2006-06-03 Michael Natterer <mitch@gimp.org>
Make better use of the available 32x32 pixels so the cursors look
less crowded:
* cursors/cursor-corner-*.png
* cursors/cursor-side-*.png
* cursors/xbm/cursor-corner-*.xbm
* cursors/xbm/cursor-side-*.xbm: made the crosshair lines one
pixel longer.
* cursors/modifier-join.png
* cursors/modifier-pattern.png
* cursors/xbm/modifier-join-mask.xbm
* cursors/xbm/modifier-join.xbm
* cursors/xbm/modifier-pattern-mask.xbm
* cursors/xbm/modifier-pattern.xbm: moved one pixel to the right.
* cursors/tool-*.png
* cursors/xbm/tool-*.png: moved closer to the bottom right corner.
* cursors/gimp-tool-cursors.xcf: ditto.
2006-06-03 Michael Natterer <mitch@gimp.org>
* app/tools/gimpnewrectselecttool.c
(gimp_new_rect_select_tool_button_press): check for function ==
INACTIVE, not CREATING || EXECUTING when checking whether to start
moving.
* app/tools/gimpselectiontool.c (gimp_selection_tool_oper_update):
changed modifiers so alt+ctrl does cut+float+move and alt+shift
does copy+float+move.
* app/tools/gimpnewrectselecttool.[ch]: use RECT_INACTIVE state,
(but doesn't seem to do anything). More importantly, change
logic so that button_release executes the tool, and button
release undo's if appropriate.
* app/tools/gimprectangletool.c (gimp_rectangle_tool_button_press):
if function is RECT_CREATING, start a new rectangle.
* app/tools/gimpnewrectselecttool.c: remove g_printerr statement that
mitch added for debugging.
2006-06-02 Michael Natterer <mitch@gimp.org>
* cursors/Makefile.am
* cursors/cursor-corner-bottom-left.png
* cursors/cursor-corner-bottom-right.png
* cursors/cursor-corner-top-left.png
* cursors/cursor-corner-top-right.png
* cursors/cursor-side-bottom.png
* cursors/cursor-side-left.png
* cursors/cursor-side-right.png
* cursors/cursor-side-top.png
* cursors/xbm/cursor-corner-bottom-left-mask.xbm
* cursors/xbm/cursor-corner-bottom-left.xbm
* cursors/xbm/cursor-corner-bottom-right-mask.xbm
* cursors/xbm/cursor-corner-bottom-right.xbm
* cursors/xbm/cursor-corner-top-left-mask.xbm
* cursors/xbm/cursor-corner-top-left.xbm
* cursors/xbm/cursor-corner-top-right-mask.xbm
* cursors/xbm/cursor-corner-top-right.xbm
* cursors/xbm/cursor-side-bottom-mask.xbm
* cursors/xbm/cursor-side-bottom.xbm
* cursors/xbm/cursor-side-left-mask.xbm
* cursors/xbm/cursor-side-left.xbm
* cursors/xbm/cursor-side-right-mask.xbm
* cursors/xbm/cursor-side-right.xbm
* cursors/xbm/cursor-side-top-mask.xbm
* cursors/xbm/cursor-side-top.xbm: new cursors for edge and corner
resizing. They perfectly align with the small crosshair and can be
used together with tool cursors and cursor modifiers.
* cursors/gimp-tool-cursors.xcf: add them here too.
* app/widgets/widgets-enums.h: add them to the GimpCursorType enum.
* app/widgets/gimpcursor.c: add them here too.
* app/display/gimpdisplayshell-cursor.c: treat them like the small
crosshair (don't replace them by the small crosshair but use them
as-is). Also allow the bad modifier with the large crosshair.
* app/tools/gimprectangletool.c
(gimp_rectangle_tool_cursor_update): use the new cursors. Don't
call gimp_tool_set_cursor() here.
(gimp_rectangle_tool_response): reset "function" to RECT_CREATING
when resetting the tool.
* app/tools/gimpselectiontool.[ch] (struct GimpSelectionTool):
added boolean member "allow_move" which defalts to TRUE.
(gimp_selection_tool_oper_update): don't move masks, floating
selections or anything when "allow_move" is FALSE. Changed
behavior of click inside a selection to simply create a new
selection, need to press alt+shift now to drag-float the
selection. Please test this, it's apretty fundamental change!
(gimp_selection_tool_cursor_update): use the tool's configured
cursor instead of always GIMP_CURSOR_MOUSE, so this function can
be called after gimp_rectangle_tool_cursor_update() to add the
plus, minus etc. modifiers.
* app/tools/gimpnewrectselecttool.c: implement
GimpTool::cursor_update() and call
gimp_rectangle_tool_cursor_update() from there. Chain up to get
the plus, minus etc. modifiers added.
