05:01:43 #startmeeting i18n 05:01:43 Meeting started Thu Sep 2 05:01:43 2010 UTC. The chair is juhp. Information about MeetBot at http://wiki.debian.org/MeetBot. 05:01:43 Useful Commands: #action #agreed #halp #info #idea #link #topic. 05:01:55 #meetingname i18n 05:01:55 The meeting name has been set to 'i18n' 05:02:13 welcome to the weekly i18n meeting 05:02:16 hi 05:02:19 hi 05:02:37 hi 05:03:54 hi 05:04:15 hi 05:05:11 hi 05:05:37 ok 05:05:45 better get started :) 05:06:12 https://fedoraproject.org/wiki/I18N/Meetings/2010-09-02 05:06:16 is the agenda page 05:06:22 #topic F14 05:06:45 F14 beta is looking fairly good currently I think 05:06:52 overall I mean 05:07:20 change/devel freeze starts 14th 05:07:57 I updated the docsbeat a bit 05:08:05 https://fedoraproject.org/wiki/Documentation_I18n_Beat 05:09:05 Looks geat. thx. 05:09:19 after the beta freeze all updates for dist-f14 need releng approval I guess - it is also the deadline for version changes 05:09:42 hi 05:09:50 In two weeks there is the https://fedoraproject.org/wiki/Test_Day:2010-09-16_AnacondaTranslationKeyboard 05:10:01 which I kind of signed myself up for... 05:10:13 though I am probably not in the right geo... 05:10:42 ok 05:10:42 3 target bugs cc i18n-bugs currently 05:10:59 hi 05:11:32 hi 05:11:45 anything else for F14? 05:11:53 we are quite close to the beta 05:12:33 #topic splitting zh comps 05:12:48 I think we discussed this before so forgive me for tabling it again... 05:13:03 I am thinking again it would be better to split zh comps if possible 05:13:55 I think we kind of shot it down before, but I am thinking if anaconda could support multiple locales for langgroups it might still work 05:14:17 and it makes a lot of sense IMHO looking at the Chinese lang group 05:14:36 almost complicate package duplication 05:14:40 erm complete 05:14:57 what do you think? 05:15:17 I could file an anaconda rfe and see what happens if we think it is a good idea 05:15:36 ping epico_laptop, dychen_, jni ! 05:16:41 juhp, quite a lot of them are common. 05:17:00 like? 05:17:25 juhp, IMHO, how about three, chinese(common), chinese(simp), chinese(trad 05:18:05 doesn't work with current comps/anaconda 05:18:20 juhp, ibus, stardict, goldendict, are example for chinese(common). 05:18:29 dychen_: any common packages could go in both 05:18:39 maybe it can reduce packages needed to install. 05:18:49 dychen_: ibus? 05:18:56 not very sure about this. 05:19:03 ok stardict is not really zh specific 05:19:33 dychen_: it is ok to duplicate across groups where appropriate 05:19:55 juhp, then I have no further questions then. :-) 05:20:00 epico_laptop: right 05:20:20 for me the main problem is reducing duplication of langpacks 05:20:42 I guess most zh users only want SC or TC langpack not both 05:20:54 * epico_laptop think so. 05:21:14 so SC would be zh_CN only ? 05:21:29 yes. 05:21:34 all other zh_* TC? 05:21:58 zh_sg is SC? 05:22:03 so problem is having one group for all TC zh locales 05:22:07 epico_laptop: ah perhaps 05:22:41 epico_laptop, Yes. 05:22:42 ok let's see what anaconda team says - not sure they will like it 05:22:54 though actually might be yum too not sure 05:23:21 epico_laptop, But I quite doubt how many people actually use zh_sg/ zh_ml. 05:24:17 hmm yeah less common perhaps 05:24:23 okay thanks 05:24:33 I can't choose zh_SG in gdm greeting screen, assume it need manually specify it? 05:24:39 well no zh_ML in glibc anyway :) 05:24:54 epico_laptop: probably no transtions? 05:24:59 la 05:25:00 dunno 05:25:23 epico_laptop: gdm filters out locales without translations anyway 05:25:34 I see. 05:25:53 so yeah maybe zh_SG less of an issue 05:26:12 anyway there is still zh_TW and zh_HK for TC 05:26:38 though not sure how many projects have zh_HK... 05:26:46 anyway let's move on :) 05:27:14 #topic bug stats 05:28:36 rawhide: 42 bugs (+5); 19 NEW (+1) 05:28:37 F14: 29 bugs (-7); 15 NEW (-3) 05:28:51 rawhide up a bit and f14 down :) 05:29:54 let me update f13 and f12 presently 05:30:30 #topic IM 05:30:48 yesterday we had the Bugs Day 05:30:59 paragan: thanks for organizing 05:31:07 welcome 05:31:22 we got some bugs updated now :) 05:31:40 * juhp had hoped for a bit more participation but we reduced the bug count a bit and triaged quite a lot 05:31:46 yeah 05:32:16 hm ah maybe I should have moved into IM section: so we may do another later for fonts? 