2006-05-13 Hans Breuer <hans@breuer.org>
* **/makefile.msc app/gimpcore.def : updated
* app/core/gimp-util.c : dont include "config/gimpbaseconfig.c", it
gives an redefinition error with msvc. Instead include
config/gimpbaseconfig.h and libgimpconfig/gimpconfig-path.h
* plug-ins/common/psd_save.c : fix c99isms (declarations only at the
start of a block)
2006-05-12 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpview.h: save 20 bytes per instance by using
single bits instead of 6 gbooleans.
* app/widgets/gimpview.c: some code cleanup.
* app/widgets/gimpviewrendererbrush.c: don't #include "gimpbrush.h".
* app/widgets/gimpviewrendererbuffer.c: #include "gimpviewable.h"
instead of "gimpbuffer.h".
* app/widgets/gimpviewrenderergradient.c
* app/widgets/gimpviewrendererimagefile.c
* app/widgets/gimpviewrendererimagefile.h: micro cosmetics.
2006-05-12 Michael Natterer <mitch@gimp.org>
* plug-ins/imagemap/imap_csim.y: fix warning about unused return
value of g_list_append().
* plug-ins/imagemap/imap_csim_parse.[ch]: regenerated (using
bison 2.1 instead of 2.0, lots of things changed, please test).
2006-05-12 Michael Natterer <mitch@gimp.org>
* configure.in: check for gmodule-no-export-2.0, not for
gmodule-2.0. Also removed gobject-2.0 from the gmodule check, i
have no idea why it was there.
* libgimpmodule/Makefile.am: s/GMODULE/GMODULE_NO_EXPORT/ because
we only want to import modules here, not export our own symbols.
(note that this change is actually useless, because libgimpmodule
exports its symbols anyway, but it's more correct this way).
2006-05-11 Michael Natterer <mitch@gimp.org>
* app/pdb/gimppdb-query.c (gimp_pdb_proc_info): restored old
version of this function, it was correct after all (spotted by
Kevin Cozens).
2006-05-11 Michael Natterer <mitch@gimp.org>
Applied modified patch from Michael J. Hammel which allows to
remove all keyboard shortcuts from the menus (fixes bug #331839):
* app/dialogs/preferences-dialog.c: added "Remove all keyboard
shortcuts" button to the "Interface" section.
* app/menus/menus.[ch]: added menus_remove() which does the
shortcut removal.
2006-05-10 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpviewrendererbrush.c
* app/widgets/gimpviewrendererbuffer.c: use
gimp_viewable_get_size() and get rid of useless
local "brush" and "buffer" variables.
2006-05-10 Michael Natterer <mitch@gimp.org>
* app/widgets/gimptoolbox.c: code cleanup, no logic changed.
* app/widgets/gimptoolbox-color-area.c: make the very first click
on the color area work as expected.
2006-05-09 Simon Budig <simon@gimp.org>
* themes/Default/images/stock-color-triangle-16.png: New icon
to reflect the corrected triangular color selector.
2006-05-09 Michael Natterer <mitch@gimp.org>
* app/core/gimpimage.c (gimp_image_new)
* app/core/gimpitem.c (gimp_item_configure)
* app/display/gimpdisplay.c (gimp_display_new): make sure IDs wrap
correctly at G_MAXINT and skip IDs which are currently in use.
It makes no sense to treat gimp->image_table and gimp->images
differently, since they keep the same set of images. And it makes
no sense to treat gimp->displays and gimp->images differently.
Moved all container adding into the objects' constructors:
* app/core/gimp.c (gimp_create_image): don't add the new image
to gimp->images here...
* app/core/gimpimage.c (gimp_image_constructor): ...but here.
* app/core/gimp-gui.c (gimp_create_display): don't add the new
display to gimp->displays here...
* app/display/gimpdisplay.c (gimp_display_new): ...but here.
2006-05-08 Simon Budig <simon@gimp.org>
* plug-ins/script-fu/siod/sliba.c: Insert a missing break that
was responsible for a weird output of integer arrays.
Untabbified.
2006-05-08 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpfiledialog.c (gimp_file_dialog_new): set the
alternative button order here...
* app/dialogs/file-open-dialog.c (file_open_dialog_new)
* app/dialogs/file-save-dialog.c (file_save_dialog_new): ...instead
of here.
2006-05-08 Simon Budig <simon@gimp.org>
* plug-ins/common/postscript.c: improve the error message when
spawning ghostscript does not work. See bug #340996. Untabbified.
2006-05-07 Simon Budig <simon@gimp.org>
* modules/colorsel_triangle.c: Make the hue angle behave
mathematically correct. Fixes bug #141922
* plug-ins/script-fu/scripts/circuit.scm: Fix multi line description.
