I18N/InputMethods/Meetings/2009-09-09

previous meeting | next meeting

Agenda

Time:2009-09-09 0500 UTC

Place: #fedora-i18n at Freenode

• ja OpenWnn
• I18N/InputMethods/Bugs
• indic phonetic dictionary based IMEs
• make IMEs in @input-methods optional instead

Summary

• bug review

Log

juhp	I18N/InputMethods/Meetings/2009-09-09	Sep 09 15:09
juhp	ping phuang, tagoh3, paragan, pravins, kaio	Sep 09 15:10
juhp	shall we have the IM meeting	Sep 09 15:10
phuang	hi	Sep 09 15:10
pravins	juhp: hi	Sep 09 15:10
paragan	juhp, pong	Sep 09 15:10
fujiwarat	hi	Sep 09 15:10
*pravins also have one topic today :)		Sep 09 15:11
dychen	juhp: hi	Sep 09 15:11
juhp	# ja OpenWnn	Sep 09 15:12
Kaio	hi	Sep 09 15:12
tagoh3	hi	Sep 09 15:12
juhp	well just thought I would mention that android.git seems to include openwnn	Sep 09 15:13
juhp	but realised it is java of course ;)	Sep 09 15:13
juhp	some maybe not terribly useful for us	Sep 09 15:13
juhp	unless someone wants to write a java layer for ibus perhaps ;)	Sep 09 15:13
juhp	originally though it might be interesting to package for fedora...	Sep 09 15:14
juhp	might still be worth looking that the dictionary though?	Sep 09 15:14
phuang	what's is openwnn?	Sep 09 15:16
juhp	phuang: ah Wnn is a japanese IM backend	Sep 09 15:16
juhp	quite old	Sep 09 15:17
fujiwarat	and a commercial product ;)	Sep 09 15:17
juhp	(originally for Unix I guess)	Sep 09 15:17
juhp	fujiwarat: right	Sep 09 15:17
phuang	thanks	Sep 09 15:17
juhp	anyway think we can move on	Sep 09 15:17
juhp	that was mostly a headsup	Sep 09 15:17
---asgeirf__ is now known as asgeirf		Sep 09 15:18
juhp	will have a look over the src files later...	Sep 09 15:20
juhp	http://android.git.kernel.org/?p=platform/packages/inputmethods/OpenWnn.git	Sep 09 15:20
dychen	juhp: Should we:	Sep 09 15:21
dychen	1) make ibus-java	Sep 09 15:21
dychen	2) make ibus-openwnn	Sep 09 15:22
dychen	3) port ibus to android	Sep 09 15:22
dychen	?	Sep 09 15:22
fujiwarat	Sounds good with me. It might be good to think about ibus-wnn and ibus-atok.	Sep 09 15:22
phuang	We could do some investigation at first	Sep 09 15:23
juhp	dychen: I think ibus-java would be a complete rewrite :-/	Sep 09 15:23
dychen	juhp: Not sure how far can JNI go.	Sep 09 15:24
juhp	dychen: jni works with android?	Sep 09 15:27
dychen	juhp: Don't know.	Sep 09 15:28
dychen	juhp: It depends whether/when we want to support android.	Sep 09 15:29
juhp	dychen: it is an upstream issue	Sep 09 15:29
juhp	at least until fedora should support android...	Sep 09 15:29
juhp	no resources for that currently anyway	Sep 09 15:30
juhp	let's come back to bugs at the end	Sep 09 15:30
dychen	juhp: sure. :-)	Sep 09 15:30
juhp	# indic phonetic dictionary based IMEs	Sep 09 15:32
juhp	again more of a headsup	Sep 09 15:32
pravins	juhp: yes	Sep 09 15:32
juhp	but I think pravins has a nice idea of IME phonetic IMEs with dictionaries	Sep 09 15:32
pravins	this IME will give permutation and combination's of user i/p in candidate windows	Sep 09 15:33
pravins	but need to developed soe intelligence based combination, so user should find his accepted string in candidate window	Sep 09 15:34
pravins	s/soe/some	Sep 09 15:34
pravins	might be need some backend dictionary for that	Sep 09 15:34
dychen	pravins: what's i/p ?	Sep 09 15:34