Re-enble selection moving:
(gimp_new_rect_select_tool_oper_update): set GimpSelectionTool's
"allow_move" to FALSE unless the rectangle tool is in an idle
state.
(gimp_new_rect_select_tool_button_press): allow a selection moving
to be started if the rectangle tool is idle. Fall back to starting
a rect select if gimp_selection_tool_start_edit() returned FALSE.
2006-06-02 Sven Neumann <sven@gimp.org>
* plug-ins/common/mkgen.pl: always look at $plugins{$_}->{cflags}.
* plug-ins/common/plugin-defs.pl: use MMX_EXTRA_CFLAGS for the
selective gaussian blur plug-in.
* plug-ins/common/Makefile.am: regenerated.
2006-06-02 Sven Neumann <sven@gimp.org>
* libgimpbase/gimpparasite.[ch]: hide the GimpParamSpecParasite
struct like we do with other custom param specs.
2006-06-02 Sven Neumann <sven@gimp.org>
* app/composite/gimp-composite.c (gimp_composite_use_cpu_accel):
need to test for GIMP_COMPOSITE_OPTION_NOEXTENSIONS.
* libgimp/gimp.c (gimp_config): call gimp_set_use_cpu_accel() from
here, not in gimp_main().
* plug-ins/common/sel_gauss.c: applied patch from Loren Merritt
that adds MMX code to boost the plug-in speed (bug #342860).
2006-06-02 Sven Neumann <sven@gimp.org>
Moved the CPU detection code to libgimpbase (see bug #342860):
* app/base/Makefile.am
* app/base/cpu-accel.[ch]
* app/base/test-cpu-accel.c: removed here...
* libgimpbase/Makefile.am
* libgimpbase/gimpbase.h
* libgimpbase/gimpcpuaccel.[ch]
* libgimpbase/test-cpu-accel.c: ... and added here again with
some API changes.
* app/composite/Makefile.am
* app/composite/make-installer.py: changed accordingly.
* app/composite/gimp-composite-*-installer.c: regenerated.
* libgimp/gimp.c (gimp_main): call gimp_set_use_cpu_accel().
2006-06-01 Kevin Cozens <kcozens@cvs.gnome.org>
* scripts/script-fu-compat.init: Fixed the definition of fmod for
real this time. Previous commit fixed the definition of truncate.
2006-06-01 Michael Natterer <mitch@gimp.org>
* cursors/Makefile.am
* cursors/modifier-bad.png
* cursors/xbm/modifier-bad-mask.xbm
* cursors/xbm/modifier-bad.xbm: new "bad" cursor
modifier. Replaces the "bad" cursor.
* cursors/gimp-tool-cursors.xcf: added it here too.
* app/widgets/widgets-enums.h: added GIMP_CURSOR_MODIFIER_BAD.
* app/widgets/gimpcursor.c: add the bad modifier. Leave the bad
cursor there for now.
* app/display/gimpdisplayshell-callbacks.c
* app/tools/gimpaligntool.c
* app/tools/gimpblendtool.c
* app/tools/gimpbycolorselecttool.c
* app/tools/gimpclonetool.c
* app/tools/gimpcolortool.c
* app/tools/gimpfliptool.c
* app/tools/gimpiscissorstool.c
* app/tools/gimpmovetool.c
* app/tools/gimptransformtool.c
* app/tools/gimpvectortool.c: use the modifier instead of the
cursor. Fixes hotspot jumping when switching between normal and
bad cursors. The changed cursor_update() functions even make more
sense IMHO. Fixes bug #158407.
2006-06-01 Sven Neumann <sven@gimp.org>
Added basic framework for plug-ins to access the use_cpu_accel
configuration (bug #342860):
* app/composite/gimp-composite.[ch]: added new function
gimp_composite_use_cpu_accel().
* libgimpbase/gimpprotocol.[ch]: added use_cpu_accel to the config
message.
* app/plug-in/gimppluginmanager-call.c: pass the return value of
gimp_composite_use_cpu_accel() for config.use_cpu_accel.
* libgimp/gimp.[ch]: make the config value accessible by means of
a new function gimp_use_cpu_accel().
* libgimp/gimp.def: updated.
2006-06-01 Sven Neumann <sven@gimp.org>
* plug-ins/common/sel_gauss.c: applied patch from Loren Merritt
that replaces the floating-point implementation of selective
gaussian blur with a fixed-point version (bug #342860).
* app/tools/gimprectangletool.[ch]: add "constrain" property
to specify whether to clip at image bounds when computing
dimensions.
* app/tools/gimpcroptool.c
* app/tools/gimpnewrectselecttool.c: set "constrain" to TRUE.
* app/tools/gimpellipseselecttool.c: set "constrain" to FALSE.
Hopefully fixes bug #329817.
