2000-04-06 Michael Natterer <mitch@gimp.org>
* app/commands.c
* app/layers_dialog.c
* app/resize.[ch]: behave like a widget:
removed resize_widget_free() and call
gtk_widget_destroy(resize->resize_shell) instead.
* app/interface.c: s/0/FALSE/
* app/color_area.c
* libgimp/gimpchainbutton.c: show the correct background also for
engine driven themes. The chainbutton's lines still don't draw
with the pixmap theme, which is probably a bug of the theme's
draw_polygon method.
* app/gimage_cmds.c
* tools/pdbgen/pdb/gimage.pdb: (gimp_image_set_resolution) guard
against NaN and infinity. There are pics out there using these
values and it seems that NaN > GIMP_MIN_RESOLUTION. Fixes#8476.
--Sven
* app/app_procs.c: if we cannot load the font we'd like to use,
use the gtk+ default font. Fixes bug #8359.
* app/about_dialog.c
* app/install.c: properly ref/unref fonts
* app/text_tool.[ch]: code cleanup (do not rely on TRUE being 1)
* app/tips_dialog.c: code cleanup and less resizing
--Sven
* app/app_procs.c: if we cannot load the font we'd like to use,
use the gtk+ default font. Fixes bug #8359.
* app/about_dialog.c
* app/install.c: properly ref/unref fonts
* app/text_tool.[ch]: code cleanup (do not rely on TRUE being 1)
* app/tips_dialog.c: code cleanup and less resizing
--Sven
2000-04-05 Michael Natterer <mitch@gimp.org>
* app/layers_dialog.c: oops, yesterday's "cleanups" contained an
unreverted hack I've tried which caused a SEGV with floating
selections. Fixed now.
2000-04-03 Michael Natterer <mitch@gimp.org>
* app/color_panel.[ch]
* app/color_picker.c
* app/qmask.c: removed the public function color_panel_free() and
fake a real widget's behaviour by connecting to the panel widget's
"destroy" signal.
* app/channels_dialog.c
* app/layers_dialog.c: cleaned up and sync'ed the code where
possible (without changing the logic).
2000-04-03 Michael Natterer <mitch@gimp.org>
* app/channels_dialog.c
* app/layers_dialog.c: when dropping layers/channels around in the
stack, Gtk sometimes continues searching a drag destination
_after_ the drop callback has been called (which is probably a
bug), causing a SEGV because the callback changed the GtkList
containing the layer/channel widgets.
The safe way is to reorder the stack in an idle function which
should also be robust against any future Gdk/X dnd protocol
change.
For the same reason, I added an idle function which lets the
active layer/channel widget grab the focus because in some dnd
cases the row holding the focus was not the active one (which is
otherwise impossible with GTK_SELECTION_BROWSE).
This commit should fix the remaining problems with layer/channel
dnd as well as all cases where the highlighted layer was not the
active one. Please report if it still doesn't work for you.
2000-04-03 Michael Natterer <mitch@gimp.org>
* app/gimpdnd.[ch]
* app/channels_dialog.c
* app/layers_dialog.c: removed the GdkGC parameter from
gimp_dnd_set_drawable_preview_icon() since it's no longer used.
* app/gimpimage.c: dirty the view correctly after applying the
layer mask. Don't gdisplays_flush() because it's the job of the
caller.
* app/layers_dialog.c: when applying a layer mask, flush either
the display or just the layer_widget, depending on the previous
visibility state of the mask. Cleanups.
* plug-ins/print/gimp_color_window.c
* plug-ins/print/gimp_main_window.c: including config.h re-enables
i18n.
german translation. Hopefully the translation is now good enough
that it won't ihave to change much anymore, since the new
"Gimp kurz und gut" reference for 1.2 will go into print soon...
--Sven
2000-04-02 Garry R. Osgood <gosgood@idt.net>
* gimp/app/colormap_dialog.i.c
As Mr. Lamb, report originator noted, an easy
fix. in ipal_set_image() line 688
Reorganized the initialization of the
indexed palette object so as not to run afoul
of gasserts() inhabiting ipal_draw() and client
routines. These entailed initialing ipal->col_index,
ipal->dnd_col_index prior to call to ipal_draw()
* app/bezier_select.c
Curves/Path tool
Fixed propblem with deleting points. You can now delete the first
and last point on any open curve (as well as mid-points).
Also fixed some problems where some points would leave the markers
on screen after they had been deleted.
Note you have always been able to delete whole curves by pressing
the "shift" key when over a point to be deleted in "remove mode".
2000-04-02 Michael Natterer <mitch@gimp.org>
* app/interface.c: completely overworked the widget packing of the
display shell. This was necessary because the table insisted on
giving the vertical scrollbar more allocation than it needed when
first displaying an image which is smaller than the minimal
statusbar width. Replaced the tables with packing boxes.
See the comments I've added to create_display_shell() if you
really want to see the ugly details.
Also, to get rid of the 2-pixel spacing which appears at the
bottom of the display whenever the statusarea is hidden,
gdisp->statusarea is now the eventbox, not the hbox it contains.
* pixmaps/navbutton.xpm: made it 2x2 pixels larger.
2000-04-01 Garry R. Osgood <gosgood@idt.net>
* gimp/app/bezier_select.c
No fooling, #6903 was not that hard to close;
in bezier_edit_point_on_curve(),when point
deletion reduces a curve below the minimum
with which the implementation can cope (2
anchors, four controls) we put it out of its
misery with an invocation of delete_whole_curve().
