Commit e2a52192 authored by Owen Taylor's avatar Owen Taylor

Merge from themes-2. See the ChangeLog for a somewhat detailed

history of the evolution of the changes involved. Most of this
is actually minor painting tweaks.
parent 3c0df19a
......@@ -10,4 +10,3 @@ aclocal.m4
configure
gtk-config
config.cache
Thu Nov 5 12:06:16 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkcheckbutton.c (gtk_check_button_draw_focus): Queue
a clear/redraw, not just a redraw, so the focus
gets undrawn properly.
* gtk/gtktogglebutton.c (gtk_toggle_button_unrealize):
Chain to gtk_widget_unrealize() instead of duplicating
code.
Wed Nov 4 14:34:51 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdk.c gdk/gdkprivate.h gdk/gdkwindow.c: Removed
last vestiges of old DND.
Tue Nov 3 12:27:52 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkentry.c: Restored inner border to entries,
erase cursors to background image.
Fri Oct 30 12:09:21 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkentry.c gtk/gtkstyle.[ch]: removed paint_entry() -
it was identical to paint_flat_box().
* gtk/gtkstyle.c: Coding style fixups.
* gtk/gtkwindow.c (gtk_window_move_resize): removed useless
test.
* gtk/gtkframe.c gtkstyle.[ch] gtknotebook.c:
Change 'side' for paint_[box/shadow]_gap and paint_extension
to be GtkPositionType instead of a random int.
* gtk/gtkaspectframe.c (gtk_aspect_frame_paint): Use
"frame" as detail - an aspectframe should draw identical
to a frame.
Tue Oct 27 09:58:11 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdkwindow.c (gdk_window_merge_child_shapes): renamed
from combine_child_shapes().
* gdk/gdkwindow.c: Merge propagate_shapes() and
propagate_combine_shapes() to remove code duplication.
Thu Nov 5 14:31:57 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/gdk.c: sometimes we get a KeyRelease before a KeyPress in
gdk_event_translate, and the USE_XIM buffer isn't allocated yet.
So allocate some mem if it's not already done.
* clist scrollbar merges from HEAD
Tue Nov 3 13:16:36 PST 1998 Manish Singh <yosh@gimp.org>
* gtk/Makefile.am: remove explict -ldl, gmodule takes care of that
* gtk/gtkdnd.c: start TARGET_MOTIF_SUCCESS from 0x40000000 instead
of 0x80000000, so we stay within the bounds of a postive signed
int (it broke some compilers)
* gtk/gtkmenushell.c
* gtk/gtkoptionmenu.c: merges from HEAD to make option menus work
when some of the menuitems are hidden
* gtk/gtkthemes.h: don't include dlfcn.h explictly, gmodule takes
care of that
* gtk/gtkwidget.c: add a dummy continue after the line label
next_rect, since some compilers don't like label: }
Sun Nov 1 20:05:55 PST 1998 Manish Singh <yosh@gimp.org>
* Okay, so we need the X11R6 keysyms so programs don't have to
worry about them. They don't conflict anyway. Reverted the #ifdef
changes to gtk/*.c
* gdk/Makefile.am: remove the BUILT_SOURCES rule, and make a
special rule for them (X-derived-headers)
Sun Nov 1 05:30:32 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/Makefile.am: changed the awk scripts to generate gdkcursors.h
and gdkkeysyms.h into sed scripts, since the awk scripts weren't
portable, and we really shouldn't be using awk anyway.
* makecursors.awk
* makekeysyms.awk: bye bye
* gtk/gtkaccelgroup.c
* gtk/gtkclist.c
* gtk/gtknotebook.c
* gtk/gtkwindow.c: don't use X11R6 keysyms if they aren't there
Sat Oct 31 15:22:14 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/Makefile.am: do the same built sources trick for gdkcursors.h
and gdkkeysyms.h. Remove those files from CVS.
