Re: GTK+ HEAD and Pango HEAD
- From: Owen Taylor <otaylor redhat com>
- To: Arnaud Charlet <charlet ACT-Europe FR>
- Cc: Keith Packard <keithp keithp com>, gtk-devel-list gnome org, jacob ximian com
- Subject: Re: GTK+ HEAD and Pango HEAD
- Date: Fri, 2 Aug 2002 17:14:29 -0400 (EDT)
Arnaud Charlet <charlet ACT-Europe FR> writes:
> > In my mind "if I haven't used it before, if I don't see it used
> > all over the place in other software, it's a potential portability
> > problem" ... too much hard experience with X stuff that really
> > should have been portable.
> Speaking of which, any reason why the following functions
> are used in Gtk+:
> they are used in gdk/x11/gdkmain-x11.c
> These are apparently X11R6 extensions, and make it impossible to
> build Gtk+ out of the box under Solaris 2.5.1
> Now, if you comment these calls out, Gtk+ seems to work fine,
> so this is a bit puzzling, and if these calls could be removed,
> that'd be great.
They are needed for correct operation with input methods.
> The other piece of code that uses non existent calls under Solaris 2.5.1
> is in modules/input/gtkimcontext.c (XIMValuesList, XIC). This file
> gets compiled, even when using --disable-xim
The --disable-xim configure option does nothing now, and should be
If people are really interested in GTK+-2.0 on X11R5, the right
thing to is probably to check for X11R6, and if not present
disable the X11R6 specific functions and gtkimcontextxim.
(There was XIM for X11R5 but it didn't work properly. Then there were
releases of Solaris that contained bits of X11R6 XIM that didn't quite
work the same as the final release.)
On the other hand, X11R6 was released on May 2, 1994, over 8
years ago, and 2 years before development of GTK+ started...
] [Thread Prev