* docs/Wilber.svg: Add xmlns attribute to the svg element, so
firefox renders it when reading it from disk.
2006-05-07 Michael Natterer <mitch@gimp.org>
* app/display/gimpdisplayshell-scale.c (gimp_display_shell_scale):
return early if the shell is already at the requested zoom level.
(paranoia addition for bug #165032, which was already fixed by the
fix for bug #164281):
2006-05-07 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpbrusheditor.c: hide the button bar, which is
useless for the brush editor. Fixes user confusion (bug #306704).
2006-05-06 Michael Natterer <mitch@gimp.org>
* app/tools/gimppainttool.c: connect to the context's
"brush-changed" signal and update the brush core's brush
accordingly. Fixes inconsistent brush preview when not moving
the mouse while the brush changes (bug #323404).
2006-05-06 Michael Natterer <mitch@gimp.org>
* app/core/gimpparamspecs.[ch]: added "gboolean none_ok" members
to GimpParamSpecImageID, GimpParamSpecItemID and
GimpParamSpecDisplayID. If none_ok is set, allow '0' and '-1' as
valid IDs. Added "none_ok" parameter to all ID param spec
constructors.
* app/xcf/xcf.c: require real image IDs, but allow images without
active drawable.
* app/pdb/gimp-pdb-compat.c: set none_ok to TRUE in all compat
param specs, so only IDs that are really random garbage don't pass
validation. Fixes validation errors with 0 and -1 IDs.
* tools/pdbgen/app.pl: set none_ok to FALSE by default. Added
support for $arg->{none_ok} to control it.
* tools/pdbgen/pdb/image.pdb
* tools/pdbgen/pdb/progress.pdb
* tools/pdbgen/pdb/text_tool.pdb: use none_ok instead of
no_success (which disables validation entirely).
* app/pdb/[many]_cmds.c: regenerated.
2006-05-06 Michael Natterer <mitch@gimp.org>
* app/core/gimpparamspecs.[ch]: added "gboolean none_ok" members
to GimpParamSpecImageID, GimpParamSpecItemID and
GimpParamSpecDisplayID. If none_ok is set, allow '0' and '-1' as
valid IDs. Added "none_ok" parameter to all ID param spec
constructors.
* app/pdb/gimp-pdb-compat.c: set none_ok to TRUE in all compat
param specs, so only IDs that are really random garbage don't pass
validation. Fixes validation errors with 0 and -1 IDs.
* tools/pdbgen/app.pl: set none_ok to FALSE by default. Added
support for $arg->{none_ok} to control it.
* tools/pdbgen/pdb/image.pdb
* tools/pdbgen/pdb/progress.pdb
* tools/pdbgen/pdb/text_tool.pdb: use none_ok instead of
no_success (which disables validation entirely).
* app/pdb/[many]_cmds.c: regenerated.
2006-05-05 Manish Singh <yosh@gimp.org>
* plug-ins/common/gifload.c (DoExtension): Revert back to unsigned
chars for the buffer, and only cast to gchar if we know we're dealing
with a comment. Fixes bug #339865.
2006-05-05 Michael Natterer <mitch@gimp.org>
* app/plug-in/gimppluginmanager.[ch]: added signals
"plug-in-opened" and "plug-in-closed". Added functions
gimp_plug_in_manager_add_open_plug_in() and _remove_open_plugin()
which maintain the list of open plug-ins and emit the signals.
* app/plug-in/gimpplugin.c (gimp_plug_in_open)
(gimp_plug_in_close): don't touch manager->open_plug_ins and don't
ref/unref the plug-in. Call above new functions instead. Don't
call gimp_pdb_dialogs_check().
* app/core/gimp-gui.[ch]
* app/gui/gui-vtable.c: removed gimp_pdb_dialogs_check().
* app/widgets/gimppdbdialog.[ch]: removed
gimp_pdb_dialogs_check_callback() and connect to the
plug-in-manager's "plug-in-closed" signal instead.
2006-05-04 Michael Natterer <mitch@gimp.org>
* app/plug-in/gimppluginshm.c: cleaned up a bit.
(gimp_plug_in_shm_new): return NULL if anything goes wrong,
instead of a GimpPlugInShm structure that contains no shm (we
don't need multiple cases of "there is no shm"),
* app/plug-in/gimppluginmanager.[ch]
(gimp_plug_in_manager_get_shm_ID)
(gimp_plug_in_manager_get_shm_addr): removed these functions.