Sat Oct 31 14:44:01 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/gdk.c: move prototype of gdk_wm_protocols_filter outside
of the USE_XIM #ifdef
* gtk/Makefile.am: cause the built sources to be rebuild when possible,
even with --include-deps. Slightly different from timj's fix, to avoid
circular dependency with gtkmarshal.c
* autogen.sh
* gtk/gtkclist.[ch]
* gtk/gtkctree.c
* gtk/gtkfontsel.[ch]
* gtk/gtklist.c
* gtk/gtklistitem.c
* gtk/gtkobject.[ch]
* gtk/testgtk.c: merges from HEAD
* gtk/gtkmenubar.c
* gtk/gtkradiomenuitem.c
* gtk/gtktreeitem.c: cleanups
Wed Oct 28 23:32:19 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/gdkrgb.c: Oops. s/BIGENDIAN/BIG_ENDIAN/
Wed Oct 28 19:04:49 PST 1998 Manish Singh <yosh@gimp.org>
* configure.in: AIX has libpthreads, to make sure you can have
more than one thread. Also add -D_THREAD_SAFE, and reorder
configure.in so CFLAGS doesn't get clobbered when we actually
want to modify it
* gdk/gdkimage.c: initialize the bpp and byte_order fields properly
in gdk_image_get
* gtk/testdnd.c: removed unused variable "buf"
Mon Oct 26 16:46:15 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkcontainer.c (gtk_container_unregister_toplevel):
remove_link(), not remove()
Thu Oct 22 02:27:30 PDT 1998 Manish Singh <yosh@gimp.org>
* gdk/gdkrgb.c: use glib endian stuff
* configure.in: don't need to check for endianness anymore
* gtk/gtkstyle.h: removed #pragma }, causing a silly warning
Wed Oct 21 17:44:27 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwindow.c (gtk_window_read_rcfiles): Reset
all toplevel widgets, not toplevel windows. (I.e.,
plugs should be reset)
* gtk/gtkwindow.c (gtk_window_read_rcfiles): Yes, we
_do_ need to notify the other toplevels - by the time
the get the client messages, the RC file dates will
have been updated.
* gtk/gtkcontainer.h: Added a function call to list
every toplevel widget.
* gdk/gdk.c (gdk_event_send_client_message_to_all_recurse):
Don't send client messages to _every_ window on display!
Tue Oct 20 16:21:04 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdnd.c (gtk_drag_source_event_cb): Don't start drags
when dragging across widgets, only in drags
starting in widget.
Tue Oct 20 15:59:49 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdnd.c (gtk_drag_source_info_destroy):
Emit "drag_end" signal at end of drag.
Tue Oct 20 13:07:54 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwindow.c (gtk_window_read_rcfiles): Look
for "gtk-embedded" data currently set by
GtkSocket, and use that to forward client message
on to embedded windows. Only reset our own toplevel -
the others will handle themselves. (Now that
GtkMenus live in GtkWindows)
Mon Oct 19 16:40:40 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdkdnd.c (get_client_window_at_coords_recurse): Modify
client window finding code so that it works with
WM's other than fvwm.
Sun Oct 18 22:48:27 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdrawwindow.c (gtk_draw_window_get_type): Added
the "GtkDrawWindow" class, which is a descendent
of the Window class which leaves its background alone.
Essentially, a toplevel drawing area, though it does
allow adding a child.
Sat Oct 17 00:40:51 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkrc.c (gtk_rc_style_init): Duplicate bg_pixmap_name when
merging RC styles. Report new share/themes path for
themes directory. Deref theme engines when appropriate.
* gtk/gtkclist.c: Set all backgrounds in _style_set().
Make sure that pixel for xor_gc is never 0.
* gtk/gtkctree.c: Choose colors for plus/minus more
carefully, in a hopefully style-independent fashion.
(Still needs a bit more work)
* gtk/gtkwidget.c: Don't send expose events to widgets
which are about to be resized and redrawn. (Needs a
bit more thought.)
* gtk/gtknotebook.c (gtk_notebook_draw_tab): Avoid
drawing on tabs for newly allocated, un-sized
pages - we'll get the size wrong.
Thu Oct 1 17:38:07 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwidget.c (gtk_widget_redraw_queue_remove): Fixed
up stupidities in last commit.
Thu Oct 1 14:25:43 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwidget.c: When handling queued resizes, account
for handleboxes, where the widget heirarchy is different
than the window heirarchy by just queueing a resize on
the entire handlebox if the detached child is changed.
- Remove widgets from the redraw queue when reparenting,
even if not unrealizing. We queue a resize/redraw in
any case.
Mon Sep 28 21:07:01 EDT 1998 The Rasterman <raster@redhat.com>
* gtk/gtkstyle.c: added check to see if the widget has been shaped
byt the app via gtk_widget_shape_combine_mask and if it has does
not reset the shape on theme change.
Tue Sep 15 11:30:03 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkfilesel.c: Keep a list of directories in which we never
want to stat the entries (like /afs), and assume everything in
those directories is a subdirectory.
- When following a path, try to open components even when
we don't find them in their parent's directory to
support automounters.
