2006-05-20 Manish Singh <yosh@gimp.org>
Fixes to address -ansi -pedantic compilation (bug #342390).
Thanks goes to Daniel Richard G. for noticing and suggesting
fixes.
* libgimpbase/gimpsignal.c: #include "config.h" and define
__POSIX_SOURCE for sigaction stuff.
* app/base/tile-swap.c (tile_swap_test): use more portable
S_IRUSR and S_IWUSR, instead of S_IREAD and S_IWRITE.
* plug-ins/common/ripple.c
* plug-ins/imagemap/imap_main.c: use C89 comments.
* plug-ins/Lighting/lighting_preview.h: don't define spin widget
variables here...
* plug-ins/Lighting/lighting_ui.[ch]: ... and instead take care
of them here.
2006-04-21 Sven Neumann <sven@gimp.org>
Applied a slightly modified version of a patch from Lode Leroy
(bug #336183):
* app/base/gimphistogram.[ch]: added new function
gimp_histogram_get_threshold() that returns an optimal
binarization threshold.
* app/tools/gimpthresholdtool.c: use the new function to
initialize the Threshold tool.
2006-04-20 Simon Budig <simon@gimp.org>
* app/base/curves.c: implemented a new interpolation for the
"smooth" curve type. This fixes the artefacts as described in
bug #169078 although there definitely is room for improvement.
This has an impact on the API as now the curves used by the
gimp-curves-spline PDB call change. I do however believe, that
the change is for good and the old behaviour was simply buggy.
2006-04-05 Sven Neumann <sven@gimp.org>
* app/base/lut-funcs.[ch]: minor code cleanup, renamed a function,
declared some functions as static.
* app/core/gimpdrawable-equalize.c: changed accordingly.
2006-03-30 Sven Neumann <sven@gimp.org>
* app/base/siox.c (siox_foreground_extract): actually increment
the tile counter used to limit the number of progress updates.
2006-03-23 Sven Neumann <sven@gimp.org>
* app/base/siox.c (siox_foreground_extract): fixed the potential
problem that the cache wouldn't have been cleared correctly if
both background and foreground areas are added at the same time.
2006-03-22 Sven Neumann <sven@gimp.org>
* app/base/siox.[ch] (siox_foreground_extract): let the caller
specify a region of interest. Call the progress function more often.
* app/core/gimpdrawable-foreground-extract.c
(gimp_drawable_foreground_extract_siox): pass a region of interest
to siox_foreground_extract() to speed up the SIOX tool.
2006-02-24 Simon Budig <simon@gimp.org>
* app/base/lut-funcs.c: change the LUT function for the contrast.
Fixes the buggy behaviour as described in bug #319872 and
bug #332068 and makes the behaviour consistent with the standard
contrast formula. However, I am leaving the bug open to discuss
further improvements.
Untabbified.
2005-11-04 Sven Neumann <sven@gimp.org>
* app/base/siox.c: applied a patch by Gerald Friedland that
changes the code to clear the cache selectively instead of dumping
it completely if the foreground or background area is changed.
2005-10-28 Sven Neumann <sven@gimp.org>
* app/base/siox.c: applied a patch by Tobias Lenz that makes use
of the state that was added to SIOX in order to speed up the tool.
2005-10-17 Manish Singh <yosh@gimp.org>
* app/base/siox.c (siox_drb): declare mindistbg, mindistfg, and alpha
float variables at the start of the block.
2005-10-17 Sven Neumann <sven@gimp.org>
* app/base/base-enums.h
* app/base/siox.[ch]
* app/core/gimpdrawable-foreground-extract.[ch]
* app/tools/gimpforegroundselecttool.[ch]: export stateful SIOX to
the core API and use it from the foreground selection tool.
2005-10-12 Sven Neumann <sven@gimp.org>
* app/base/base-types.h
* app/base/siox.[ch]: applied patch from Gerald Friedland that
introduces a state to the SIOX operation.
* app/core/gimpdrawable-foreground-extract.c: adapted to the new API.
2005-09-29 Sven Neumann <sven@gimp.org>
* app/base/tile-cache.c
* app/base/tile-manager.c
* app/base/tile-private.h
* app/base/tile-swap.c
* app/base/tile.c: keep the data size of a tile in the Tile struct
instead of recalculating it over and over again. Adds an overhead
of 4 bytes per tile but speeds up the tile cache significantly.
2005-09-25 Sven Neumann <sven@gimp.org>
* app/base/siox.[ch]: applied patch from Gerald Friedland and
Tobias Lenz. Introduces a hash table to speed up the algorithm
and changes the post-processing so that multiple disjunct objects
can be selected.
