: I was talking about 624158 above
05:22:25 <juhp> paragan: do you know anything about 246325?
05:22:44 <tagoh> juhp: aha. maybe
05:23:17 <pravins> juhp: hmm, it required lots of changes in s-c-l
05:23:54 <pravins> tagoh: i think in spare time Naveen can work on this
05:24:26 <paragan> I think nkumar  worked on this fallback issue
05:25:04 <juhp> paragan: yes I know
05:25:14 <juhp> I think iso-codes is wrong anyway...
05:25:45 <tagoh> hm, one more outstanding issue for s-c-l
05:25:48 <tagoh> .bug 493604
05:25:50 <zodbot> tagoh: Bug 493604 Tracker: System Configuration Tools Cleanup Project - https://bugzilla.redhat.com/show_bug.cgi?id=493604
05:25:58 <paragan> ok will check iso-codes
05:26:14 <tagoh> looks like it needs some loves to be improved.
05:27:21 <pravins> tagoh: yeah, this was created long time bac for improve system-config packages, some enhancements are done, but some not  doable for taking time [RFE]
05:27:31 <pravins> s/back
05:27:33 <tagoh> .bug 493912
05:27:35 <juhp> I think better to simplify it just to be a system locale changer tool
05:27:35 <zodbot> tagoh: Bug 493912 S-c-tools cleanup: port to PolicyKit - https://bugzilla.redhat.com/show_bug.cgi?id=493912
05:29:53 <juhp> and nkumar is not here...
05:30:24 <tagoh> .bug 659214
05:30:26 <zodbot> tagoh: Bug 659214 Don't have U+040D (and U+0400) needed for Bulgarian language - https://bugzilla.redhat.com/show_bug.cgi?id=659214
05:31:21 <tagoh> suppose it may be worth adding into the target?
05:31:38 <juhp> dychen: 654929 is fixed?
05:31:56 <juhp> tagoh: is there a patch?
05:32:50 <juhp> dychen: I closed it already
05:33:05 <tagoh> juhp: seems not
05:33:15 <juhp> so probably unlikely then?
05:33:23 <juhp> it is an old bug I think?
05:33:39 <tagoh> not really
05:33:40 <juhp> ah no
05:33:51 <tagoh> filed 1 week ago
05:33:59 <juhp> yeah sorry
05:34:21 <juhp> I think we need a contribution for that
05:34:42 <tagoh> right
05:35:02 <juhp> pravins: ?
05:35:41 <tagoh> .bug 654929
05:35:43 <zodbot> tagoh: Bug 654929 ibus-table not working with ibus-1.4 - https://bugzilla.redhat.com/show_bug.cgi?id=654929
05:35:52 <dychen> juhp, as dueno said, there's a fix in ibus-1.4 which address this problem.
05:36:34 <juhp> dychen: so please close...
05:37:02 <dychen> juhp, it is closed as rawhide.
05:37:25 <juhp> yeah I just closed it ;)
05:37:50 <pravins> juhp: sorry, was in other discussion
05:38:09 <nkumar> hi
05:39:46 <tagoh> nkumar: please check the logs later - that would be nice to work on several RFE bugs for s-c-l if you have a time.
05:40:34 <nkumar> tagoh: thanks, I will do that... :)
05:41:01 <tagoh> ok, anything else on f15?
05:43:26 <tagoh> shall we move on then?
05:43:55 <tagoh> #topic Package Updates
05:44:18 <pravins> zodbot: whoowns ibus-table
05:44:18 <zodbot> pravins: dchen (orphan in Fedora EPEL)
05:44:39 <tagoh> new ipa-ex-mincho-fonts are available in testing btw
05:44:54 <juhp> aha
05:45:02 <juhp> new release?
05:45:06 <tagoh> right
05:45:18 <pravins> dychen: i have some more queries for you on indic-tables
05:45:27 <pravins> :)
05:45:40 <dychen> pravins, yes?
05:46:18 <pravins> tagoh: i am working on ibus-indic-tables
05:46:37 <pravins> https://github.com/pravins/indic-table/blob/master/trans.txt
05:46:44 <tagoh> aha
05:46:59 <juhp> pravins: cool - we have an ibus section (maybe should be Input methods?) coming up
05:48:07 <tagoh> anything else packages in testing that want to pay some attentions?
05:48:12 <nkumar> I am working on statistical transliteration through moses & ibus
05:48:32 <nkumar> got some results after training moses...
05:48:34 <pravins> juhp: yes
05:48:52 <nkumar> will fine tune it today, expecting good results after fine tuning
05:49:15 <tagoh> otherwise let's move on next topic
05:49:32 <tagoh> which you guys may want to talk about more :)
05:50:18 <tagoh> #topic ibus
05:50:43 <tagoh> ok, any progress on input methods related?
05:51:09 <pravins> tagoh: yes, work is started
05:51:30 <pravins> i have created initial table at https://github.com/pravins/indic-table/
05:52:30 <tagoh> pravins: aha. are you gonna replace the existing indic table with it eventually?