05:32:18 so remaining bugs I see few abrt and few feature request bugs. 05:32:31 aha right 05:32:39 (shouldn't) 05:32:41 I will check to request for detail input to reproduce abrt 05:32:46 bugs 05:32:51 ok 05:33:26 i think bugs days with testing will help may b 05:33:27 down to 36 bugs which is not too bad :) 05:33:28 be 05:33:33 yeah 05:33:48 maybe we can do the fonts one after beta released 05:33:52 so users can also contribute little more 05:34:02 aha 05:34:05 yeah, font is really required one 05:34:23 whenever i see some bugs for font, while checking for it i found so many other 05:34:29 not reported for long time 05:34:42 so pulling people for testing fonts will help lot to improve font quality 05:35:03 okay 05:35:16 yeah getter testers is the hard part 05:35:33 at least now bodhi encourages more testing before release :) 05:35:35 and having all expert same time can quickly decide which is fonts and which is rendering engine problem 05:35:43 right 05:36:28 any IM topics to discuss? 05:36:34 I have a couple actually 05:37:57 I have been looking a bit more at key compose under gtk (XIM, gtk-im-context-simple, ibus) but not come to conclusion yet 05:38:23 though currently might be leaning towards making XIM default again for most locales like we/I did for F13 05:39:05 though gnome Desktop team like gtk-im-context-simple 05:39:19 going to compare the rules to get a better picture 05:39:35 not sure what rules ibus uses 05:40:04 fujiwarat: tried asking before ^ 05:40:21 fujiwarat: but more importantly wanted to discuss about xkb support in ibus 1.3.7 05:40:33 which I am not quite getting :) 05:40:52 Currently I'm thinking migration of gtk-xim to an ibus engine. 05:41:22 fujiwarat: I think that only works if you make that engine the off state 05:42:14 Currently ibus engine off is same as gtk-im-context-simple. 05:42:15 though I am increasingly thinking that the MacOS IME toggling switching UI is the right one 05:42:28 fujiwarat: ah I see 05:43:00 fujiwarat: how about the xkb layout support in 1.3.7? 05:43:14 there are a lot of layouts now in the config 05:43:23 fujiwarat: is that from google? 05:44:11 Yes, it shows some layouts. 05:44:27 we can also discuss more in detail afterwards - just don't feel it is quite right yet perhaps 05:44:37 The default is: 05:44:39 XKB_PRELOAD_LAYOUTS=""\ 05:44:39 "us,us(chr),ad,al,am,ar,az,ba,bd,be,bg,br,bt,by,"\ 05:44:39 "de,dk,ca,ch,cn,cn(tib),cz,ee,epo,es,et,fi,fo,fr,"\ 05:44:39 "gb,ge,ge(dsb),ge(ru),ge(os),gh,gh(akan),gh(ewe),gh(fula),gh(ga),gh(hausa),"\ 05:44:39 "gn,gr,hu,hr,ie,ie(CloGaelach),il,"\ 05:44:41 "in,in(ben),in(guj),in(guru),in(jhelum),in(kan),in(mal),in(ori),in(tam),"\ 05:44:43 "in(tel),in(urd-phonetic),in(bolnagri),iq,iq(ku),ir,ir(ku),is,it,jp,"\ 05:44:45 some is an understatement :) 05:44:47 "kg,kh,kr,kz,la,latam,lk,lk(tam_unicode),lt,lv,ma,ma(tifinagh),mal,mao,"\ 05:44:49 "me,mk,mm,mt,mv,ng,ng(hausa),ng,ng(igbo),ng(yoruba),nl,no,no(smi),np,"\ 05:44:51 "pk,pl,pl(csb),pt,ro,rs,ru,ru(cv),ru(kom),ru(sah),ru(tt),ru(xal),"\ 05:44:53 "se,si,sk,sy,sy(ku),th,tj,tr,ua,uz,vn" 05:44:55 Sorry, long output. 05:45:00 thanks 05:45:15 only Latin layouts? 05:45:59 No, it includes ru, e.g. 05:46:06 aha 05:46:11 no kana say? 05:46:24 hi 05:46:32 Yes, the default doesn't include kana. 05:46:40 why? :) 05:46:57 trying to understand the criteria for listing :) 05:47:07 Ah, I have no specific reasons. 05:47:12 hm 05:47:49 Showing all layouts is too many so I just reduced some layouts by default. 05:47:54 if it was me I would almost do the opposite first - only list non-Latin layouts... 05:48:37 anyway looks good start :) 05:49:04 fujiwarat: also seems several places one can add layouts 05:50:03 ? sorry, where can add layouts? 