2006-05-31 Michael Natterer <mitch@gimp.org>
* libgimpwidgets/gimpcolorscale.c: removed #define SHADOW 1, it's
not a constant value. Use the GtkRange::trough-border style
property instead.
2006-05-30 Raphael Quinet <raphael@gimp.org>
* plug-ins/metadata/xmp-parse.c: protect against broken XMP blocks
using rdf:Alt instead of rdf:Seq. Fixes bug #343315.
* plug-ins/metadata/xmpdump.c (main): call g_set_prgname() to get
better messages from glib.
* data/images/gimp-splash.png: new splash screen, waiting for the
old animation feature to be restored.
2006-05-30 Michael Natterer <mitch@gimp.org>
Honor active components when pasting. Fixes bug #150845:
* app/core/gimplayer-floating-sel.c (floating_sel_composite):
don't temporarily set all image components to active while
compositing.
* app/core/gimpimage.c (gimp_image_set_component_active):
relax()/rigor() the floating selection around setting the "active"
flag. Also make sure the projection is updated correctly.
2006-05-30 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpdataeditor.c (gimp_data_editor_name_activate)
* app/widgets/gimpdatafactoryview.c
(gimp_data_factory_view_tree_name_edited): strip the newly
entered name from whitespace and reject empty names.
2006-05-30 Manish Singh <yosh@gimp.org>
* app/batch.c (batch_run): GIMP_BATCH_INTERPRETER can be set in
the environment to override the default batch interpreter procedure.
* plug-ins/pygimp/plug-ins/happy-valley-relief.py: basic python
code evaluator function.
* plug-ins/pygimp/plug-ins/Makefile.am: add above file.
2006-05-30 Manish Singh <yosh@gimp.org>
* app/file/gimprecentlist.c: #define _GNU_SOURCE instead of
_SVID_SOURCE, so we get all the declarations we need. Fixes
bug #342390.
2006-05-29 Sven Neumann <sven@gimp.org>
* configure.in: use PANGO_DISABLE_DEPRECATED for pango < 0.13.0.
* app/text/gimptext-vectors.c: added const qualifiers to the
FT_Vector parameters of the FT_Outline_Funcs.
2006-05-29 Sven Neumann <sven@gimp.org>
* app/widgets/gimpdialogfactory.c
(gimp_dialog_factory_dialog_new_internal): code cleanup; only call
gtk_window_present() if called with present == TRUE.
2006-05-29 Sven Neumann <sven@gimp.org>
* app/core/gimpbrushgenerated.c (gimp_brush_generated_load): don't
choke on brush files with empty names (bug #343140).
(gimp_brush_generated_save): warn about brushes with empty names.
2006-05-29 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpdock.c (gimp_dock_key_press_event): make sure
text widgets get all key events first. Fixes bug #301006.
2006-05-28 Michael Natterer <mitch@gimp.org>
* plug-ins/common/pnm.c: applied patch from Martin Collins which
adds PBM (bitmap) support. Fixes bug #167578. Did some additional
cleanups and sprinkled some gimp_progress_update(1.0).
2006-05-28 Michael Natterer <mitch@gimp.org>
Applied patch from David Gowers which adds actions to select
palette and colormap colors with actions. Modified the patch quite
a bit. Fixes bug #130123.
* app/widgets/gimpcolormapeditor.[ch]
* app/widgets/gimppaletteeditor.[ch]: add functions get_index()
which gets the currently selected color's index (optionally the
index of a passed color), set_index() which sets the selected
color by index, and max_index() which returns the maximum possible
color index.
* app/dialogs/dialogs-constructors.c: changed accordingly.
* app/actions/context-actions.c
* app/actions/context-commands.[ch]: actions and callbacks which
use the new functions.
2006-05-28 Michael Natterer <mitch@gimp.org>
* app/core/gimppalette-import.[ch]: added support for extracting
colors from the selected pixels only.
* app/dialogs/palette-import-dialog.c: added "Sample merged" and
"Selected Pixels only" toggles. Fixes bug #316212. Cleaned up the
code quite a bit.
2006-05-28 Michael Natterer <mitch@gimp.org>
* app/tools/gimpcolorpickertool.c (gimp_color_picker_tool_picked):
removed all code except info dialog updating and chain up instead.
* app/tools/gimpcolortool.c (gimp_color_tool_real_picked): newly
added default implementation. Also updates the colormap dialog and
picks into a palette.
* app/tools/gimppainttool.c (gimp_paint_tool_color_picked):
removed. The default impl. does this and much more now. Hopefully
fixes bug #320660.
* plug-ins/script-fu/scripts/reverse-layers.scm: if any layers
are opaque, make them transparent when reversing. Current CVS
allows moving opaque layers above the bottom, but that might change.
Michael Natterer