Fri Sep 11 15:36:33 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdkinputgxi.h (gdk_input_gxi_grab_pointer): Hopefully
now at least compiles again.
* gdk/gdk.c (gdk_event_translate): Don't pass events
on destroyed widgets to the gdk_input layer.
Thu Sep 3 19:45:34 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkthemes.c: Use g_module instead of dlopen()
Thu Sep 3 19:30:11 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtktooltips.c (gtk_tooltips_expose): Check
if we've changed widgets since the expose event
was queued.
Mon Aug 31 15:33:13 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkstyle.h (struct _GtkStyle): Reordered a bit
to restore approximate bin compatibility with the
non-themes branch. (This does break bin compat
within the themes branch.)
Sun Aug 30 16:48:27 1998 Owen Taylor <otaylor@gtk.org>
* gtk/gtklabel.c (gtk_label_set): Eliminate a
gdk_window_clear_area() that snuck in with the
menu underlines merge.
Mon Aug 24 23:37:26 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtktext.c: Fixed some remaining background drawing,
merges from main branch, cleanups.
Mon Aug 24 10:55:51 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtknotebook.c (gtk_notebook_draw_tab): Don't
dereference NULL focus tab.
* gtk/gtkscale.c (gtk_scale_[un]map): Add map/unmap
functions since we need to both show/hide and
queue a redraw. (There's some duplication of drawing here,
we probably should only be queuing a redraw on the
"value" portion of the scale.)
Fri Aug 21 19:02:30 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkrange.[ch] gtk/gtkhscale.c gtk/gtkvscale.[ch]: Fit scales
into the theme-drawing framework. We do this by making them,
like check/radio-buttons, no-window widgets with subwindows. We
then move the value by calling gtk_widget_queue_clear() and only
redraw the value in our draw() / expose() handler. A virtual
function clear_background is added which queues a clear on the
"background" portion of the widget.
Fri Aug 21 13:04:27 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtktext.c (mark_bg_gc): Make background colors work
again. (We can't always paint the background with a
paint() function - only do that for selected text)
Mon Aug 17 20:31:01 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtktearoffmenuitem.c gtk/gtkclist.[ch] gtk/gtkctree.[ch]
gtk/gtklabel.c gtk/gtkmenu.c: Merges from main branch.
* Makefile.am (SRC_SUBDIRS): Removed themes from SRC_SUBDIRS.
the themes can't be built as part of the same build since
they depend on gdk_imlib.
Fri Aug 14 12:44:30 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkthemes.c (gtk_theme_engine_get): Removed sample engine code
from gtkthemes.c.
* configure.in ltconfig.in ltmain.sh gdk/Makefile.am gtk/Makefile.am
gtk/
- Upgrade to libtool-1.2b + patch to add shared library dependencies
for Linux. (Needs to be extended to other platforms where they
are safe)
- Check for -ldl so this branch should build on all platforms, though
themes will work on all platforms they could work on.
- Use -version and -release simultaneously: sonames stay as they are
in the current scheme, but -lgtk-1.1 changes to -lgtk.
- Added in extra -l flags to library LDFLAGS so we get shared library
dependencies against X libs and -ldl. (But not libgtk against
libgdk or libglib yet, since dependencies against uninstalled
libraries aren't portable)
Fri Aug 21 16:10:38 EDT 1998 The Rasterman <raster@redhat.com>
* gtkrc.c - added api calls to find module and theme paths from gtk.
nuked gtk_install_prefix global.
Fri Aug 21 12:36:46 EDT 1998 The Rasterman <raster@redhat.com>
* fixed gtk_install_prefix global.
Thu Aug 20 19:27:49 EDT 1998 The Rasterman <raster@redhat.com>
* added gtk_install_prefix global const variabel to find where gtk is
installed.
Thu Aug 20 18:27:29 EDT 1998 The Rasterman <raster@redhat.com>
* added ~/.gtk/lib/themes/engines/ to the default searchpath for
gtk theme engines
Thu Aug 20 18:27:29 EDT 1998 The Rasterman <raster@redhat.com>
* made gtkrc scanner append dir that gtkrc is read from to pixmap path.
Mon Aug 10 23:03:55 1998 The Rasterman <raster@redhat.com>
* Fixed minor problem in gtkhandlebox.c with drawing, and some
notebook stuff. Also pixmap theme is now "clean" again.
Mon Aug 10 23:03:55 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwidget.c (gtk_widget_clip_rect): Fixed up
queued area resizing once more.
