Please do not reply directly to this email. All additional comments should be made in the comments box of this bug.
https://bugzilla.redhat.com/show_bug.cgi?id=598289
--- Comment #11 from Joe Smith jes@martnet.com 2010-06-28 23:04:51 EDT --- Fantastic! Thanks for your time to look at this.
The none.conf checks the existence of im-xim.so so if you uninstall gtk2-immodule-xim, gtk-im-context-simple could be used.
This works--thanks for the tip!
This will set the default for every user account on the system, right?
Obviously, uninstalling xim means that no user on the system can use xim, but I suppose the ibus input methods will still be available, correct?
You could create $HOME/.xinputrc newly to specify GTK_IM_MODULE=gtk-im-context-simple.
Sorry, this method does not seem to work: I could not see that environment variable from a gnome-terminal shell and gtk+ applications did not provide the C-S-u input method.
I can now get the behavior I need, but I was hoping to also get some consideration of what the policy will be for the future in these two areas:
1) What is the standard configuration going to be for Fedora users in locales that use the Latin alphabet, like en_US. Will an input method be part of the standard configuration or not? If no input method is installed, will gtk-im-context-simple be used or not?
2) How will Fedora users input characters by their code? Will there be one standard method for doing this?
#1 is important because applications need to know what input method(s) to support.
I recently filed a bug against OpenOffice because it sometimes records the digit keystrokes from the C-S-u character input. Now that I've upgraded to F13, this method is no longer part of the default Fedora configuration, and the OOo devs will not fix the bug. Some kind of policy is needed, so developers know what they should support.
#2 is important because this an important task. Many users will need some way to enter characters by their code, and it's always better to have a single, standard method for doing a task.
This is an important task because many users will need access to some characters that are not part of their home language. For example, entering punctuation (bullets or dashes), or technical symbols, or developers needing access to an exact code point. Sometimes, you really just want to enter a character by its code.
I've provided a lot of support for OpenOffice users working in many different locales, and the need to enter a specific character comes up more often than you might expect. Having one standard method for doing it is a big win.
Thanks again for your time.
i18n-bugs@lists.fedoraproject.org