dychen	input?	Sep 09 15:34
pravins	yeah input :)	Sep 09 15:35
pravins	for example say i/p is 'kitab'	Sep 09 15:35
pravins	so candidate windows will show	Sep 09 15:35
dychen	pravins: Sounds quite similar with ibus-pinyin and ibus-chewing.	Sep 09 15:36
pravins	kitab, kiitab, kitaab, kittab i.e	Sep 09 15:36
pravins	dychen: yes	Sep 09 15:36
dychen	pravins: fuzzy match?	Sep 09 15:36
juhp	sure	Sep 09 15:36
pravins	yes	Sep 09 15:36
juhp	it will be first proper indic IME I think	Sep 09 15:37
pravins	yes	Sep 09 15:37
juhp	ok move on?	Sep 09 15:37
pravins	juhp: yes	Sep 09 15:37
juhp	# make IMEs in @input-methods optional instead	Sep 09 15:37
juhp	okay this is actually about fedora ;o)	Sep 09 15:37
dychen	pravins: sounds great.	Sep 09 15:38
juhp	having discussed a bit about @input-methods with clumens in the anaconda rfe...	Sep 09 15:38
pravins	dychen: will talk more once i will start development	Sep 09 15:38
tagoh3	juhp: does it affects only people who use IM on English locale say?	Sep 09 15:38
juhp	I get a new idea - which I think turns out to "copy" ubuntu	Sep 09 15:38
dychen	pravins: Sure.	Sep 09 15:38
juhp	ie just install ibus and immodule but no IMEs by default	Sep 09 15:39
juhp	tagoh3: right it only affects non-CIJK installs say	Sep 09 15:39
tagoh3	ok	Sep 09 15:39
dychen	juhp: I do think M$ at this point do a bit better job. :-P	Sep 09 15:39
juhp	dychen: ?	Sep 09 15:40
juhp	then "English" who wants to input japanese say needs to "yum install ibus-anthy"	Sep 09 15:41
dychen	They have some thing like "East Asian" and "Complex support (which covers Thai and Indic),"	Sep 09 15:41
juhp	and it should justwork	Sep 09 15:41
dychen	juhp: ibus-anthy will/should pull in ibus and ibus-libs anyway. :-)	Sep 09 15:41
dychen	Mmm, but not ibus-gtk and ibus-qt.	Sep 09 15:42
juhp	right	Sep 09 15:42
juhp	since they will already be installed ;)	Sep 09 15:43
juhp	dychen: aren't they deps? ;)	Sep 09 15:43
dychen	juhp, IMHO, I am not sure how the *purist* will say about this.	Sep 09 15:44
*dychen thinks *purists* wants to get rid of ibus completely.		Sep 09 15:44
juhp	dychen: anyway the whole point being to avoid having to install the immodules afterwards :)	Sep 09 15:44
juhp	dychen: well at least we get rid of the IMEs	Sep 09 15:44
juhp	which are the biggest part by far	Sep 09 15:44
dychen	juhp: But, like we said before, what's the point to install ibus without IM?	Sep 09 15:45
juhp	and most users only want one at most	Sep 09 15:45
juhp	so that immodules work	Sep 09 15:45
juhp	until that is fixed...	Sep 09 15:45
juhp	I am saying it is a step in the right direction, not perfect :)	Sep 09 15:45
juhp	so what do people think of actually doing this for f12? :)	Sep 09 15:46
juhp	not sure about Live though	Sep 09 15:46
tagoh3	well, just one thing - don't they who use English locale and who want to input some languages, do yum groupinstall any languages first? and IMEs are pulled in with it right?	Sep 09 15:47
juhp	dychen: there are still edge cases of course I know	Sep 09 15:47
dychen	juhp: At some point during the installation, we need to ask which input method they want anyway.	Sep 09 15:47
juhp	tagoh3: why groupinstall ?	Sep 09 15:47
juhp	well that was our traditional approach	Sep 09 15:47