05:52:53 <pravins> tagoh: yes, we can
05:53:10 <pravins> but first we must get satisfying results
05:53:52 <tagoh> so this is just out of curiosity, what's issues on current tables and what are you trying to improve on that?
05:54:15 <pravins> tagoh: current table's are not working
05:54:18 <pravins> i tried those
05:54:32 <pravins> and there only only few tables
05:54:33 <tagoh> oh, not working?
05:54:36 <pravins> yes
05:54:41 <pravins> i tried malayalam one
05:54:53 <pravins> but there are just created nothing in inside it
05:55:22 <tagoh> do we have a bug for that?
05:55:53 <tagoh> that sounds like a blocker, doesn't it?
05:55:54 <pravins> no, but there is not enough data as well in it
05:56:01 <pravins> and no one using it
05:56:10 <tagoh> heh
05:56:24 <pravins> :)
05:56:25 <juhp> to clarify these are completely new tables
05:56:42 <juhp> actually word tables
05:56:59 <juhp> so there are no current tables in fedora :)
05:57:15 <pravins> juhp: thanks :)
05:57:20 <juhp> (:
05:57:34 <nkumar> I guess these are like predictive text entry with all words listed and ibus implicitly doing frequency thing...
05:57:57 <nkumar> pravin: am i right?
05:57:58 <tagoh> so how does those who isn't well supported by the tables input their languages??
05:58:00 <juhp> ibus-table yes
05:58:07 <pravins> nkumar: yes
05:58:10 <juhp> tagoh: m17n
05:58:41 <juhp> these are new input methods basically
05:58:47 <pravins> indic user uses .mim inscript and itrans
05:59:54 <tagoh> juhp: ok, let me rephrase then. what's issues in m17n and what are you expecting to the indic table making it better? :)
06:00:19 <pravins> tagoh: :)
06:00:22 <juhp> tagoh: it is word based not character based :)
06:00:45 <tagoh> aha
06:00:51 <juhp> so actually our first real input-method for indic...
06:00:57 <pravins> tagoh: it will show you candidates (optional words) based on user input
06:01:44 <tagoh> I see
06:02:20 <pravins> juhp: tagoh dychen but i think there might be need some options in ibus-table, or may be i am not aware of some options :(
06:03:16 <pravins> it is accepting invalid key's as well, it should commit words, if next character pressed not forming valid key
06:03:39 <juhp> pravins: right wildcard option might be good - I recommend testing with scim-tables too and filing an rfe
06:04:10 <pravins> my table was not detected in scim, need to check again
06:04:34 <juhp> might also be worth comparing with m17n though not sure it will scale in size
06:04:48 <pravins> yeah, that can be major problem
06:05:25 <juhp> pravins: you used scim-make-table ?
06:06:23 <pravins> think need to paste those table in different place
06:06:34 <juhp> of course
06:06:48 <juhp> scim-tables has .bin format iirc
06:07:08 <pravins> juhp: thanks, got it i will try
06:07:18 <juhp> :)
06:07:28 <tagoh> pravins: are you planning to have some releases during f15 timeframe?
06:07:29 <juhp> I think scim-tables is more featureful
06:08:14 <pravins> yes, i can quickly, but i am not satisfied with some result
06:08:28 <pravins> so may be 1 week more, i can release mr_IN atleast
06:09:17 <tagoh> ok, just wonder if we should have some note in rel-note or in the feature page?
06:09:27 <tagoh> anyway
06:09:43 <pravins> tagoh: yes, it should come under feature page
06:09:47 <pravins> i will report one
06:10:30 <tagoh> ok
06:10:39 <pravins> can we add action item in this meeting? does zodbot do that?
06:11:01 <juhp> sure
06:11:05 <juhp> would be good
06:11:23 <tagoh> # action does
06:12:07 <tagoh> without a space after # of course
06:13:04 <tagoh> anyway, we are running over a time on this meeting. anything else?
06:13:08 <nkumar> my turn... ;)
06:13:23 <nkumar> Learned how to train Moses for statistical transliteration between any two languages
06:13:30 <nkumar> will fine tune it today, and expecting good results after fine tuning.
06:13:37 <nkumar> After that can package moses for fedora and combine the trained statistical model to run with ibus & moses engine
06:13:54 <dychen> pravins, I sense the need of End key support
06:14:20 <pravins> dychen: yes
06:14:29 <tagoh> nkumar: cool
06:14:29 <pravins> dychen: can you try to install it once
06:14:39 <pravins> so i can explain you exact problem
06:14:53 <dychen> pravins, the git?
06:15:42 <pravins> yes
06:17:29 <tagoh> hm, that may be good to postpone next topic..
06:17:59 <tagoh> ok, anything else before closing?
06:20:11 <tagoh> thanks everyone for the meeting. feel free to keep in the discussion at #fedora-i18n or anywhere else if you want.
06:20:21 <tagoh> #endmeeting