06:18:02 <tagoh_> #startmeeting i18n
06:18:02 <zodbot> Meeting started Thu Aug 30 06:18:02 2012 UTC.  The chair is tagoh_. Information about MeetBot at http://wiki.debian.org/MeetBot.
06:18:02 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
06:18:06 <tagoh_> #meetingname i18n
06:18:06 <zodbot> The meeting name has been set to 'i18n'
06:18:15 <tagoh_> #topic agenda and roll call
06:18:24 <paragan> hi
06:18:24 <tagoh_> https://fedoraproject.org/wiki/I18N/Meetings/2012-08-30
06:18:27 <mfabian> hi
06:18:29 <anish_> hi
06:18:37 <tagoh_> sorry for late start.
06:18:37 <juhp> hi
06:18:41 <epico> hi
06:18:56 <tagoh_> shall we have a meeting quickly
06:19:49 <dueno> hi
06:20:29 <tagoh_> okay, let's get started.
06:20:34 <tagoh_> #topic F18
06:21:42 <tagoh_> though the release schedule is slipped, no changes in the test day schedule so far.
06:22:04 <juhp> is it an optional to change the dates?
06:22:09 <juhp> option I mean
06:22:12 <tagoh_> so 4th September for i18n.
06:22:22 <juhp> or are later slots already taken?
06:23:00 <tagoh_> juhp: well, if we want to do.
06:23:48 <juhp> ok dunno just asking/exploring :)
06:24:38 <tagoh_> no problem. is there any issues that we can't do it next Tuesday say?
06:26:38 <juhp> main thing might be f18 stability still perhaps though it is improved considerably in last days
06:26:58 <tagoh_> sure
06:27:35 <tagoh_> https://fedoraproject.org/wiki/Test_Day:2012-09-04
06:28:55 <Alam> we need to use Live ISO, which must be ready to test atleast by Monday (I am not sure) to test
06:29:15 <Alam> there is all slot booked till Nov 1
06:29:23 <Alam> if need to change
06:29:38 <juhp> aha then I guess we can't really change the date easily
06:30:05 <tagoh_> Alam: no exception to have any extra slots?
06:30:14 <juhp> maybe we should go for a slightly later slot next time...
06:30:34 <Alam> yes, there can be, we can do on Tuesday
06:30:40 <juhp> yeah unless we could do it on another day of the week later say
06:31:21 <juhp> https://fedoraproject.org/wiki/QA/Fedora_18_test_days
06:31:44 <Alam> I will take care during f19 test day
06:32:03 <juhp> I see OpenStack is also on 4th
06:32:14 <tagoh_> hm, indeed the availability of the live image may be the point to go or no-go.
06:32:44 <juhp> aha
06:32:58 <juhp> dunno if 13th would be better for us?
06:33:23 <juhp> at least by then alpha would helpfully be pretty ready
06:33:47 <juhp> oh
06:34:05 <juhp> s/oh//
06:34:05 <tagoh_> 13th? is it for FreeIPA?
06:34:10 <juhp> yeah
06:34:29 <juhp> is it allowed?
06:34:46 <Alam> or 18?
06:34:59 <juhp> aha
06:35:33 <tagoh_> we can ask on the ticket I guess.
06:35:34 <juhp> competing with OpenStack or IPA seems similar but sure could also well be on our own separate day :)
06:35:39 <juhp> good idea
06:35:41 <Alam> alpha may widely available to test
06:36:29 <tagoh_> nod
06:36:38 <juhp> around alpha time would be ideal
06:37:45 <tagoh_> okay, 18 may be safe because there seems a lot of alpha blocker yet. let's try to ask for it on the ticket then.
06:37:58 <juhp> Alam, has live spinning been fixed yet, do you know?
06:39:06 <Alam> juhp: I had not tired test 3 live iso
06:39:51 <Alam> livecd-fedora-livecd-desktop-201208271518.iso was not working to run from live
06:40:03 <Alam> I used that
06:40:30 <juhp> Alam, where did that come from?