tagoh3	juhp: because no idea how to use that language :)	Sep 09 15:47
juhp	tagoh3: but I think lot of users don't want langpacks say	Sep 09 15:47
juhp	ah no idea which ime	Sep 09 15:48
juhp	true	Sep 09 15:48
juhp	tagoh3: anyway, and? :)	Sep 09 15:48
juhp	see we add conditional immodules to every lang group?	Sep 09 15:48
juhp	so	Sep 09 15:49
tagoh3	juhp: in other words, we don't need to install ibus too by default for them, no?	Sep 09 15:49
juhp	immodules?	Sep 09 15:49
tagoh3	could be conditional with the desktop and ibus?	Sep 09 15:49
juhp	or go the old ubuntu way and use xim?	Sep 09 15:49
juhp	tagoh3: ?	Sep 09 15:49
dychen	juhp: Or "Input method solution"? :-P	Sep 09 15:49
juhp	both	Sep 09 15:49
juhp	?	Sep 09 15:49
tagoh3	put a conditional with ibus in each desktop section?	Sep 09 15:50
juhp	tagoh3: it is already there	Sep 09 15:50
tagoh3	doesn't it work then?	Sep 09 15:50
juhp	how does it help with lang support?	Sep 09 15:50
juhp	tagoh3: at install time?	Sep 09 15:50
juhp	dychen: ?	Sep 09 15:51
juhp	in anaconda you mean?	Sep 09 15:51
tagoh3	hmm, yum doesn't manage which groups are now installed and evaluate conditions with it?	Sep 09 15:51
juhp	tagoh3: indeed not	Sep 09 15:51
tagoh3	aha	Sep 09 15:51
dychen	juhp: That was your original suggestion.	Sep 09 15:51
juhp	dychen: yeah but getting push back ;)	Sep 09 15:51
juhp	so not sure if it will happen	Sep 09 15:52
dychen	juhp: You will get pushed back, no matter what. ^.^B	Sep 09 15:52
juhp	we could also have meta-packages for ibus but not terribly keen	Sep 09 15:52
juhp	hehe	Sep 09 15:52
*dychen thinks he pulled it back. :-)		Sep 09 15:53
juhp	I think installing just ibus by default is reasonable compromise	Sep 09 15:53
juhp	phuang: would be even better if ibus could see newly installed IMEs of course ;)	Sep 09 15:54
phuang	juhp, yeah. but many details need considering	Sep 09 15:54
juhp	I know it is a bit hard	Sep 09 15:55
phuang	juhp, including update of of ime, install of m17n sub packages	Sep 09 15:55
juhp	on the other hand one of the reasons we wanted for moving to ibus imho	Sep 09 15:55
phuang	We need to observe all related files	Sep 09 15:55
juhp	phuang: at least for new IME would be nice	Sep 09 15:55
dychen	And for server only installation, ibus but neither ibus-gtk nor ibus-qt	Sep 09 15:56
phuang	how to handle m17n sub packages	Sep 09 15:56
phuang	if user installs some m17n maps	Sep 09 15:56
juhp	dychen: yeah - not like ibus is the only case there fedora	Sep 09 15:56
juhp	in fedora	Sep 09 15:56
juhp	phuang: I can forgive m17n for now :)	Sep 09 15:57
phuang	The most easiest way it kill ibus-daemon in post script of IME's rpm	Sep 09 15:58
juhp	dychen: better than current ibus-* anyway ;)	Sep 09 15:58
juhp	phuang: is it safe?	Sep 09 15:58
phuang	and then imsetting will restart it,	Sep 09 15:58
tagoh3	phuang: implement any methods to reload it with a trigger outside ibus and kick it from the package?	Sep 09 15:58
dychen	juhp: have't test when you are inputting something.	Sep 09 15:58
juhp	killing doesn't sound right to me	Sep 09 15:59
juhp	hehe	Sep 09 15:59
phuang	Or send other signal to ibus-daemon, ibus-daemon could restart by self	Sep 09 15:59
juhp	phuang: reload?	Sep 09 15:59
juhp	right	Sep 09 15:59
phuang	no reload, restart	Sep 09 15:59
