2000-05-02 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdrawable-win32.c (gdk_win32_draw_arc): Don't try to
draw too narrow or too low arcs, they seem to fail, at least with
some display drivers.
* gdk/gdk.c (gdk_init_check): Use G_DIR_SEPARATOR.
2000-05-02 Tor Lillqvist <tml@iki.fi>
Large changes to the Win32 backend, partially made necessary by
the changes to the backend-independent internal
structures. Attempts to implement similar backing store stuff as
on X11. The current (CVS) version of the Win32 backend is *not* as
stable as it was before the no-flicker branch was merged. A
zipfile with that version is available from
http://www.gimp.org/win32/. That should be use by "production"
code until this CVS version is usable. (But note, the Win32
backend has never been claimed to be "production quality".)
* README.win32: Add the above comment about versions.
* gdk/gdkwindow.c: Don't use backing store for now on Win32.
* gdk/gdk.def: Update.
* gdk/gdkfont.h: Declare temporary Win32-only functions. Will
presumably be replaced by some more better mechanism as 1.4 gets
closer to release shape.
* gdk/makefile.{cygwin,msc}: Update.
* gdk/win32/*.c: Correct inclusions of the backend-specific and
internal headers. Change code according to changes in these. Use
gdk_drawable_*, not gdk_window_* where necessary.
* gdk/win32/gdkdnd-win32.c: Use MISC selector for GDK_NOTE, not
our old DND.
* gdk/win32/gdkdrawable-win32.c (gdk_win32_draw_text): Don't try
to interpret single characters as UTF-8. Thanks to Hans Breuer.
Use correct function name in warning messages.
* gdk/win32/gdkevents-win32.c: Use correct parameter lists for the
GSourceFuncs gdk_event_prepare and gdk_event_check.
(gdk_event_get_graphics_expose): Do implement, use
PeekMessage. Thanks to Hans Breuer.
(event_mask_string): Debugging function to print an GdkEventMask.
(gdk_pointer_grab): Use it.
* gdk/win32/gdkfont-win32.c: The Unicode subrange that the
(old) book I used claimed was Hangul actually is CJK Unified
Ideographs Extension A. Also, Hangul Syllables were missing.
Improve logging.
* gdk/win32/gdkgc-win32.c: Largish changes.
* gdk/win32/gdkim-win32.c (gdk_set_locale): Use
g_win32_getlocale() from GLib, and not setlocale() to get current
locale name.
* gdk/win32/gdkprivate-win32.h
* gdk/win32/gdkwin32.h: Move stuff from gdkprivate-win32.h to
gdkwin32.h, similarily as in the X11 backend.
* gdk/win32/gdkwindow-win32.c (gdk_propagate_shapes): Bugfix,
assignment was used instead of equals in if test. Thanks to Hans
Breuer.
* gdk/win32/makefile.{cygwin,msc}
* gtk/makefile.{cygwin,msc}: Updates. Better kludge to get the
path to the Win32 headers that works also with the mingw compiler.
* gtk/gtkstyle.c: Include <string.h>.
2000-03-07 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkmain-win32.c: Internal GDK error reporting changes:
(gdk_win32_gdi_failed) New function for reporting errors from GDI,
for which it is no use to call GetLastError on Win9x.
(gdk_other_api_failed) New function, for
general error reporting without calling GetLastError.
(gdk_win32_api_failed) OTOH, this function always calls
GetLastError. (gdk_win32_last_error_string) Remove this function,
GLib has the equivalent now.
* gdk/win32/gdkprivate-win32.h: Declare the above, and macros to
call them with function name, file name and line number in the
arguments.
* gdk/win32/*.c: Use the new macros for reporting errors from GDI
functions.
* gtk/gtk.def: Add some missing entry points.
* gtk/gtkcompat.h.in: Define gtk_paned_gutter_size and
gtk_paned_set_gutter_size, which don't exist any longer, as void.
Fixes by Hans Breuer:
* gdk/makefile.msc: Update for debugging.
* gdk/win32/gdkgc-win32.c (gdk_win32_gc_set_dashes): Fake
implementation, just use the PS_DASH, PS_DOT, PS_DASHDOT or
PS_DASHDOTDOT or PS_DASH styles depending on the number of entries
in the dash_list.
* gdk/win32/gdkwin32.h (GDK_FONT_XFONT): New macro, similar as in
the x11 backend.
* gdk/win32/gdkprivate-win32.h (IS_WIN_NT): New macro.
* gtk/testgtk.c: Include config.h, guard inclusion of unistd.h.
2000-03-04 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkprivate-win32.h (struct _GdkWindowWin32Data): Store
just pixel value of background colour.
