2000-08-23 Garry R. Osgood <gosgood@idt.net>
* app/image_map.c
A bit of refresh drawable code was misplaced
these many years ago, leading to the persistence
of color tool effects, as reported in Mihail
Milushev's nicely composed bug report. Closes#7626
which has the gory details.
2000-08-24 Sven Neumann <sven@gimp.org>
* plug-ins/gap/iter_ALT/*/*.inc: removed COMPAT_CRUFT
* libgimp/gimpcompat.h
* libgimp/gimpenums.h
* tools/pdbgen/enumcode.pl: reversed the COMPAT_CRUFT logic:
Default is not to include compatibility definitions. Define
GIMP_ENABLE_COMPAT_CRUFT to make old plug-ins compile against
the 1.2 API.
* plug-ins/print/print_gimp.h: define GIMP_ENABLE_COMPAT_CRUFT
here since we will update gimp-print in a few days anyway.
2000-08-23 Sven Neumann <sven@gimp.org>
* app/fileops.c: make the file_save dialog sensitive again if the
user clicked "No" when asked if he wishes to overwrite a file.
* app/gimpimage.c: when saving an undoable parasite, just do it
without calling undo_push_cantundo(). Undoable parasites are used
by many save plug-ins to attach a comment, but this operation
shouldn't prevent the user from using the undo_stack afterwards.
The JPEG plug-in has an additional bug that I couldn't solve, so
undoing after a Save As JPEG is still impossible.
* app/undo_history.c: lame attempt to draw the clean symbol at the
correct undo step. Not perfect but much better as before...
2000-08-23 Sven Neumann <sven@gimp.org>
* libgimp/gimpimage.[ch]
* libgimp/gimpproceduraldb.[ch]: pass the return value of the
PDB call through the wrapper functions.
2000-08-23 Sven Neumann <sven@gimp.org>
* plug-ins/common/threshold_alpha.c: avoid floating point exception
on small images (reported by Steven Grady)
* plug-ins/common/CML_explorer.c
* plug-ins/common/max_rgb.c
* plug-ins/common/nlfilt.c
* plug-ins/common/scatter_hsv.c: same fix here
* configure.in: $srcdir != $builddir fix for GIMP_CONTRIBUTORS
* plug-ins/common/plugin-defs.pl: More tidying up after Daniel Egger
* tools/pdbgen/Makefile.am: parse pdb defs only once for app and lib
* tools/pdbgen/lib.pl
* tools/pdbgen/pdb/procedural_db.pdb: all void rets now return a
boolean for success/failure
-Yosh
2000-08-21 Sven Neumann <sven@gimp.org>
* plug-ins/common/ps.c: applied gimp-kirchgessner-000820-0,
a patch by Peter Kirchgessner which fixes a problem with the
calculation of extent for Mac files (bug report #10547) and
a problem with the loop when not reading all images of a
multipage file.
Sun Aug 20 21:02:22 BST 2000 Austin Donnelly <austin@gimp.org>
* emboss.c
* jigsaw.c
* max_rgb.c
* nlfilt.c
* wind.c
* waves.c: When a radio button in GTK is clicked, the
widget that used to be selected gets a signal as well as the
one that was clicked. Mostly people cope with this by
guarding their signal handers with a check to run to the code
only if the widget is active. However, the callback functions
registered with gimp_radio_group_new2() in the above files
don't have this guard. In most cases, this is a bad idea
since it causes previews to be updated too many times. In the
emboss case it would also get the preview contents wrong when
first displaying the dialog, since the act of
gtk_widget_show()ing the dialog causes an event to be sent to
the widget that's just been de-selected, causing the preview
to redraw while the settings are inconsistent.
* emboss.c: The emboss plugin also used uninitialised data the
first time it is started, which wasn't biting currently, but
might in the future or on other architectures which don't zero
their stacks. Also made active region estimation more robust.
I still haven't been able to reproduce Bug #17660: Embross
plugin crashes, but I will admit that there's what could be
uninitialised memory in the preview bitmap the very first time
it is shown.
2000-08-20 Sven Neumann <sven@gimp.org>
* plug-ins/imagemap/Makefile.am
* plug-ins/imagemap/imap_cern_parse.[ch]
* plug-ins/imagemap/imap_csim_parse.[ch]
* plug-ins/imagemap/imap_ncsa_parse.[ch]: recreated from yacc
files and added those files to the distribution. Supposed to
close bugs #8076, #9212 and #10090.
* plug-ins/imagemap/imap_default_dialog.c
* plug-ins/imagemap/imap_file.c
* plug-ins/imagemap/imap_main.c: set help accelerator and window icon.
Fri Aug 18 18:14:35 BST 2000 Austin Donnelly <austin@gimp.org>
* app/fileops.c: fix nasty bug where if overwriting a file and
also changing the format, gimp would use the old format
rather than the newly selected one. To repeat: save a new
image as (eg) jpeg by typing foo.jpg<RET>, then Save As XCF by
using the pop-up menu and clicking OK, then Save As jpeg by
using pop-up and clicking OK, and OK again to allow the file
to be overwritten. You should find that foo.jpg contains an
XCF version of the image. This was because the "overwrite"
path through the code was subtly different from the standard
way through, meaning that gimage_set_save_proc() was never
called. Now unified into new function file_save_with_proc().
Thu Aug 17 22:07:50 BST 2000 Andy Thomas <alt@gimp.org>
* plug-ins/sel2path/globals.h
Fixed compilation problems. Some really old compilers had problems
with the macros in sel2path. Thanks to
Dougal_R_Scott@nag.national.com.aui for pointing it out.
* app/layers_dialog.c
Fixes#16583: "new view" not being updated correctly. I still think
theres is a problem in the idle rendering when images that
have multiple views require updating.
I could only get the deleting layers operation
to leave the "other views" out of date. This fix forces the views
to be updates as soon as the layer is deleted. I think this is
acceptable since we are doing a "large scale" operation and the
performance loss by not doing the update in the background is OK.
* plug-ins/common/plugindetails.c
Fixed problem when doing searches for plugins containing a
given string caused gtk warnings to be issued
2000-08-16 Garry R. Osgood <gosgood@idt.net>
* app/gdisplay.c
* app/image_map.c
Addresses Bug Report #14704, and an
undisclosed undo system bug. Freezes
undo stack before interactive tools
commence idle rendering; thaws stack
on cleanup and immediately sets menu
sensitivity.
Not an image lock, but Yet Another UI
Lock. Oh well; merrilly we patch along...
See http://bugs.gnome.org/db/14/14704.html
for gory details. Closes#14704.