juhp	might be better	Sep 09 15:59
juhp	but frankly we should aim for ibus to do it right not some hack	Sep 09 15:59
phuang	restart can resolve all issues,	Sep 09 16:00
juhp	hmm ok	Sep 09 16:00
juhp	phuang: but it is system resetting user's daemon	Sep 09 16:00
juhp	but suppose it might be good enough	Sep 09 16:00
juhp	for now	Sep 09 16:01
phuang	reload still has some problems, including how to handle the current activated input contexts	Sep 09 16:01
tagoh3	how to restart from the package's postinst, which isn't running under the expected user's role?	Sep 09 16:01
dychen	Or yet another systray icon says "new IME available, restart ibus to enable it." :-)	Sep 09 16:01
juhp	tagoh3: killall ;o)	Sep 09 16:01
tagoh3	heh	Sep 09 16:01
tagoh3	it's not "restart" nor "reload", you said :)	Sep 09 16:01
juhp	dychen: if it can do it that should be able to just add the IME to its lists	Sep 09 16:02
juhp	wonder if I should propose my @input-methods idea to f-i-l and f-d-l	Sep 09 16:03
juhp	nearly out of time here...	Sep 09 16:03
dychen	juhp: the applet can be active by post install script, but add into IME list, like phuang said, has many issues.	Sep 09 16:03
juhp	any bugs to discuss?	Sep 09 16:03
juhp	dychen: true	Sep 09 16:03
juhp	well I am not asking for a solution tomorrow but agreed that it is something we should tackle	Sep 09 16:04
tagoh3	dychen: again, how to communicate with ibus from postinst? how to know an abstract adress of dbus and ibus?	Sep 09 16:04
juhp	I don't accept that killall is a acceptable solution anyway	Sep 09 16:04
juhp	s/agreed/agreement/	Sep 09 16:05
phuang	tagoh3, may kill with a signal	Sep 09 16:05
juhp	phuang: ok that might be better	Sep 09 16:05
dychen	tagoh3: No need, it simply shows that users need to restart ibus themselves.	Sep 09 16:06
juhp	killall -USR1 or something	Sep 09 16:06
juhp	dychen: I don't think it is a dream solution	Sep 09 16:07
dychen	juhp: Agreed. It's actually dirty hack. :-)	Sep 09 16:07
juhp	perhaps ibus might even be able to wait for user input to complete??	Sep 09 16:07
phuang	Or another way, is add a button [restart ibus for new ime] in ibus-setup	Sep 09 16:07
juhp	dychen: haha	Sep 09 16:07
-->krishnababu (n=kkrothap@nat/redhat-us/x-uolqhxxlqrnhmdct) has joined #fedora-i18n		Sep 09 16:08
juhp	phuang: to what extent can the real problem be tackled?	Sep 09 16:08
phuang	If users do not find the IME they want, just need click the icon	Sep 09 16:08
phuang	s/icon/button/	Sep 09 16:08
juhp	phuang: but it is just restart?	Sep 09 16:09
juhp	maybe tooltip could mention it?	Sep 09 16:09
phuang	juhp, yeah. or just put some label 'If you can not find newly installed ime, please restart ibus by ....'	Sep 09 16:09
juhp	phuang: I think IMEs should be able to update themselves too ideally	Sep 09 16:10
tagoh3	just put updates with a restart flag. give up to think for rawhide. doesn't it sound good, eh? :)	Sep 09 16:10
juhp	tagoh3: hmm that is desktop restart	Sep 09 16:11
phuang	tagoh3, yeah. we could use restart flags too	Sep 09 16:11
juhp	back to scim	Sep 09 16:11
phuang	:)	Sep 09 16:11
juhp	no thanks	Sep 09 16:11
juhp	but for now we may have to...	Sep 09 16:11
tagoh3	juhp: we used to restart desktops to change IM! don't see dreams! :p	Sep 09 16:12