* gdk/win32/gdkgc-win32.c (gdk_colormap_color): New function
containing code snippet previously duplicated in a couple of
places.
* gdk/win32/gdkdrawable-win32.c (gdk_win32_draw_points): Use
gdk_colormap_color.
* gdk/win32/gdkevents-win32.c (gdk_event_translate): Generate
GDK_SCROLL events for WM_MOUSEWHEEL messages. Don't generate
events for autorepeated Shift, Control and Alt keys. Use
gdk_colormap_color.
* gdk/win32/gdkwindow-win32.c (gdk_window_set_cursor): If the
pointer is inside the window the cursor of which we are setting,
call SetCursor immediately.
* gdk/win32/makefile.cygwin
* gtk/makefile.cygwin: If we don't have the build number stamp
file, use zero.
* gtk/gtkcompat.h.in: Define gtk_paned_gutter_size and
gtk_paned_set_gutter_size as empty.
* gtk/gtk.def: Remove gtk_paned_set_gutter_size.
2000-02-14 Tor Lillqvist <tml@iki.fi>
* gdk/gdkdraw.c (gdk_draw_segments): Move the test for no segments
before the assertion for non-NULL segment list.
* gdk/win32/gdkgc-win32.c: (gdk_win32_gc_values_to_win32values):
Seems that pattern brushes *must* be 8x8 pixels! At least on my
machine, but it might be display driver dependent. Sigh, so make
sure the stipple is that size. Does Windows suck or what?
* gdk/win32/gdkdrawable-win32.c: Improve error handling in a few
places.
2000-02-13 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkgc-win32.c (gdk_gc_predraw): Add an extra parameter
so that we don't have to do unnecessary settings to the HDC.
* gdk/win32/gdkdrawable-win32.c
* gdk/win32/gdkimage-win32.c: Corresponding changes to the calls of
gdk_gc_predraw.
* gdk/win32/gdkdrawable-win32.c: Special-casing for GdkGCs where
the fill style is GDK_OPAQUE_STIPPLED. In those cases we construct
GDI paths, call WidenPath to get the outline of the stroken path,
and then fill the outline (with the brush that was built from the
stipple).
* gdk/win32/gdkgc-win32.c: Factor out common code from
_gdk_win32_gc_new and gdk_win32_gc_set_values into
gdk_win32_gc_values_to_win32values.
* gdk/win32/gdkmain-win32.c (gdk_win32_last_error_string,
gdk_win32_api_failed): New functions for error logging.
* gdk/win32/gdkprivate-win32.h: Declare them, and a macro
WIN32_API_FAILED to call them, passing function or file name and
line number.
* gdk/win32/*.c: Use the WIN32_API_FAILED macro.
* gdk/win32/gdkprivate-win32.h: Store just the pixel values from
GdkColor for foreground and background in GdkGCWin32Data.
* gdk/makefile.cygwin: Link in the resource object separately.
* gdk/win32/rc/gdk.rc
* gtk/gtk.rc (New file)
* gdk/win32/makefile.cygwin
* gtk/makefile.cygwin: Update build number in DLLs automatically,
as in GLib.
1999-11-21 Tor Lillqvist <tml@iki.fi>
* gdk/gdkconfig.h: Remove, as this is a file generated by configure
(or copied from gdkconfig.h.win32 on Win32, sigh).
* gdk/gdkprivate.h: Mark gdk_parent_root for export from DLL.
* gdk/gdkgc.c (gdk_gc_alloc): Remove duplicated line.
1999-11-08 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdraw.c
* gdk/win32/gdkfont.c
* gdk/win32/gdkgc.c: Various minor cleanups and bugfixes
following yesterday's changes. (Unrelated to Owen's reorg.)
1999-11-07 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkprivate.h: New font private structures, related to
fontsets.
* gdk/win32/gdkfont.c: New functions gdk_font_list_new() and
gdk_font_list_free(). On X11, will just be wrappers to
XListFonts() and XFreeFontNames(). On Win32, the code previously
in gtkfontsel.c is now here.
New function gdk_font_xlfd_create(). On X11 will get the FONT
property of the font (for GDK_FONT_FONTs), or call
XBaseFontNameListOfFontSet (for GDK_FONT_FONTSETs), on Win32
builds a XLFD style name from the font information in the LOGFONT
struct(s).
New function gdk_font_xlfd_free(), which correspondingly frees the
string returned by gdk_font_xlfd_create().
Implement fontsets on Win32. Add a function that iterates over a
wide char string and calls a callback function for each substring
of wide chars from the same Unicode subrange (and thus probably
available in the same real font).
Improve the XLFD emulation a bit.
