Description of problem: when you type following keystroke with scim-sihala in kedit (correct) and gedit (wrong) e+e you can see the actual problem Version-Release number of selected component (if applicable): scim-sinhala-0.2.0-1.fc6 How reproducible: Everytime during typing above combintion Steps to Reproduce: 1. Open kedit 2. select Scim-sinhala for input 3. type e e 4 repeat step with gedit Actual results: gedit is not showing correct output as kedit Expected results: gedit's output should be like kedit Additional info: Screenshot attached
Created attachment 136141 [details] issue in gedit/kedit
(In reply to comment #1) > Created an attachment (id=136141) [edit] > issue in gedit/kedit > this feature works fine on kedit , but the problem exists in gedit
Reassigning to pango then.
I have to admit that I don't see a big difference in the screenshot.
I don't see the difference either.
Created attachment 137131 [details] Problem is clearly visible here Gedit & kedit shown in the picture with "ee" being pressed using scim-sinhala scim-1.4.4-34.fc6 scim-sinhala-0.2.0-1.fc6 pango-1.14.4-1.fc6 gedit-2.15.9-1.fc6 kdeutils-3.5.4-3.fc However, copy & paste from kedit to gedit pastes the same ligerature in gedit as it appears in kedit.
C #7 provides the required needinfo.
This can not be a bug. Its just a coincidence that shape of 0D92(ee) matches with the shape of 0D91+0DCA (e + virama). We cannot expect combination of 0D91 with 0DCA to result 0D92. 0D91 and 0D92 are both independent vowels. Ideally pango should not allow any vowel to be followed by virama(0DCA in case of sinhala).
Copy paste from kedit to gedit works, hence this seems like a scim-sinhala bug. Also, the bug summary is misleading... leading to Comment #9. Changing bug summary to more meaningfull sentence. Jens... any comments?
Does it happen even without scim-bridge or with GTK_IM_MODULE=scim?
This is a definitely not a Pango bug. The rawcodes work correctly on pango. Its mostly scim-sinhala bug if it has different implementation for gnome and kde. scim-sinhala on gnome, gives 0d91 on first keystroke on 'o'. It is expected it generates 0d92 on second 'o' keystroke and remove the earlier 0d91. This does not happen on Gnome application. But since pango is able to render 0d91 and 0d92 both properly, using rawcode, and pango does not involve in the input mechanism, its not a pango bug. Latest cvs-repository build of scim-sinhala does not have this problem. Similar problem was present with bug #207110, and that too is working fine.
This sounds very similar to bug 206261? (In reply to comment #12) > Latest cvs-repository build of scim-sinhala does not have this problem. Err, which cvs repo?
*** This bug has been marked as a duplicate of 206261 ***