juhp	seriously this is something that ibus should do right at least by 2.0 say	Sep 09 16:12
phuang	yeah	Sep 09 16:12
juhp	ok.....	Sep 09 16:12
juhp	long discussion for futurefeature.... sigh	Sep 09 16:13
juhp	any bugs to discuss?	Sep 09 16:13
juhp	dychen: how about bug 509330?	Sep 09 16:14
juhp	.bug 509330	Sep 09 16:14
zodbot	juhp: Bug 509330 Press [,] or [.] keyboard have no function to page down or up. - https://bugzilla.redhat.com/show_bug.cgi?id=509330	Sep 09 16:14
juhp	phuang: should we open another bug for	Sep 09 16:15
juhp	.bug 501214	Sep 09 16:15
zodbot	juhp: Bug 501214 need better logo and panel icons: light blue ibus icon is not obvious - https://bugzilla.redhat.com/show_bug.cgi?id=501214	Sep 09 16:15
juhp	i mean for the logo?	Sep 09 16:15
phuang	yeah	Sep 09 16:15
Kaio	like the logo on the rawhide	Sep 09 16:16
juhp	ah but see update	Sep 09 16:16
juhp	maybe it is ok	Sep 09 16:16
juhp	kaio: you mean icon?	Sep 09 16:16
dychen	juhp: it's actually ibus-pinyin. phuang, do you really want me to mock with it? :-)	Sep 09 16:16
Kaio	juhp: trayicon..	Sep 09 16:16
juhp	ah	Sep 09 16:16
juhp	kaio: yeah we are now asking for better project logo	Sep 09 16:17
Kaio	as marketing?	Sep 09 16:17
phuang	dychen, ?	Sep 09 16:17
dychen	juhp: Actually one community guy just submit an ibus logo?	Sep 09 16:17
dychen	phuang: for bug 509330.	Sep 09 16:18
juhp	dychen: fixed	Sep 09 16:18
juhp	(assignment)	Sep 09 16:18
juhp	sigh	Sep 09 16:18
juhp	that bug has been waiting long time	Sep 09 16:18
juhp	dychen: next place reassign	Sep 09 16:18
juhp	please	Sep 09 16:18
dychen	juhp: Sure.	Sep 09 16:19
juhp	kaio: when will you finish ibus-table-setup gconf?	Sep 09 16:20
dychen	Though myself doesn't mind to "light phuang's load". :-)	Sep 09 16:20
juhp	.bug 513901	Sep 09 16:20
zodbot	juhp: Bug 513901 ibus-table setup does not store config settings - https://bugzilla.redhat.com/show_bug.cgi?id=513901	Sep 09 16:20
phuang	I will check it in rawhide, but it does not happen on my f11 box	Sep 09 16:20
juhp	ok	Sep 09 16:20
juhp	dychen: need to keep bugs moving anyway :)	Sep 09 16:20
phuang	f11 with ibus-1.2.x	Sep 09 16:20
juhp	(not sure why I assigned to dchen anyway :)	Sep 09 16:21
juhp	oh	Sep 09 16:21
juhp	phuang: maybe ask?	Sep 09 16:21
juhp	.bug 521591	Sep 09 16:22
zodbot	juhp: Bug 521591 ibus: check if the icon filename is a real file before trying to open it - https://bugzilla.redhat.com/show_bug.cgi?id=521591	Sep 09 16:22
juhp	phuang: is the patch ok?	Sep 09 16:22
juhp	.bug 509518	Sep 09 16:23
zodbot	juhp: Bug 509518 ibus-anthy should only override to jp layout for kana input - https://bugzilla.redhat.com/show_bug.cgi?id=509518	Sep 09 16:23
phuang	juhp, does not the patch	Sep 09 16:24
fujiwarat	I guess it's not for f12.	Sep 09 16:25
phuang	fujiwarat, yeah	Sep 09 16:25
juhp	ok	Sep 09 16:26
juhp	f11?	Sep 09 16:26
juhp	I think we have to fix 509518 for f12	Sep 09 16:26
juhp	if no comments I will make it a F12Blocker	Sep 09 16:26
phuang	sure	Sep 09 16:26
juhp	since it is a regression	Sep 09 16:26
juhp	ok	Sep 09 16:27
fujiwarat	Oh, I thought it's f13..	Sep 09 16:27
juhp	fujiwarat: depends which bug we are talking about :)	Sep 09 16:27
fujiwarat	Oh sorry, I mean 509518	Sep 09 16:28
fujiwarat	I thought it would be a RFE since we have a workaround.	Sep 09 16:29