* app/plug-in/gimppluginmanager-call.c (gimp_plug_in_manager_call_run)
* app/plug-in/gimpplugin-message.c (gimp_plug_in_handle_tile_req):
get the shm ID and addr directly from manager->shm if it exists,
use -1 and NULL otherwise.
Unrelated:
* app/plug-in/gimppluginmanager.c: move most stuff from
gimp_plug_in_manager_exit() to gimp_plug_in_manager_finalize().
Simplify plug-in killing in _exit().
2006-05-04 Sven Neumann <sven@gimp.org>
* app/core/gimp-user-install.[ch]: added a "verbose" parameter.
* app/app_procs.c (app_run): abort if the user installation fails.
2006-05-04 Michael Natterer <mitch@gimp.org>
* app/plug-in/gimpplugin.[ch] (struct GimpPlugIn): remove member
"name" and unused cruft "ref_count". Reordered other members a bit.
(gimp_plug_in_new): set GimpObject's name to
g_filename_display_basename(path) instead.
Use GimpObject's name instead of plug_in->name for error messages
and pass the full path to gimp_plug_in_debug_argv()
* app/plug-in/gimpplugindebug.c (gimp_plug_in_debug_argv): extract
the path's basename ourselves.
* app/plug-in/gimpplugin-message.c: GimpObject's name for
error messages.
2006-05-04 Michael Natterer <mitch@gimp.org>
* app/core/gimpparamspecs.c (gimp_string_array_new)
(gimp_param_string_array_validate): restore the strict checks
on string arrays.
* app/plug-in/gimppluginmanager-help-domain.c
(gimp_plug_in_manager_get_help_domains): return NULL string arrays
when there are 0 help domains to fix the inconsistency at its root.
2006-05-04 Michael Natterer <mitch@gimp.org>
* app/plug-in/gimpplugin.c
* app/plug-in/gimppluginmanager-call.c: use the right procedure
types in calls to gimp_plug_in_proc_frame_foo().
2006-05-03 Michael Natterer <mitch@gimp.org>
* app/widgets/gimpselectiondata.c
(gimp_selection_data_get_tool_info): port to using
gimp_selection_data_get_object(), it was simply forgotten.
Fixes tool dropping (bug #336402).
2006-05-03 Michael Natterer <mitch@gimp.org>
Made an object out of the plug-in struct. Also change refcounting
considerably:
- gimp_plug_in_open() adds a reference that is only dropped by
gimp_plug_in_close().
- temporarily ref the plug-in while handling messages.
- remporarily ref the plug-in while a recursive main loop is
running.
- each caller of gimp_plug_in_new() also unrefs the plug-in before
returning, the only reference that persists across functions
(that keeps the plug-in alive) is the one added by open().
* app/plug-in/plug-in.[ch]
* app/plug-in/plug-in-context.[ch]
* app/plug-in/plug-in-message.[ch]
* app/plug-in/plug-in-progress.[ch]: removed these files...
* app/plug-in/gimpplugin.[ch]
* app/plug-in/gimpplugin-context.[ch]
* app/plug-in/gimpplugin-message.[ch]
* app/plug-in/gimpplugin-progress.[ch]: ...and added here as GObject.
* app/plug-in/plug-in-proc-frame.[ch]: removed...
* app/plug-in/gimppluginprocframe.[ch]: ...and added with a namespace.
* app/plug-in/Makefile.am
* app/plug-in/plug-in-types.h
* app/plug-in/gimppluginmanager-call.c
* app/plug-in/gimppluginmanager-file.c
* app/plug-in/gimppluginmanager.[ch]
* app/pdb/gimppluginprocedure.c
* app/pdb/gimptemporaryprocedure.c
* app/pdb/gimptemporaryprocedure.h
* tools/pdbgen/pdb/context.pdb
* tools/pdbgen/pdb/drawable.pdb
* tools/pdbgen/pdb/help.pdb
* tools/pdbgen/pdb/message.pdb
* tools/pdbgen/pdb/plug_in.pdb
* tools/pdbgen/pdb/progress.pdb
* tools/pdbgen/pdb/undo.pdb: changed accordingly.
* app/pdb/context_cmds.c
* app/pdb/drawable_cmds.c
* app/pdb/help_cmds.c
* app/pdb/message_cmds.c
* app/pdb/plug_in_cmds.c
* app/pdb/progress_cmds.c
* app/pdb/undo_cmds.c: regenerated.
2006-05-03 Sven Neumann <sven@gimp.org>
* app/core/gimpparamspecs.c (gimp_string_array_new,
gimp_param_string_array_validate): relaxed the checks so that
GimpStringArray deals correctly with (length == 0 && data != NULL).
Fixes bug #340365.
Hans Breuer