Issue 381597 in chromium: Chrome doesn't handle Linux X11 forward key event anymore (breaks our language's IME)

[chro...@googlecode.com]

Status: Unconfirmed
Owner: ----
Labels: Pri-2 Via-Wizard Type-Bug-Regression OS-Linux

New issue 381597 by phaik...@gmail.com: Chrome doesn't handle Linux X11
forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

UserAgent: Mozilla/5.0 (X11; Linux i686) AppleWebKit/537.36 (KHTML, like
Gecko) Chrome/35.0.1916.114 Safari/537.36

Steps to reproduce the problem:
..

What is the expected behavior?
Chrome should handle Linux X11 forward key event like the past versions.

What went wrong?
Why doesn't chrome handle forward key event any more since 35? Some bugs in
input method support or is it intended?
Inputting Vietnamese is different from CJK that the majority prefer the
faking backspaces and inserting method to preediting like CJK. It's just
much much better. I have been very happy with the input method Bogo
(https://github.com/BoGoEngine/ibus-bogo-python) for Vietnamese. Just so
much better compared to preedit-based solutions, and most other Vietnamese
Linux users also feel like me.
Now with new chrome versions since 35 Bogo just doesn't work anymore
because the fake backspace doesn't work, chrome doesn't handle X11 forward
key event and I don't see any reasons why it's that way, a regression?

Did this work before? N/A

Chrome version: 35.0.1916.114 Channel: stable
OS Version: Ubuntu 14.04
Flash Version: Shockwave Flash 13.0 r0

--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[chro...@googlecode.com]

Updates:
Cc: yukishi...@chromium.org
Labels: Cr-UI-Input-Text-IME Proj-DesktopAura

Comment #1 on issue 381597 by yuk...@chromium.org: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

Thank you for the report. This is probably a regression caused by the Aura
transition in M35. See the following announcement for details about this
transition.
https://groups.google.com/a/chromium.org/d/msg/chromium-dev/Zpu9801pPRc/mUu_8Fyq_DcJ

Adding Cr-UI-Input-Text-IME and Proj-DesktopAura labels so that relevant
people can pick this up.

[chro...@googlecode.com]

Comment #2 on issue 381597 by phaik...@gmail.com: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

This bug has been around for such a long time, someone please fix it, it
affects a lot of people here in my country.

[chro...@googlecode.com]

Comment #4 on issue 381597 by yukishi...@chromium.org: Chrome doesn't

handle Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

I'm not a native speaker and not 100% sure, but it seems I can type
Vietnamese with ibus-unikey + Chrome Version 37.0.2062.3 dev (64-bit). As
I searched, there seems no package named ibus-bogo on ubuntu or debian (it
seems in preparation stage for 1+ year).

[chro...@googlecode.com]

Updates:
Cc: song...@chromium.org

Comment #6 on issue 381597 by manoranj...@chromium.org: Chrome doesn't

handle Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

(No comment was entered for this change.)

[chro...@googlecode.com]

Comment #7 on issue 381597 by yuk...@chromium.org: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

In reply to #5.
> I think it's a very tiny bug, would take no time to fix, but it's serious
> for people out here.

Please feel free to send us a patch to fix this issue in Chromium side if
you can fix it easily.

You can find contribution guide here.
http://www.chromium.org/developers/contributing-code

Thanks!

[chro...@googlecode.com]

Updates:
Labels: Needs-Feedback

Comment #8 on issue 381597 by hmupp...@chromium.org: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

phaikawl@, are you still seeing the issue on chrome 36.0.1985.125 stable?

[chro...@googlecode.com]

Comment #9 on issue 381597 by phaik...@gmail.com: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

@hmupp: Still the same, doesn't work.

[chro...@googlecode.com]

Comment #10 on issue 381597 by phaik...@gmail.com: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
http://code.google.com/p/chromium/issues/detail?id=381597

@yukawa: sorry I don't have enough skill to fix it myself, would take me
ages to do so, I just guess that it's easy for the people involved in the
change that broke it, they probably know what went wrong.

[chro...@googlecode.com]

Comment #11 on issue 381597 by rnimmaga...@chromium.org: Chrome doesn't

handle Linux X11 forward key event anymore (breaks our language's IME)

https://code.google.com/p/chromium/issues/detail?id=381597

@yukawa: Could you please have a look into this issue as commented in #10.

Thank you.

[chro...@googlecode.com]

Comment #13 on issue 381597 by su...@chromium.org: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)

https://code.google.com/p/chromium/issues/detail?id=381597

I don't understand the background reason which broke this IME, but IMHO, if
the IME sends Chrome faked X11 events, then Chrome should handle those
faked X11 events correctly, if not, then we should fix Chrome.
I think this bug is not like Issue 383093, where Unikey on Windows doesn't
send faked key events at all. It sends faked char events, which should
never work in Chrome.

[chro...@googlecode.com]

Comment #14 on issue 381597 by yuk...@chromium.org: Chrome doesn't handle

Linux X11 forward key event anymore (breaks our language's IME)
https://code.google.com/p/chromium/issues/detail?id=381597

> #13.

Yep. That's a good point. I agree that Chromium should be able to handle
X11 key events correctly no matter how the event is generated or
synthesized. Perhaps the issue might have already been fixed by
http://crrev.com/1068093002.

Having said that, keeping the correct order of key events is indeed
difficult if you are trying to consume a key event and insert another fake
key event at the same time, especially when the target application and the
IME engine (ibus-bogo-python in this case) are working in different
processes asynchronously.

Here are some examples between IBus and X11 that look to not be completely
resolved yet. I just wondered if some of them might be related.
- http://tools.android.com/knownissues/ibus
- https://github.com/ibus/ibus/issues/1786
- https://github.com/ibus/ibus/issues/1713
- https://bugs.freedesktop.org/show_bug.cgi?id=76626