juhp	fujiwarat: me too	Sep 09 16:29
fujiwarat	ok	Sep 09 16:29
juhp	not saying how to fix but Shift-2 -> " is not acceptable on us kbd IMHO	Sep 09 16:29
juhp	actually I thought fix should be in IME	Sep 09 16:30
juhp	phuang: anyway please have another look	Sep 09 16:30
juhp	was hoping we could get rid of keyboard override in ibus-setup	Sep 09 16:31
juhp	.bug 514192	Sep 09 16:32
zodbot	juhp: Bug 514192 ibus toolbar should use config icon for setup - https://bugzilla.redhat.com/show_bug.cgi?id=514192	Sep 09 16:32
juhp	I started looking at this	Sep 09 16:32
juhp	looks straightward but needs changes to all IME I guess :-\	Sep 09 16:32
phuang	get rid of keymap in ibus? and only uses xkb ?	Sep 09 16:32
phuang	juhp, yeah	Sep 09 16:32
juhp	phuang: no I am saying it should be done through IME	Sep 09 16:32
juhp	not globally	Sep 09 16:33
phuang	ic	Sep 09 16:33
juhp	but dunno how much we can do in time for f12	Sep 09 16:33
juhp	I would say might be better to turn off ibus-anthy override for now if we can't find a proper solution quickly	Sep 09 16:34
phuang	ok. to use xkb as default setting	Sep 09 16:34
phuang	ok?	Sep 09 16:35
fujiwarat	schemas file can be set the default value by locale.	Sep 09 16:35
fujiwarat	If you want hardcoded us layout, probably it might be better to set ja only.	Sep 09 16:36
phuang	we don't like it. jensp, right?	Sep 09 16:36
tagoh3	I thought that feature isn't yet implemented	Sep 09 16:36
tagoh3	in gconf	Sep 09 16:36
fujiwarat	no	Sep 09 16:36
fujiwarat	oh, yes done	Sep 09 16:36
<--Kaio has quit (Read error: 104 (Connection reset by peer))		Sep 09 16:37
juhp	phuang: well should only be jp for kana input - but still wonder how many people are going to input kana without jp kbd anyway	Sep 09 16:38
juhp	so value seems small to me?	Sep 09 16:38
juhp	phuang: yes	Sep 09 16:39
juhp	I mean just follow system kbd	Sep 09 16:39
juhp	(sorry irc is hard:)	Sep 09 16:39
fujiwarat	ok, I'll give the example change later.	Sep 09 16:40
juhp	okay	Sep 09 16:40
juhp	cool	Sep 09 16:40
juhp	paragan:	Sep 09 16:40
-->Kaio (n=cchance@fedora/Kaio) has joined #fedora-i18n		Sep 09 16:40
juhp	.bug 501673	Sep 09 16:40
zodbot	juhp: Bug 501673 [indic] Shortcut key should be ctrl+shift for changing keymaps - https://bugzilla.redhat.com/show_bug.cgi?id=501673	Sep 09 16:40
juhp	.bug 239414	Sep 09 16:41
zodbot	juhp: Bug 239414 illogical mapping for non-QWERTY keyboard with m17n - https://bugzilla.redhat.com/show_bug.cgi?id=239414	Sep 09 16:41
juhp	just pinging those	Sep 09 16:41
paragan	ok	Sep 09 16:41
juhp	any plans there? :)	Sep 09 16:41
juhp	wish we could fix 239414 for f12	Sep 09 16:41
juhp	it is a long-standing bug and would highlight improve with ibus over scim say	Sep 09 16:42
paragan	yes need to work on it	Sep 09 16:42
paragan	ok	Sep 09 16:42
juhp	but I fear it needs upstream changes to label maps that assume qwerty	Sep 09 16:42
dychen	Mmm, not quite sure a table for keylayout <-> input map helps.	Sep 09 16:42
juhp	dychen: why not? :)	Sep 09 16:43
paragan	juhp, is alt + shift was decided as default key for changing keymaps in ibus	Sep 09 16:43
juhp	s/qwerty/US/	Sep 09 16:43
dychen	paragan: Changing input methods, per se.	Sep 09 16:44
juhp	phuang: can we make Alt+Shift next hotkey for Indic?	Sep 09 16:44