06:40:39 <Alam> http://dl.fedoraproject.org/pub/alt/anaconda/
06:40:43 <juhp> aha
06:40:51 <juhp> thanks
06:41:02 <juhp> Alam, I think there was no live for tc3 btw
06:41:06 <tagoh_> we can presume that alpha has any live images?
06:41:26 <Alam> yes, no live
06:41:33 <juhp> tagoh_, hopefully :)
06:41:52 <juhp> Alam, so livecd-fedora-livecd-desktop-201208271518.iso  doesn't boot or ?
06:41:59 <Alam> juhp: tagoh_: https://fedorahosted.org/fesco/ticket/941
06:42:30 <tagoh_> hmm
06:42:31 <Alam> juhp: it boot, but failed to install and crash during use of some application
06:43:20 <juhp> hmm "Assuming an extra week's slip from the current schedule (which is a given at this point)"
06:43:35 <juhp> Alam, thanks
06:43:40 * juhp dl's
06:44:03 * Alam is adding comment in ticket for Sep 18 request for i18n test day
06:44:16 <tagoh_> Alam: thanks!
06:45:01 <tagoh_> #agreed i18n test day to slip to 18th September
06:45:02 <tagoh_> anyway
06:45:34 <tagoh_> fujiwarat: do you want to update your feature?
06:45:48 <tagoh_> or anything more from anyone else on f18?
06:46:05 <juhp> wow that ticket re-enforces my view that anaconda and other such core packages should be frozen well before Alpha time
06:46:28 <paragan> I will try to check with desktop people to see if we can get new pango release pushed to testing early.
06:46:40 <fujiwarat> Probably nothing specially.
06:47:31 <juhp> paragan, great
06:48:07 <tagoh_> fujiwarat: any changes or improvements related to this feature on new release from GNOME upstream? do you know?
06:50:08 <fujiwarat> new release?
06:51:23 <fujiwarat> I think there are no big changes since people attended GUADEC.
06:51:27 <tagoh_> fujiwarat: 3.5.90 I meant.
06:55:04 <tagoh_> btw we have finished most feature implementation. that would be good time to concentrate bug fixes in next month to improve/polish the beta.
06:56:38 <fujiwarat> It seems libnm-gtk is added newly but no effect ibus, I think.
06:56:56 <tagoh_> fujiwarat: okay, thanks.
06:57:43 <tagoh_> hope we'll see more activities on bugs and reduce the amount of bugs we own by beta.
06:58:02 <tagoh_> anyway, let's move on if no further topics to discuss for f18?
06:58:46 <tagoh_> #topic Input Methods
06:59:22 <tagoh_> anyone has anything to share for IM?
07:01:01 <tagoh_> anish_: I may asked you this before but what was the reason why we can't make i-t-b default?
07:02:07 <pravins> tagoh_: i think, the mail reason will be testing and user feedback
07:02:16 <tagoh_> aha
07:02:20 * pravins thought he missed the meeting
07:02:25 <fujiwarat> new ibus rpm ibus-1.4.99.20120822-1 is integrated in f18 and f17 which includes some bug fixes.
07:02:29 <pravins> s/mail/main
07:02:29 <anish_> users are not willing to use it as default
07:02:50 <tagoh_> anish_: why? :)
07:03:10 <anish_> default means replacing itb with existing methods right
07:03:12 <pravins> tagoh_:  i think, we can take this as a hypothesis for upcoming testing event and see how it goes
07:03:19 <tagoh_> anish_: do we have any missing feature or any difficulty to do?
07:04:06 <tagoh_> pravins: sure. yeah, that would gives us a good opportunity to see the sort of this.
07:04:16 <juhp> I think tagoh_ is trying to probe on what it would take to make it default in the feature, if that is desirable etc?