Mon Aug 10 23:09:30 EDT 1998
* Added a few drawing fixes for notebook and other widgets. finished
pixmap theme up to the level that can be supported so far. (need to
add path discovery 9ie where is the rc fiel from - add to pixmap
path so we can get theimages from there too). then we can package
themes. After that need to add to theme seartch path stuff in users
home dir.
Thu Aug 6 16:49:25 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtktooltips.c (gtk_tooltips_expose): Draw into a
drawing area, draw in an expose handler.
* gtk/gtknotebook.c (gtk_notebook_expose): Check for
empty current page.
Wed Aug 5 16:13:15 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkcolorsel.c (gtk_color_selection_draw_wheel): Moved
creation of GC's so that they will be set before
they are used for map events (why are they used there?)
* gtk/gtkwidget.c: Fixed mistake that caused widgets to
be queued in multiple times in the queue.
Wed Aug 5 14:23:52 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkstyle.c (gtk_default_draw_shadow): Don't set
clip rects (or do anything) for shadow_none.
Wed Aug 5 13:14:57 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkbutton.c (gtk_button_paint): Restored paint function.
Yes, it was needed. draw() and expose() are _not_
identical.
Sat Aug 1 17:44:31 EDT 1998 EDT 1998 The Rasterman <raster@redhat.com>
* Wheeeeeeeeeeeeeeee fixed radiobittons to work in gimp again.
Fri Jul 31 19:45:12 EDT 1998 The Rasterman <raster@redhat.com>
* YES YES YES YES YES clist borders work again! YES! YES YES! gimme
more baby oooh yeah more more MORE YES YES.
Fri Jul 31 19:45:12 EDT 1998 The Rasterman <raster@redhat.com>
* hopefully fixed problem wiht togglebuttons acting like
radiobuttons in gtktogglebutton.c
Thu Jul 30 23:36:46 EDT 1998 The Rasterman <raster@redhat.com>
* fixed another drawing buglet in the frame drawing code (actually
in the clearing code but anyway in gtkstyle.c)
Thu Jul 30 19:23:48 EDT 1998 The Rasterman <raster@redhat.com>
* Fixed warning from colorsel widget...
Thu Jul 30 19:10:42 EDT 1998 The Rasterman <raster@redhat.com>
* Fixed defautl drawing code to no longer clobber the original style
pixmap bgs as it did and added a helper funciton og gtkstyle.c to
help with this.
Thu Jul 30 18:09:21 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkrc.c gtk/gtkstyle.c: More mm fixups
* gtk/gtktooltips.c: Pass a widget (the window) to the
drawing function instead of the GtkTooltips.
Thu Jul 30 17:27:10 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdkwindow.c: Abstract out checking for shape
extension into a function.
* gtk/gtkwindow.c: Fix up gtk_window_expose()/draw().
Remove extra callls to gtk_window_draw().
Thu Jul 30 11:18:43 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkstyle.c gtk/gtkrc.c: Memory management fixups.
* themes/th_main.c: Moved messages around so they are
printed before destroying strings.
Wed Jul 29 16:56:07 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdk.c: Fairly simple implementation of expose compression.
(No GraphicsExpose compression yet, always compress
with the equivalent of XtCompressMultiple - no compression
of Expose events across non-Expose events)
* gdk/gdk.h gdk/gdkrectangle.c: gdk_rectangle_union()
Find rectangle bounding two rectangles.
* gtk/gtkstyle.c: Set clip rects for a few extra GC's
in gdk_default_draw_shadow()
Wed Jul 29 16:03:20 EDT 1998 The Rasterman <raster@redhat.com>
* Fixed drawing functions to handle switching themes on the fly
gracefully and look decent.
Tue Jul 28 22:52:10 EDT 1998 The Rasterman <raster@redhat.com>
* Added all the theme drawing code to all widgets, fixed up minor
nigglies in some widgets when it comes to themes, seem to be
running into some funky behavior with clipping rects in GC's. Too
many files to list here. (all files that now use gtk_paint_* are
affected).
Tue Jul 28 22:49:16 1998 Owen Taylor <otaylor@redhat.com>
* themes/th_main.c themes/th_draw.c gtk/testgtkrc:
Set up the demo theme so it is loaded from the RC file.
