UbuntuIRC / 2012 /01 /25 /#ayatana.txt
niansa
Initial commit
4aa5fce
[00:00] <bschaefer> jaytaoko, as that env var is the cause of the bug in unity because the Korean IM has a bug in it...
[00:08] <jaytaoko> bschaefer: I tried with IBUS_ENABLE_SYNC_MODE but I wasn't quite sure what I was doing...
[00:09] <jaytaoko> bschaefer: I am actually looking forward to your inputs on this.
[00:09] <jono> folks, just to let you know - we are re-directing this channel to #ubuntu-unity
[00:10] <bschaefer> jaytaoko, yeah what the IBUS_ENABLE_SYNC_MODE does is tell ibus to push ALL key events through the IM engine; then if that fails it pushs it through gkt_filter_keypress
[00:10] <bschaefer> or XIM if you are using that with ibus
[00:15] <bschaefer> jaytaoko, otherwise if it isn'
[00:15] <bschaefer> t set it goes through an aysnc method in ibus and unity cant handle normal text because it never gets through gtk_filter_keypress
[00:16] <jaytaoko> bschaefer: I see
[00:16] <jaytaoko> thanks for this update
[00:17] <bschaefer> jaytaoko, yeah, I spent last week digging though ibus and ibus-hangul...
[00:18] <bschaefer> jaytaoko, the bug with that env var is for Korean when the IM engine doesn't handle a key event it sends it to gtk_filter_keypress before it commit the preedit (causing a space to commit before the preedit)
[00:19] <jaytaoko> bschaefer: ah!
[00:19] <bschaefer> jaytaoko, so I would much rather get this working in nux then going upstream to fix bugs in all the korean im engines...
[00:20] <jaytaoko> bschaefer: this is my first foray into input method. I have found the documentation was not helping much.
[00:21] <jaytaoko> bschaefer: but it seems you know a great deal more than I do
[00:21] <bschaefer> jaytaoko, yeah input methods are very annoying I spent a lot of time working with njpatel on it with unity; and I couldnt find any helpful documentation
[00:22] <bschaefer> jaytaoko, haha, its because ibus seems to hate unity!
[00:22] <bschaefer> maybe it will like nux
[00:22] <jaytaoko> lets hope so:-D
[00:23] <bschaefer> and lets hope I know enough now to help!
[00:24] <jaytaoko> bschaefer: well more eyes on the matter is always a good thing
[00:24] <bschaefer> jaytaoko, that is very true! Im really curious how the code looks with out using GTK
[00:25] <jaytaoko> bschaefer: I am preparing a branch, you will find out soon!
[00:26] <bschaefer> jaytaoko, sweet! Im installing ibus and stuff from source right now so no rush. (Finally re installed ubuntu with the new daily build)
[00:27] <bschaefer> from 11.10 haha
[00:40] <bschaefer> jaytaoko, well ibus also uses X11 (instead of gtk) to handle input through XIM, so Im guessing that is what nux will have to be doing!
[00:52] <jaytaoko> bschaefer: yes, nux relies on X11 , however, I don't remember using XIM in the input method code...
[00:53] <jaytaoko> bschaefer: this could be a problem them
[00:53] <jaytaoko> then
[00:54] <bschaefer> jaytaoko, hmm I think that is what ibus uses if useing X11 rather then gtk
[00:54] <bschaefer> jaytaoko, so I dont think you need to have any XIM code but Im not sure. It is something I can look at
[00:55] <bschaefer> jaytaoko, since unity uses the gtk method it would be good to try and push it through the ibuses X11 code
[00:55] <bschaefer> as IBUS_ENABLE_SYNC_MODE works correctly if using the X11 code. Ibus is pretty interesting on how it handles things
[00:59] <bschaefer> cd
[00:59] <bschaefer> ops
=== dyams|away is now known as dyams
=== dyams is now known as dyams|lunch
=== dyams|lunch is now known as dyams
=== MacSlow is now known as MacSlow|lunch
=== dyams is now known as dyams|away
=== greyback is now known as greyback|lunch
=== bil21al_ is now known as bil21al
=== MacSlow|lunch is now known as MacSlow
=== Saviq is now known as Saviq|afk
[14:20] <Cimi> me
=== yofel_ is now known as yofel
=== bil21al_ is now known as bil21al
=== Saviq|afk is now known as Saviq
=== chaoticuk_ is now known as chaoticuk
=== bil21al is now known as bil21al_
=== bil21al_ is now known as bil21al
=== _thumper_ is now known as thumper
=== bil21al is now known as fst8r12
=== m_conley_away is now known as m_conley