07:04:17 <anish_> they say they typing speed is faster they do not want suggestions
07:04:26 <pravins> anish_: IMHO we should try bit hard to make it default, without having it default it will not make much impact
07:05:03 <pravins> anish_: we have solution for that, i.e. disable suggestion option :)
07:05:19 <tagoh_> anish_: aha
07:05:38 <juhp> so only show suggestions with tab completion?
07:05:40 <anish_> then what is the difference between existing input method and i-t-b
07:05:54 <juhp> say
07:06:02 <pravins> it should be optional
07:06:12 <juhp> aha
07:06:31 <juhp> pravins, so you mean just pretend it is ibus-m17n like? :)
07:06:46 <juhp> by default
07:06:52 <pravins> juhp: yes, if user do not want suggestion, he can simply disable it from check box.
07:06:58 <juhp> sure
07:06:59 <anish_> e.g if your english typing speed is good then you do not use i-t-b to type english
07:07:05 <pravins> backward compatibility :)
07:07:24 <juhp> currently there is no checkbox, right?
07:07:47 <tagoh_> anish_: how about for Indic say?
07:08:04 <juhp> I think I would prefer using Tab say to show suggestions/completions
07:08:14 <anish_> actually i got this feed back from Indic
07:08:33 <anish_> they say it is good input method
07:08:52 <anish_> it is really useful for new users
07:09:11 <tagoh_> to input English? I meant to input Indic on i-t-b.
07:09:30 <anish_> input Indic on i-t-b
07:09:39 <tagoh_> sure
07:09:49 <anish_> they suggested me to install it as default package
07:09:52 <tagoh_> so they think ibus-m17n is better than i-t-b so far?
07:10:14 <anish_> yeah
07:10:41 <tagoh_> and due to the slow suggestion? anything else?
07:10:59 <anish_> nope only suggestions
07:12:06 <juhp> anish_, pravins: what do you think about tab completion?
07:12:22 <pravins> i feel experienced l10n users can disable suggestions from i-t-b, but making it default will make news users life happy and motivate them to type more in indian language
07:12:39 <juhp> seems to me turning off suggestions completely seems too extreme
07:12:51 <pravins> juhp: since, yet we do not have powerful dictionary in the background, i dont think tab completion will give good result
07:13:02 <tagoh_> hmm, I don't know about what's better condition to enable or disable the suggestion feature by default but may be not that bad idea to have any option to turn it on/off.
07:13:15 <juhp> pravins, but I mean it will be less intrusive that full suggestions mode
07:13:21 <juhp> than
07:13:39 <anish_> juhp, i do not get exact tab completion means?
07:13:42 <juhp> so better for exactly that reason
07:13:54 <tagoh_> particularly if it's the case people don't want to make it default.
07:14:07 <juhp> anish_, it means no candidates shown until requested
07:14:34 <juhp> hence my suggestion to default to tab completion instead of current way
07:15:10 <anish_> so it will be same as that of ibus-m17n
07:15:11 <juhp> if I knew I could complete longer words I might be interested in using it :)
07:15:19 <pravins> but juhp it will increase number of key strokes, like for using suggestion list i need to first press tab, then corresponding word in suggestion list
07:15:28 <juhp> anish_, erm no - m17n doesn't show any suggested words
07:15:42 <juhp> pravins, sure
07:15:59 <juhp> mostly I don't need to see suggestions all the time...
07:16:05 <anish_> then that is the wrong with showing words every time
07:16:10 <juhp> maybe with vkbd it will be different
07:16:18 <juhp> anish_, it is annoying :)
07:16:33 <pravins> i feel disable suggestion option is best for people familiar or expert typing with particular layout
07:16:39 <juhp> anyway an idea
07:16:44 <anish_> nod , in some cases
07:16:47 <pravins> even ankit_ once recommended it
07:17:02 <juhp> pravins, then no point to switch to i-t-b :)
07:17:03 <anish_> may be we need to integrate google suggestions as well
07:17:15 <tagoh_> anyway, what I wanted to know is if we have any plans/possibility to make i-t-b default in the future... otherwise we may be better to spend a time to improve ibus-m17n as default engine for Indic then.