* gdk/win32/gdkim.c (gdk_nmbstowchar_ts): Small bugfix.
* gdk/win32/gdkevents.c: Workaround for suspected bug on Win2k
Beta3, WM_IME_CHAR messages don't seem to contain the composed
multi-byte char as with the Active IMM on Win9x. Oh well, handle
WM_IME_COMPOSITION with GCS_RESULTSTR instead, use
ImmGetCompositionStringW() to get the composed Unicode chars.
* gdk/win32/gdkgc.c
* gdk/win32/gdkdraw.c: Changes needed because of the font private
struct changes.
* gdk/win32/gdk.def: Add the new functions.
1999-10-31 Tor Lillqvist <tml@iki.fi>
* gdk/gdkkeysyms.h: Add new keysyms from X11R6.4 (including
EuroSign).
* gdk/gdktypes.h: Add note about wchar_t not necessarily being the
same type as GdkWChar, especially on Win32.
* gdk/win32/*.c: Change gdk_root_parent to be a pointer.
* gdk/win32/*.c: Assume all strings are UTF-8. Convert to Unicode
before passing to Windows GDI for drawing etc. Convert to the
system default codepage before passing to Windows as window
titles.
* gdk/win32/gdkprivate.h: Add more fields to GdkWindowPrivate to
support changing input locale on the fly.
* gdk/win32/gdkevent.c: Support input language (keyboard locale)
on-the-fly changes. Convert incoming characters from the current
codepage to Unicode (and then to a UTF-8 multi-byte string) based
on the current input language. Use keysym<->Unicode mapping tables
and functions borrowed from xterm sources.
Support IMEs (Input Method Editors) for CJK languages. On non-CJK
editions of Win9x, use the ActiveX-based Active IMM (Input Method
Manager) if available. IMEs and the Active IMM are available under
the disguise of Chinese, Korean and Japanese support for IE and
Outlook Express from "Windows Update" for Win98. On Win2k, the CJK
support is present in all editions (as long as you install it).
Call DispatchMessage from gdk_events_queue() (and thus
gdk_WindowProc()), instead of duplicating the code in
gdk_WindowProc().
Reworked the grab handling and propagation code, factored out
duplicated code snippets into separate functions. Other cleanups,
too.
* gdk/win32/surrogate-dimm.h: Provide just the bits we need from
the <dimm.h> header describing the Active IMM.
* gdk/win32/gdkfont.c: Pretend to support fontsets, but so far
just do the same as for "single" fonts.
* gdk/win32/gdk.c: Call CoInitialize() (COM initialisation) from
gdk_init_check, and CoUninitialize() from gdk_exit_func. Handle
the new keysyms from gdkkeysyms.h.
* gtk/gtkfontsel.c (Win32): Load the font for the preview as a
fontset, so that gtkentry uses wide characters.
* gtk/gtkrc.c (Win32): Get the locale with gwin_getlocale(). Call
GTk+'s system directory "gtk+", not "gtk".
1999-10-21 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkprivate.h: Add more font private data.
* gdk/win32/gdkfont.c
* gdk/win32/gdkdraw.c: Revamped handling of multi-byte charset
fonts and strings. Now works much better. You still have to
have a correct font selected, though. No fontset emulation yet.
1999-10-14 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdraw.c (gdk_draw_text_wc): Don't use TextOutW for
GDK_FONT_FONT fonts (which is all we have for now, we don't
emulate fontsets). The X11 version uses plain XDrawString in that
case, too. The string passed to gdk_draw_text_wc seems to be in
fact (at least, when used by gtkentry and gtktext) either in a
single-byte charset, or a DBCS. Not Unicode.
This fixes the problem in gtkfontsel, where even if you had
selected a font with a non-Latin1 charset (windows-greek, for
instance), the preview still used Latin-1 glyphs.
* gdk/win32/gdkfont.c (gdk_text_width_wc): Similar change. Don't
use GetTextExtentPoint32W, use GetTextExtentPoint32A.
(gdk_font_load): Recognize the demibold etc weights, even if we
don't have the corresponding constants in the headers.
(gdk_font_hash_insert): Use same hash mechanism as in the X11
version. Should save font resources a bit, when we don't have
multiple HFONTs for the same font.
* gdk/win32/gdkprivate.h: Add the names field as in the X11
version.
1999-10-05 Tor Lillqvist <tml@iki.fi>
* gdk/win32/*.[ch]: Corresponding changes as in X11 backend.
* gdk/win32/gdkcompat.c: New file, actually provide an
implementation for the deprecated functions. (Just temporarily.)
* gtk/gtkfilesel.c: Fix an #ifdef syntax botch.