* themes/th_draw.c: Alleviate the hardcoded paths a wee
bit. (use $HOME/themes/)
* gtk/gtkthemes.[ch] gtk/gtkmain.c: Load engines specified in
the RC file via dlopen()
Tue Jul 28 20:30:56 1998 Owen Taylor <otaylor@redhat.com>
Internal handling of styles and RC styles pretty much
completely reworked for Theme work.
* gtkthemes.[ch]:
- Added routines for virtualization of RcStyle and Style handling.
- Temporary "sample" inline theme-engine. (Does nothing
but print out a few messages.
* gtkrc.[ch]
- Made GtkRcStyle public, added flags to tell if particular
colors have been set.
- RC Styles are "transparent", and the final style for
a widget is the composition of all matching styles
- Added pointer to GtkThemeEngine, and gpointer engine_data
to RC style and style.
- Parse "engine" keyword, load the appropriate theme engine
and call the engine's parser.
* gtkstyle.[ch]
- Keep a GtkRcStyle * in styles created from Rc Styles.
- Load up images from RC file at style attachment time.
- Call ->engine->foo() at appropriate times.
- Don't use a GCache. All incarnations for a style are
just linked together in a list, depend on the RC style
code to find the existing copy.
Tue Jul 28 14:41:54 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwidget.[ch]: Added the concept of queued redraws
of partial areas. This can be done either relative
to the existing windows (gtk_widget_queue_clear[_area]())
or relative to the future allocation of the widget
(gtk_widget_queue_draw[_area]).
The initial implementation is pretty braindead, but
can be improved.
* gtk/gtkwidget.[ch]: Removed calls to gdk_window_clear_area()
in favor of gtk_widget_queue_clear().
* gtk/gtklabel.[ch] gtk/gtkpixmap.[ch]: Removed needs_clear
flags in favor of gtk_widget_queue_clear()
Tue Aug 4 19:12:14 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwidget.c (gtk_widget_idle_draw): Fixed up
handling of partially clipped areas.
Thu Nov 5 18:52:35 1998 Lars Hamann <lars@gtk.org>
* gtk/gtkclist.h (struct _GtkCListClass)
......
Thu Nov 5 12:06:16 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkcheckbutton.c (gtk_check_button_draw_focus): Queue
a clear/redraw, not just a redraw, so the focus
gets undrawn properly.
* gtk/gtktogglebutton.c (gtk_toggle_button_unrealize):
Chain to gtk_widget_unrealize() instead of duplicating
code.
Wed Nov 4 14:34:51 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdk.c gdk/gdkprivate.h gdk/gdkwindow.c: Removed
last vestiges of old DND.
Tue Nov 3 12:27:52 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkentry.c: Restored inner border to entries,
erase cursors to background image.
Fri Oct 30 12:09:21 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkentry.c gtk/gtkstyle.[ch]: removed paint_entry() -
it was identical to paint_flat_box().
* gtk/gtkstyle.c: Coding style fixups.
* gtk/gtkwindow.c (gtk_window_move_resize): removed useless
test.
* gtk/gtkframe.c gtkstyle.[ch] gtknotebook.c:
Change 'side' for paint_[box/shadow]_gap and paint_extension
to be GtkPositionType instead of a random int.
* gtk/gtkaspectframe.c (gtk_aspect_frame_paint): Use
"frame" as detail - an aspectframe should draw identical
to a frame.
Tue Oct 27 09:58:11 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdkwindow.c (gdk_window_merge_child_shapes): renamed
from combine_child_shapes().
* gdk/gdkwindow.c: Merge propagate_shapes() and
propagate_combine_shapes() to remove code duplication.
Thu Nov 5 14:31:57 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/gdk.c: sometimes we get a KeyRelease before a KeyPress in
gdk_event_translate, and the USE_XIM buffer isn't allocated yet.
So allocate some mem if it's not already done.
* clist scrollbar merges from HEAD
Tue Nov 3 13:16:36 PST 1998 Manish Singh <yosh@gimp.org>
* gtk/Makefile.am: remove explict -ldl, gmodule takes care of that
* gtk/gtkdnd.c: start TARGET_MOTIF_SUCCESS from 0x40000000 instead
of 0x80000000, so we stay within the bounds of a postive signed
int (it broke some compilers)
* gtk/gtkmenushell.c
* gtk/gtkoptionmenu.c: merges from HEAD to make option menus work
when some of the menuitems are hidden
* gtk/gtkthemes.h: don't include dlfcn.h explictly, gmodule takes
care of that
* gtk/gtkwidget.c: add a dummy continue after the line label
next_rect, since some compilers don't like label: }
Sun Nov 1 20:05:55 PST 1998 Manish Singh <yosh@gimp.org>
* Okay, so we need the X11R6 keysyms so programs don't have to
worry about them. They don't conflict anyway. Reverted the #ifdef
changes to gtk/*.c
* gdk/Makefile.am: remove the BUILT_SOURCES rule, and make a
special rule for them (X-derived-headers)
Sun Nov 1 05:30:32 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/Makefile.am: changed the awk scripts to generate gdkcursors.h
and gdkkeysyms.h into sed scripts, since the awk scripts weren't
portable, and we really shouldn't be using awk anyway.