07:17:35 <anish_> its default as guess
07:17:59 <pravins> juhp: yes agree, but that option will be only used by users not interested in getting suggestion, we will enable suggestion by default.
07:18:11 <juhp> hmm
07:18:23 <juhp> I still feel Tab would be a good compromise
07:18:49 <juhp> in theory...
07:19:35 <tagoh_> juhp: providing any trigger key to enable the feature sounds good to me if it affects to the performance etc.
07:20:10 <juhp> aha
07:20:16 <juhp> good idea
07:20:29 <juhp> so how about a mode switch as tagoh_ suggested?
07:20:46 <anish_> good idea
07:20:54 <pravins> yes, looks good to me. i.e switch from booster to m17n?
07:21:11 <tagoh_> does i-t-b use m17n?
07:21:16 <juhp> well switch between suggestions and no suggestions
07:21:18 <anish_> yes
07:21:28 <tagoh_> aha
07:21:38 <anish_> i-t-b uses m17n db
07:21:49 <tagoh_> and suggestion feature is came from hunspell? is it only case i-t-b uses hunspell?
07:21:50 <juhp> right - not ibus-m17n I guess
07:22:21 <juhp> dueno, any thoughts on this btw?
07:22:27 <anish_> yeah suggestions come from hunspell dictionary
07:22:36 <tagoh_> anish_: I see
07:23:46 <anish_> juhp, right not ibus-m17n
07:25:38 <tagoh_> okay, we can try to improve i-t-b in f18+ time frame and make it default possibly then.
07:26:08 <juhp> yeah perhaps targetting f19 is more realistic?
07:26:19 <anish_> making it default means to replace i-t-b with existing input methods
07:26:22 <anish_> right?
07:26:31 <pravins> juhp: tagoh_ yes, anish_ lets try to get more feedback during f18 beta i18n test days
07:26:38 <juhp> ok
07:26:46 <tagoh_> anish_: ibus-m17n so far for Indic yes.
07:26:58 <juhp> anish_, well replacing m17n with i-t-b
07:27:17 <pravins> anish_: it will be not replacement, but will be become default option, users can still use old input method by manually selection
07:27:47 <juhp> hmm
07:27:56 <anish_> okay thanks
07:28:00 <pravins> we can not replace m17n as i-t-b uses it in background through littranslit
07:28:10 <pravins> libtranslit
07:28:10 <juhp> well that might be ok for f18 but we don't need duplication in long term
07:28:29 <juhp> pravins, heh i mean ibus-m17n usage
07:28:52 <pravins> juhp: ohh, i misunderstood :) yes got it
07:29:09 <tagoh_> juhp: upstream is different so that would means we stop commitment to ibus-m17n afterwards.
07:29:22 <juhp> for Indic yes
07:29:52 <tagoh_> but we can't prevent users to use ibus-m17n if they prefer IMHO
07:29:58 <juhp> sure
07:30:25 <juhp> s/sure/right/
07:30:59 <juhp> so we will try for i-t-b default for Beta?
07:31:28 <juhp> with suggestions defaulting to on/off?
07:31:31 <anish_> tagoh_, yeah we cant prevent users __not__ to use ibus-m17
07:31:40 <juhp> heh
07:31:58 <tagoh_> juhp: sounds good to me.
07:32:06 <juhp> but not installing it would help to do that...
07:32:30 <juhp> not that I have anything against ibus-m17n personally - I think it is still cool :)
07:33:20 <anish_> any anyway we use m17n db :)
07:33:26 <juhp> is it more accurate to say that i-t-b is using libtranslit?
07:33:41 <pravins> tagoh_: looks good to me as well, but your switch mode suggestion looks very cool to me
07:33:46 <anish_> juhp,  yes
07:34:09 <tagoh_> pravins: it's too late to implement in f18. so we should postpone it to f18+ anyway.