05:50:14 well 2 maybe 05:50:24 fujiwarat, I see some button in ibus-setup "Configure Layouts" 05:50:26 I mean as IME and layout perhaps 05:50:43 I am not clear on the distinction 05:50:48 fujiwarat, does it also adds IME? 05:50:54 Are, one is to add layouts to default another is to delete layouts from default. 05:51:24 aha 05:51:49 can Configure Layouts button be renamed to Delete Layouts? 05:51:57 ok I'll play bit more with it and maybe have some suggestion 05:53:02 I mean the configure layouts has two button 'add' and 'delete'. 05:53:14 right 05:53:26 layout is different to IME? 05:53:34 oh 05:53:41 Yes, it's same. 05:53:56 No, it's same. 05:54:00 so why the button? 05:54:18 filter? 05:55:02 there is also System Layout 05:55:20 Yes, the default layouts list is included in engine lists. And the configure button customize the default layout list. 05:55:42 aha 05:56:07 maybe I need to play with it more - feel like that button is a bit superfluous 05:56:27 or perhaps I am missing the whole picture :) 05:56:51 fujiwarat: thanks 05:57:01 perhaps we should move on? 05:57:04 I agree. Any ideas are welcome. 05:57:10 ok cool 05:57:31 fujiwarat: so better to drop it? :) 05:57:48 or not list layouts in IME list? 05:58:05 hm 05:58:16 but then they can be ordered 05:58:19 OK, let me think it. 05:58:20 can't 05:58:26 sure 05:58:36 #topic Fonts 05:58:45 running short of time 05:59:02 any fonts issues today to discuss? 05:59:58 I wanted to discuss .bug 624158 06:00:16 . bug 624158 06:00:18 nkumar: Bug 624158 Idea for setting individual fallback language - https://bugzilla.redhat.com/show_bug.cgi?id=624158 06:00:46 sure 06:01:19 nkumar: so any more thoughts? 06:01:35 there is a file /etc/profile.d/lang.sh 06:02:08 if we patch it something like this: 06:02:10 if [ -n $LANG ]; then 06:02:10 case $LANG in 06:02:10 hi*) export LANGUAGE=hi_IN:bn_IN ;; 06:02:10 bn*) export LANGUAGE=bn_IN:bn:hi_IN ;; 06:02:10 esac 06:02:11 fi 06:02:22 tagoh: btw did we include mozc in comps yet? :) 06:02:52 nkumar: yeah that is going to be a big patch... 06:03:00 yep 06:03:11 juhp: no not yet. I was about to do that - not even in stable though 06:03:17 cool 06:03:20 ah 06:03:23 sure 06:03:37 because though it has a line: [ -n "$LANGUAGE" ] && export LANGUAGE || unset LANGUAGE 06:03:42 just occurred to me :) (I could test it:) 06:03:51 aha 06:04:01 it is never set 06:04:13 never? :) 06:04:29 well I agree better than just hardcoding for system loacle 06:04:31 actually I am not sure 06:04:31 locale 06:04:48 3 more solutions 06:04:54 Probably $HOME/.i18n could override the setting. 06:05:04 but they require changes in gettext or glibc 06:05:04 actually I forget if we still owner ~/.i18n properly 06:05:57 the other solutions? 06:06:12 change in dcigettext.c 06:06:23 nkumar: how about asking drepper for advice on the glibc side - I am sure it has probably come up before... 06:06:56 in function guess_category_value 06:06:57 yep 06:07:13 guess_category_value 06:07:26 aha 06:07:55 from our pov glibc would be best I feel but dunno if they consider it an acceptable change 06:08:08 if not only way seems to be to do it at the OS level 06:08:10 yep there was a bug which may be of relevence https://bugzilla.redhat.com/show_bug.cgi?id=372151 06:08:30 .bug 372151 06:08:32 juhp: Bug 372151 /etc/sysconfig/i18n locale settings not propagating into X session - https://bugzilla.redhat.com/show_bug.cgi?id=372151 06:09:07 I think that is a different story - but looks like good summary by jakub 06:09:48 ah ok it talks about LANGUAGE 06:09:56 yep 06:10:22 anyway 06:12:11 I think we are out of time unless any other matters need attention 06:12:43 This is taking time : https://fedorahosted.org/fedora-infrastructure/ticket/2360 06:13:22 nkumar: you could try asking #fedora-admin 06:13:29 not sure if it is possible 06:13:47 alternative would be add new project and delete old perhaps 06:14:07 ok, should I make a fresh request 06:14:17 or use github for now or something ;) 06:14:25 I dunno better to ask first 06:14:29 ok :) 06:15:07 ok then I will put it on github or gittorios 06:16:39 ok thank you for the meeting! 06:16:47 thanks 06:17:19 now wonder if I can close ;) 06:17:40 #endmeeting 06:18:50 #endmeeting