2000-04-01 Michael Natterer <mitch@gimp.org>
* app/disp_callbacks.c: when dropping a drawable, do the
undo_push_group_start() before creating the new layer, otherwise
the undo system will show obscure misbehaviour (the undo group is
pushed correctly but no undo will be possible until an other undo
is pushed).
* app/paint_funcs.c: Fixed the off by one error that was the cause
of bug #4980. Fixed a bug that caused images to have their
centers shifted slightly when resizing. Removed a couple of
unused variables.
2000-03-31 Michael Natterer <mitch@gimp.org>
* app/gimpdnd.[ch]: minor cleanups.
* libgimp/gimpwidgets.c: don't set a border width for the
radio_group's vbox if no frame around it was requested.
* plug-ins/common/sample_colorize.c: add a separator between
image and sample.
* plug-ins/common/iwarp.c: use the libgimp ui functions, removed
callbacks, a global s/gfloat/gdouble/, replaced the booleans
toggling the transform mode with a single variable containing an
enum value, indentation.
2000-03-29 Michael Natterer <mitch@gimp.org>
* app/bezier_select.h
* app/bezier_selectP.h
* app/by_color_select.[ch]
* app/ellipse_select.[ch]
* app/free_select.[ch]
* app/move.[ch]
* app/rect_select.[ch]: kindof selection tools code review:
- use SelectOps instead of int.
- removed some unused prototyped and callbacks.
- don't show the SELECTION_MOVE_MASK cursor if there is no
selection and don't try to move the mask in that case.
- re(?)-enabled moving the selection mask even if there is a
floating selection.
- usual bunch of cleanups.
2000-03-27 Michael Natterer <mitch@gimp.org>
* app/transform_core.c: the button_press and cursor_update
functions were still doing checks on the active layer instead of
the active drawable.
Fixing this automatically made the layers mask transformable.
2000-03-26 Garry R. Osgood <gosgood@idt.net>
* gimp/app/bezier_select.c
in bezier_add_point(), when a BEZIER_MOVE type
point is being added to a path with at least
one extant segment, the point is re-typed
to a BEZIER_ANCHOR (which is old behavior)
and designated as the current anchor
(which is new behaviour). With this change,
motion events move the (second) control point
and indicia are properly updated. This closes
#6225.
2000-03-26 Michael Natterer <mitch@gimp.org>
* app/gimage_mask.[ch] (gimage_mask_extract): add an "add_alpha"
parameter to allow selected regions to be extracted without having
an alpha channel added.
* app/global_edit.c: pass add_alpha = TRUE.
* app/transform_core.[ch]: made the transform core work on
non-layer drawables even if no selection is present. Fixes#7485
and #7555.
- transform_core_cut(): extract the mask without alpha if
operating on a non-layer without having a selection.
- transform_core_paste(): return a boolean indicating success
instead of a layer and handle channels correctly.
- transform_core_do(): if the "floating_tiles" passed to the
function are from an un-floated non-layer, treat the whole
non-layer as alpha channel and never enter the loop which
transforms the (not present) color channels.
Also clip the result to ensure that the channel never grows
larger then the image.
* app/tools_cmds.c
* tools/pdbgen/pdb/tools.pdb: transform_core_paste() returns a
gboolean now.
2000-03-25 Michael Natterer <mitch@gimp.org>
* app/cursorutil.[ch]: new global variable "gimp_busy" which gets
set/unset whenever busy cursors are added/removed.
* app/info_dialog.c: register the info dialogs with the dialog
handler.
* app/fuzzy_select.[ch]: cleanups.
Here starts the ugly workaround which simulates something like
locking. If it works, it will close lots of bugs, if not, it's
easy to remove again.
So far, I didn't find strange side effects but Gimp is told to be
a complex program :-) Please test this.
* app/context_manager.c: don't allow tool changes if gimp_busy
is TRUE.
* app/disp_callbacks.c: don't allow mouse and key events in the
display_canvas if gimp_busy is TRUE.
(except if the current tool is FUZZY_SELECT and it is ACTIVE,
which is very ugly)
Also block other stuff like dropping colors/patterns etc.
* app/gdisplay_ops.c: don't close any display while Gimp is
busy. This is not really what we want but at least it prevents
crashes.
2000-03-24 Michael Natterer <mitch@gimp.org>
* Makefile.am
* gimp1_0_splash.ppm: added
* gimp1_1_splash.ppm: removed
* gimp_splash.ppm: the image formerly known as gimp1_1_splash.ppm
* app/app_procs.c: load gimp_splash.ppm
The idea if shearing the data_dir between Gimp versions has
vanished, so this seems like the right thing to do (TM).
Keep the 1.0 splash around for sentimental reasons but don't
install it anymore.
* plug-ins/common/autocrop.c: indentation paranoia.
2000-03-24 Michael Natterer <mitch@gimp.org>
* app/plug_in.c (plug_in_proc_def_remove): pass FALSE to
plug_in_proc_def_destroy() so the proc_def structure gets actually
freed.
* app/fileops.c
* app/gdisplay.[ch]: when reverting an image, reconnect all
affected views to the reverted version. This fixes one of the
bugs Tigert pointed out at GUADEC.
* app/gimage_mask.[ch]
* app/flip_tool.c: cleanups
--Sven & Mitch
Michael Natterer