IME composition options window misplaced in OOPIFs (Japanese IME on Mac) |
||
Issue descriptionVersion: 55.0.2875.0 (Official Build) canary (64-bit) OS: Mac OSX What steps will reproduce the problem? (1) Focus an out of process <iframe> (make sure --site-per-process is one). Alternatively, you should be able to do it with an extension (hangout) with --isolate-extensions. (2) Activate Japanese IME and type some words to have multiple clauses committed. (3) Move back and forth with arrow keys between clauses. What is the expected output? The composition options window should be placed close to currently focused clause (bold underline) What do you see instead? Window appears on the bottom of the screen (left).
,
Oct 11 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/71d0ae311054949697cf26b354b6b9a4efbbbaa4 commit 71d0ae311054949697cf26b354b6b9a4efbbbaa4 Author: ekaramad <ekaramad@chromium.org> Date: Tue Oct 11 16:03:57 2016 Request composition info from the focused RenderWidget Currently on Mac, we do not setup monitoring composition range which leads to misplaced composition window for Japanese IME (When moving back and forth between composition clauses). This CL makes the changes to route the monitoring IPC to the currently active RWH. BUG= 651604 Review-Url: https://codereview.chromium.org/2373243005 Cr-Commit-Position: refs/heads/master@{#424443} [modify] https://crrev.com/71d0ae311054949697cf26b354b6b9a4efbbbaa4/content/browser/renderer_host/render_widget_host_view_mac.mm [modify] https://crrev.com/71d0ae311054949697cf26b354b6b9a4efbbbaa4/content/browser/renderer_host/render_widget_host_view_mac_unittest.mm
,
Oct 11 2016
Discussion in the CL above cont'd: IME is still usable. The issue is only with Japanese IME and often happens when we try to change a composition by going back (press left arrow). Then the window is misplaces but as far as I can tell, it is usable.
,
Oct 11 2016
Sounds like this should be fixed by r424443. Let's verify it in tomorrow's Canary (likely 56.0.2888.0). Comment 3 is about whether we want to merge this to M54/M55. I think it might be a little late to merge to M54, though it's a small change and may be worthwhile if it's disruptive enough for users. We should definitely request a merge to M55 once the fix has been verified and baked for a few days on Canary (since --isolate-extensions is enabled by default in M55). Thanks! |
||
►
Sign in to add a comment |
||
Comment 1 by ekaramad@chromium.org
, Sep 29 2016