* gtk/makefile.{cygwin,msc}: Update gdk_headers.
* gdk/win32/gdk.def gtk/gtk.def: Updates.
1999-10-03 Tor Lillqvist <tml@iki.fi>
* gdk/gdkimage.h gdk/gdkpixmap.h: Change GDK_WINDOWING_WIN32 usage
to #ifdef also here.
* gdk/win32/*.h gdk/win32/*.c: Make corresponding changes as those
Owen did to the X11 backend.
* gdk/win32/gdkdraw.c (gdk_draw_pixmap): Fix it again, don't use
ScrollWindowEx when blitting inside a window, it can't be correct
in the general case.
* gdk/win32/gdkevents.c: Don't handle WM_SIZING, handling
WM_GETMINMAXINFO is easier.
* gdk/win32/gdkimage.c (gdk_image_new): Create new image with
depth equal to the bitspixel value, not the visual's depth.
* gdk/win32/gdkvisual.c (gdk_visual_init): Set the visual's depth
to 24 even if the bitspixel value is 32.
* gdk/gdkrgb.c (gdk_rgb_select_conv): After the above change, no
need to check for depth==32 when bpp==32, depth will always be 24.
* gdk/win32/gdkevents.c (gdk_event_translate): Fix from Simon
Kelley: Set expose_count in GdkEventExposes correctly.
* gdk/win32/gdkwindow.c: Remove dead code (#ifdef
MULTIPLE_WINDOW_CLASSES).
* gdk/win32/gdkdraw.c (gdk_draw_line): Workaround from Hans Breuer
for bug in NT, apparently NT *does* draw the end pixel, too, in
LineTo with a one-pixel pen, so we don't have to do it ourselves.
* gdk/win32/gdkdraw.c (gdk_draw_pixmap): When blitting inside one
window, use ScrollWindowEx, and call UpdateWindow. This prevents
bugs when for instance part of the window was outside the
display. Thanks to Philippe Colantoni for finding and fixing this.
* gdk/win32/gdkdraw.c (gdk_draw_arc): Fix start and end radial
endpoint calculations which were totally wrong. (A little RTFMing
helps a lot ;-)
* gtk/makefile.{cygwin,msc}: Use libintl extracted from glibc
from a separate directory, not from gettext, because of licensing
issues (we want to use the LGPL version).
* README.win32: Mention the intl from glibc vs from gettext issue.
* gdk/win32/gdkdraw.c (gdk_draw_pixmap): Less logging verbiage.
* gdk/win32/gdkevents.c: Fix long-standing bug in key
events. The key.string wasn't zero-terminated, still we strdup'ed
it in gdk_event_copy(). Synthesize crossing events for button
events before possible propagation.
* gdk/win32/gdkwindow.c: Log gdk_window_set_title.
* gdk/win32/makefile.cygwin
* gtk/makefile.cygwin: Use new DLL naming style for the
GCC-compiled ones.
* gdk/win32/makefile.msc
* gtk/makefile.msc: Cosmetics mostly.
* gtk/gtk.def: Add missing entry points.
* gtk/gtksocket.c: Add dummy gtk_socket_get_type() for Win32.
* gdk/win32/gdkdraw.c (gdk_draw_arc): Don't draw anything if width
or height is zero. Don't print a warning if Pie or Arc fails, they
always fail (?) for very narrow ellipses.
* gdk/win32/gdkdraw.c (gdk_draw_pixmap): Call InvalidateRgn for
the part or the destination window corresponding to source area
outside of the source drawable's boundary.
* gdk/win32/gdkdraw.c (gdk_draw_lines, gdk_draw_polygon): Don't do
anything if less than two points.
* gdk/win32/gdkselection.c (gdk_selection_owner_get): Always
return NULL. Gtk cut-and-paste inside a single program works
better this way. (It always gets the clipboard contents from
Windows, not from its own copy, which is cleared anyway. I can't
say I fully understand what happens... Emulating the X selection
and property stuff is a bit of a mess.)
* gdk/win32/gdkevents.c
* gdk/win32/gdkproperty.c: A bít more verbose logging.
* gdk/win32/gdkregion.c: Fix some memory leaks (temporary regions
that never got deleted). Revamp gdk_region_shrink.
* gdk/win32/gdkregion.c: Fix memory leak, delete temporary regions
after use.
* gtk/gtk.def: Add some missing entry points.
* gtk/gtkrc.c: Strip trailing directory separator from pixmap path
component.
directory hierarchy like this, with common gdk.h and gdktypes.h
for all backends. gdkrectangles.c and gdkrgb.c is copied before
compilation from the x11 backed (which still is in .., but
eventually will be in ..\x11).