07:34:17 <pravins> tagoh_: yes
07:34:23 <juhp> I still think it would be nice to try Tab completion too but anyway
07:34:24 <pravins> anish_: whats you planning then?
07:34:45 <anish_> we will make default in f18+
07:34:53 <juhp> tagoh_, okay then target should be f19
07:35:08 <tagoh_> sure.
07:35:34 <juhp> personally I also feel that is more realistic
07:35:35 <pravins> ok, but we should make sure, we should get more and more feedback to be at better stage for f19
07:35:43 <juhp> true
07:35:59 <tagoh_> right
07:36:06 <anish_> nod
07:36:11 <pravins> either way we will never able to understand weakness of our concept
07:36:52 <tagoh_> #agreed ibus-typing-booster to make default in f19
07:36:58 <tagoh_> okay, anything more?
07:37:31 <juhp> it would be nice to test i-t-b in the test day anyway
07:37:32 <tagoh_> if not, move on.
07:37:43 <tagoh_> yeah
07:37:57 <juhp> do we have any testcases for it?
07:38:44 <tagoh_> I think it's in QA:English-typing-booster
07:38:48 <anish_> yes
07:39:04 <juhp> cool
07:39:04 <anish_> they are in for English typing booster
07:39:04 <tagoh_> the page should be renamed perhaps
07:39:08 <juhp> aha
07:39:15 <juhp> true
07:39:15 <tagoh_> but anyway
07:39:30 <tagoh_> move on then.
07:39:33 <tagoh_> #topic Fonts and Rendering
07:39:40 <tagoh_> anything for fonts and rendering?
07:40:35 <pravins> i was facing one issues of paktype fonts, i.e. upstream removed all the separate tarball of different family and clubbed it together
07:40:47 <tagoh_> aha
07:41:05 <tagoh_> what does upstream say about it?
07:41:18 <pravins> I am able to convince upstream and he releases all with old structure :)
07:41:34 <pravins> about to build it for fedora
07:41:41 <tagoh_> okay
07:42:33 <pravins> pango is built with harfbuzz but still not able to test it :(
07:42:43 <pravins> my F18 virtual machine has some problem
07:42:49 <tagoh_> sure
07:43:44 <pravins> so will test and update in next meeting
07:43:53 <tagoh_> okay, thanks
07:44:18 <pravins> Alam: did you able to test new pango build with harfbuzz in F18?
07:44:52 <Alam> pravins: no, latest pango avaialble in updates-testing is not with harfbuzz
07:45:43 <tagoh_> hopefully we'll get new pango for testing shortly
07:45:51 <pravins> yes
07:45:55 * Alam is waiting for it
07:46:18 <tagoh_> shall we move on?
07:46:28 <pravins> yes
07:46:31 <tagoh_> #topic Open Floor
07:46:51 <tagoh_> anything topics we are missing in the agenda if anyone has?
07:47:21 <tagoh_> otherwise let's close the meeting shortly
07:47:35 <mfabian> Why is there no en_GB in the English typing booster?
07:47:35 <mfabian> 
07:48:35 <pravins> mfabian: you should have ibus-hunspell-table and hunspell dictionary for en_GB install on you machine
07:48:41 <pravins> then you will get that option
07:48:55 <mfabian> hunspell seems to have it but is is missing in $ ls /usr/share/ibus-hunspell-table/hunspell-tables/en_G*.conf
07:48:56 <mfabian> /usr/share/ibus-hunspell-table/hunspell-tables/en_GH.conf
07:48:56 <mfabian> mfabian@ari:~
07:48:56 <mfabian> 
07:49:51 <tagoh_> maybe better to file a bug then? :)
07:49:56 <mfabian> OK.
07:49:58 <juhp> anish_, I think it would be good to turn on many more languages and then fix them if needed
07:51:05 <tagoh_> anything more?
07:52:17 <tagoh_> okay, stop here then. thanks everyone for the meeting!
07:52:23 <tagoh_> #endmeeting