dychen	paragan, what's Windows keybinding for them?	Sep 09 16:45
paragan	dychen, sorry?	Sep 09 16:45
phuang	juhp, current is alt+shift, right?	Sep 09 16:45
juhp	paragan: btw xkb uses Shift+Caps_Lock to switch layout	Sep 09 16:45
juhp	ah	Sep 09 16:45
paragan	https://bugzilla.redhat.com/show_bug.cgi?id=501673#c3	Sep 09 16:45
dychen	paragan: what's windows keybind for switching keylayout/input method?	Sep 09 16:46
paragan	looks to me its alt+shift in ibus and its ctrl+shift in scim	Sep 09 16:46
juhp	phuang: so it's already fixed?	Sep 09 16:46
paragan	dychen, sorry I don't know	Sep 09 16:46
phuang	juhp, no	Sep 09 16:46
dychen	:-)	Sep 09 16:46
juhp	paragan: ah ok (actually reading;) so you want Ctrl-Shift back hmm	Sep 09 16:47
juhp	phuang: so can we have Ctrl-Shift for indic? ;)	Sep 09 16:47
paragan	juhp, phuang yes. That was request got from scim users	Sep 09 16:47
phuang	juhp, we can	Sep 09 16:47
dychen	juhp: Doesn't warren against that idea? :-)	Sep 09 16:47
*juhp doesn't think Ctrl-Shift is a good hotkey		Sep 09 16:47
phuang	juhp, actually, we like ctrl-shift too	Sep 09 16:48
juhp	phuang: indic only?	Sep 09 16:48
juhp	ah yeah	Sep 09 16:48
phuang	juhp, why indic only?	Sep 09 16:48
juhp	ok let's close wontfix ;-)	Sep 09 16:48
phuang	juhp, we don't like locale different settings. right	Sep 09 16:48
juhp	or any acceptable solution?	Sep 09 16:48
dychen	juhp: I think it Ctrl-Shift is from win95, Alt-Shift is from winxp.	Sep 09 16:48
juhp	hmm	Sep 09 16:49
juhp	I think Alt-Shift is better	Sep 09 16:49
phuang	dychen, really? I think it is ctrl+shift too	Sep 09 16:49
juhp	paragan: I think better indic users got used to it	Sep 09 16:49
juhp	hmm	Sep 09 16:49
juhp	Ctrl is too soft	Sep 09 16:49
juhp	next to shift IMHO	Sep 09 16:50
dychen	phuang: Um, ctrl-shift is for IM within the same locale, alt-shift is for IM "among" different locale.	Sep 09 16:50
juhp	or we leave it open?	Sep 09 16:50
juhp	I don't see a short-term global solution	Sep 09 16:50
dychen	juhp: Maybe stay as-is.	Sep 09 16:51
juhp	dychen: for all langs?	Sep 09 16:51
juhp	okay	Sep 09 16:51
phuang	juhp, I don't know. If me, i like change to the old one 'Ctrl - Shift' :)	Sep 09 16:51
juhp	so wontfix...	Sep 09 16:51
juhp	phuang: yeah I know it is popular in zh	Sep 09 16:51
phuang	ok	Sep 09 16:51
juhp	or deferred?	Sep 09 16:52
juhp	paragan: any more comment?	Sep 09 16:52
paragan	juhp, seems Windows uses Alt+shift as per http://www.conversationexchange.com/resources/keyboard-language.php	Sep 09 16:52
juhp	okay	Sep 09 16:52
juhp	so should be acceptable	Sep 09 16:52
paragan	ok then	Sep 09 16:53
paragan	I will leave resolution for this bug to juhp and phuang	Sep 09 16:54
juhp	closed :)	Sep 09 16:54
juhp	thanks :)	Sep 09 16:54
phuang	ok	Sep 09 16:54
juhp	ok ok	Sep 09 16:54
paragan	ok	Sep 09 16:54
juhp	let's finish - didn't plan to go on so long...	Sep 09 16:54
juhp	but glad we did cover some bugs in detail :)	Sep 09 16:54
juhp	let's try to keep pushing on the bugs for beta	Sep 09 16:55
juhp	# meeting closed	Sep 09 16:56
juhp	phew	Sep 09 16:56
paragan	thanks all	Sep 09 16:56
dychen	Thanks.	Sep 09 16:57
tagoh3	thanks all	Sep 09 16:57
phuang	thanks	Sep 09 16:59
fujiwarat	thx	Sep 09 16:59

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!