* makecursors.awk
* makekeysyms.awk: bye bye
* gtk/gtkaccelgroup.c
* gtk/gtkclist.c
* gtk/gtknotebook.c
* gtk/gtkwindow.c: don't use X11R6 keysyms if they aren't there
Sat Oct 31 15:22:14 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/Makefile.am: do the same built sources trick for gdkcursors.h
and gdkkeysyms.h. Remove those files from CVS.
Sat Oct 31 14:44:01 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/gdk.c: move prototype of gdk_wm_protocols_filter outside
of the USE_XIM #ifdef
* gtk/Makefile.am: cause the built sources to be rebuild when possible,
even with --include-deps. Slightly different from timj's fix, to avoid
circular dependency with gtkmarshal.c
* autogen.sh
* gtk/gtkclist.[ch]
* gtk/gtkctree.c
* gtk/gtkfontsel.[ch]
* gtk/gtklist.c
* gtk/gtklistitem.c
* gtk/gtkobject.[ch]
* gtk/testgtk.c: merges from HEAD
* gtk/gtkmenubar.c
* gtk/gtkradiomenuitem.c
* gtk/gtktreeitem.c: cleanups
Wed Oct 28 23:32:19 PST 1998 Manish Singh <yosh@gimp.org>
* gdk/gdkrgb.c: Oops. s/BIGENDIAN/BIG_ENDIAN/
Wed Oct 28 19:04:49 PST 1998 Manish Singh <yosh@gimp.org>
* configure.in: AIX has libpthreads, to make sure you can have
more than one thread. Also add -D_THREAD_SAFE, and reorder
configure.in so CFLAGS doesn't get clobbered when we actually
want to modify it
* gdk/gdkimage.c: initialize the bpp and byte_order fields properly
in gdk_image_get
* gtk/testdnd.c: removed unused variable "buf"
Mon Oct 26 16:46:15 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkcontainer.c (gtk_container_unregister_toplevel):
remove_link(), not remove()
Thu Oct 22 02:27:30 PDT 1998 Manish Singh <yosh@gimp.org>
* gdk/gdkrgb.c: use glib endian stuff
* configure.in: don't need to check for endianness anymore
* gtk/gtkstyle.h: removed #pragma }, causing a silly warning
Wed Oct 21 17:44:27 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwindow.c (gtk_window_read_rcfiles): Reset
all toplevel widgets, not toplevel windows. (I.e.,
plugs should be reset)
* gtk/gtkwindow.c (gtk_window_read_rcfiles): Yes, we
_do_ need to notify the other toplevels - by the time
the get the client messages, the RC file dates will
have been updated.
* gtk/gtkcontainer.h: Added a function call to list
every toplevel widget.
* gdk/gdk.c (gdk_event_send_client_message_to_all_recurse):
Don't send client messages to _every_ window on display!
Tue Oct 20 16:21:04 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdnd.c (gtk_drag_source_event_cb): Don't start drags
when dragging across widgets, only in drags
starting in widget.
Tue Oct 20 15:59:49 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdnd.c (gtk_drag_source_info_destroy):
Emit "drag_end" signal at end of drag.
Tue Oct 20 13:07:54 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwindow.c (gtk_window_read_rcfiles): Look
for "gtk-embedded" data currently set by
GtkSocket, and use that to forward client message
on to embedded windows. Only reset our own toplevel -
the others will handle themselves. (Now that
GtkMenus live in GtkWindows)
Mon Oct 19 16:40:40 1998 Owen Taylor <otaylor@redhat.com>
* gdk/gdkdnd.c (get_client_window_at_coords_recurse): Modify
client window finding code so that it works with
WM's other than fvwm.
Sun Oct 18 22:48:27 1998 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdrawwindow.c (gtk_draw_window_get_type): Added
the "GtkDrawWindow" class, which is a descendent
of the Window class which leaves its background alone.
Essentially, a toplevel drawing area, though it does