* app/tools/gimpforegroundselecttool.c: changed tooltip, the tool
doesn't any longer extract single objects only.
2005-09-24 Hans Breuer <hans@breuer.org>
* **makefile.msc : updated
* app/dialogs/user-install-dialog.c : only add the migrate page if
there is something to migrate from. Avoids on version being NULL.
* app/dialogs/file-save-dialog.c : the g_print() output was crashing
on the assumption that ->menu_label != NULL. It is for colorhtml.py.
* app/widgets/gimpselectiondata.c : use HAVE_UNISTD_H and move
* process.h definition by G_OS_WIN32 below it being defined
* app/widgets/gimpwidgets-utils.c(gimp_window_get_native) : cast
return value to (GdkNativeWindow) it is not necessary an int.
* libgimpwidgets/gimpwidgets.def : added gimp_zoom_type_get_type
* plug-ins/help/gimp-help-lookup.c : dynamic lookup of help_root
instead of hard-coding DATADIR/GIMP_HELP_PREFIX
* plug-ins/xjt/xjt.c : there is no pid_t with msvc, typedef one.
2005-09-03 Michael Natterer <mitch@gimp.org>
* app/base/pixel-region.[ch]: some general cleanup.
(pixel_region_init_temp_buf)
(pixel_region_init_data): new functions which initialize pixel
regions on TempBufs and on raw contiguous arrays of pixel data.
(pixel_region_configure): fixed a bug that has probably been there
forever: when processing contiguous (non-tiled) data, interpret
the original x and y coordinates of the region as offsets into
the data. Before this fix, the initial x and y were simply ignored
(by using them in a broken way), thus always forcing the upper
left corner of the region being the beginning of the passed data.
Lots of code was working around this problem by setting the
pixel_region's data pointer to the proper starting pixel of the
region in the middle the buffer.
* libgimp/gimppixelrgn.c: some general cleanup.
(gimp_pixel_rgn_configure): same fix as above. Fortunately, nobody
seems to know that libgimp pixel regions can be used on arrays of
data, just as core ones. Only two plug-ins were using this
feature, and they are antique and written by spencer and federico,
respectively. They both don't use offsets into the buffers and are
not affected by this change. It's highly unlikely that anybody out
there knows/uses this feature, so it can IMHO be safely changed.
* app/base/temp-buf.c
* app/core/gimpbuffer.c
* app/core/gimpdrawable-combine.c
* app/core/gimpdrawable-preview.c
* app/core/gimpimage-preview.c
* app/core/gimplayer.c
* app/paint/gimpbrushcore.c
* app/paint/gimpclone.c
* app/paint/gimpconvolve.c
* app/paint/gimpdodgeburn.c
* app/paint/gimppaintcore.c
* app/paint/gimpsmudge.c
* app/tools/gimpiscissorstool.c
* app/tools/gimppainttool.c: use the pixel_region_init_foo()
functions instead of initializing regions of TempBufs and raw data
manually. Removed lots of workarounds for the broken offset
handling. The changed places of code are much more readable now.
2005-08-23 Sven Neumann <sven@gimp.org>
* app/base/siox.c: minor optimizations.
* app/actions/gradients-actions.c
* app/actions/patterns-actions.c
* plug-ins/metadata/interface.c: applied patch from Stephan Binner
that fixes wrong use of ellipsis (bug #309657).
2005-08-20 Michael Natterer <mitch@gimp.org>
* app/base/boundary.[ch]: renamed puclic functions, defines and
enums to live in a proper "boundary" namespace. Cleaned up and
simplified internal code even more.
* app/core/gimpchannel.c
* app/core/gimpdrawable-stroke.c
* app/core/gimplayer-floating-sel.c
* app/paint/gimppaintcore-stroke.c
* app/tools/gimpeditselectiontool.c
* app/tools/gimpfuzzyselecttool.c
* app/tools/gimppainttool.c: changed accordingly.
2005-08-20 Michael Natterer <mitch@gimp.org>
* app/base/boundary.c: moved the entire internal state of boundary
creation and simplification from <eek>global variables</eek> to a
new "Boundary" struct that needs to be newly allocated for each
boundary. Yes this allocates and frees memory more often than
before but the old code was simply one big eek.
2005-08-20 Michael Natterer <mitch@gimp.org>
* app/base/boundary.c (sort_boundary): look at the passed "ns"
variable instead of at the global "num_segs" one to figure if the
passed array of BoundSegs is empty. Fixes bug #313922 at its root.
2005-08-17 Sven Neumann <sven@gimp.org>
* app/base/siox.c: applied patch by Kristian Jantz. Reimplements
find_max_blob() so that it needs less memory. Slows things down a
bit, but that can be improved later.
