Apart from being less code, this actually gives us a nice performance
improvement. Up until a few years ago, if you pass `NULL` as the
marshaller for a signal, GLib would fall back to
`g_cclosure_marshal_generic` which uses libffi to pack/unpack its
arguments. One could avoid this by specifying a more specific
marshaller which would then be used to immediately pack and unpack into
GValues with the correct type.
Lately however, as a way of optimizing signal emission (which can be
quite expensive), GLib added a possibility to set a va_marshaller, which
skips the unnecessary GValue packing and unpacking and just uses a
valist variant.
Since the performance difference is big enough, if the marshaller
argument is NULL, `g_signal_new()` will now check for the simple
marshallers (return type NONE and a single argument) and set both the
generic and the valist marshaller. In other words, less code for us with
bigger optimizations.
In case you also want va_marshallers for more complex signals, you can
use `g_signal_set_va_marshaller()`.
`g_object_notify()` actually takes a global lock to look up the property
by its name, which means there is a performance hit (albeit tiny) every
time this function is called. For this reason, always try to use
`g_object_notify_by_pspec()` instead.
if the same name is set via gimp_object_set_static_name() or
gimp_object_take_name() (gimp_object_set_name() and
gimp_object_set_name_safe() already had the check).
Since this is a slight behavior change it might have subtle side
effects; it definitely fixes GimpThumbBox thumbnail state to not
always re-create the thumbnail on click, which was happening since
porting to GFile (which switched from gimp_object_set_name() to
gimp_object_take_name()).
- don't include <gdk-pixbuf/gdk-pixbuf.h> in headers in app/
- instead, include it in many .c files instead of <glib-object.h>,
finally acknowledging the fact that app/ depends on gdk-pixbuf almost
globally
- fix up includes as if libgimpbase depended in GIO, which it soon will
Begin to consider GimpObject::name as private and always use
gimp_object_get_name(). Change gimp_object_get_name() to take an
untyped pointer so we don't have to do so awfully many casts. There is
a runtime check for the type inside the function anyway.
2009-01-17 Michael Natterer <mitch@gimp.org>
* all files with a GPL header and all COPYING files:
Change licence to GPLv3 (and to LGPLv3 for libgimp).
Cleaned up some copyright headers and regenerated the parsers in
the ImageMap plugin.
svn path=/trunk/; revision=27913
2007-03-04 Michael Natterer <mitch@gimp.org>
* app/core/gimpobject.c (gimp_object_get_memsize): don't pass
random strings to g_print() without running them through %s.
Fixes crash/warning when the string contains '%'.
svn path=/trunk/; revision=22042
2006-04-07 Michael Natterer <mitch@gimp.org>
* app/core/gimpobject.h: added "guint disconnected : 1", we
have enough space for flags now.
* app/core/gimpobject.c (gimp_object_dispose): use the new
object->disconnected flag instead of attaching that boolean to the
object.
2006-04-05 Tor Lillqvist <tml@novell.com>
* app/core/gimpobject.c (gimp_object_get_memsize): Check if
object->name is NULL before attempting to pass it to
g_strdup_printf(). Use "anonymous" for unnamed objects.
2006-01-18 Michael Natterer <mitch@gimp.org>
* app/config/config-types.c: define GIMP_PARAM_STATIC_STRINGS
which is G_PARAM_STATIC_NAME|NICK|BLURB. Also define
GIMP_PARAM_READABLE, _WRITABLE and _READWRITE which include
GIMP_PARAM_STATIC_STRINGS.
* app/*/*.c: use them for all object properties so their
strings are not copied.
2005-12-10 Michael Natterer <mitch@gimp.org>
* app/config/*.c
* app/core/*.c
* app/display/*.c
* app/text/*.c
* app/vectors/*.c: port to G_DEFINE_TYPE() and friends. Some related
core reordering and cleanup.
2005-10-10 Michael Natterer <mitch@gimp.org>
* app/core/gimpviewable.c (gimp_viewable_get_description): set the
"tooltip" return value to NULL before calling the virtual function.
* app/core/gimpbrush.c
* app/core/gimpbuffer.c
* app/core/gimpchannel.c
* app/core/gimplayer.c
* app/core/gimppaintinfo.c
* app/core/gimppalette.c
* app/core/gimppattern.c
* app/core/gimptoolinfo.c (get_description): don't do it here.
* app/core/gimpbuffer.c
* app/core/gimppaintinfo.c: untabbify, cleanup.
* app/core/gimpobject.c: emit "notify::name" whenerver the name
changes, clenup.
2003-11-25 Michael Natterer <mitch@gimp.org>
* app/core/gimpobject.[ch]: removed gimp_g_object_get_memsize()...
* app/core/gimp-utils.[ch]: ...and added it here along with
some other memsize utilities for GHashTables and G(S)Lists.
* app/core/gimp.c
* app/core/gimpimage.c
* app/core/gimpparasitelist.c
* app/core/gimpviewable.c
* app/vectors/gimpstroke.c (GimpObject::get_memsize): use the new
functions.
2003-11-16 Sven Neumann <sven@gimp.org>
* app/core/gimpobject.[ch]: exposed gimp_object_name_free() and
added documentation.
* app/core/gimplist.[ch] (gimp_list_uniquefy_name): use
gimp_object_name_free() when changing the object name silently.
Renamed "use_set_name" parameter to "notify" and documented the
function.
2003-11-16 Sven Neumann <sven@gimp.org>
* app/core/gimpobject.c: fixed a bad oversight in the
implementation of normalized object names. Introduced the helper
function gimp_object_name_free() that takes care of freeing
the normalized name. Fixes bug #127075.
2003-10-20 Sven Neumann <sven@gimp.org>
* app/core/gimpobject.c (gimp_object_name_collate): must use
strcmp() not strcoll() on the collation keys obtained from
g_utf8_collate_key().
2003-10-09 Sven Neumann <sven@gimp.org>
* app/core/gimpobject.[ch]: added new function
gimp_object_name_collate() which compares two object names for
ordering using the linguistically correct rules for the current
locale and does some caching to speed up subsequent calls.
* app/core/gimpdatalist.c (gimp_data_list_data_compare_func): use
gimp_object_name_collate() from here.
* app/core/gimplist.[ch]: added convenience function
gimp_list_sort_by_name.
* app/text/gimpfontlist.c (gimp_font_list_restore): use
gimp_list_sort_by_name() instead of g_utf8_collate.
2003-08-25 Michael Natterer <mitch@gimp.org>
* app/core/gimpobject.[ch]: changed GimpObject::get_memsize() to
return a second value named "gui_size", where the primary return
value is the "constant" actual size (as long as no operation is
performed on the object), and the second "gui_size" return value
is the size of temporary stuff like preview caches or boundary
segments (which may change asynchronously, even if the object is
on the undo stack).
* app/core/gimp.c
* app/core/gimpbrush.c
* app/core/gimpbrushpipe.c
* app/core/gimpbuffer.c
* app/core/gimpchannel.c
* app/core/gimpcontainer.c
* app/core/gimpcontext.c
* app/core/gimpdata.c
* app/core/gimpdatafactory.c
* app/core/gimpdrawable.c
* app/core/gimpgradient.c
* app/core/gimpimage-undo.c
* app/core/gimpimage.c
* app/core/gimpitem.c
* app/core/gimplayer.c
* app/core/gimplist.c
* app/core/gimppalette.c
* app/core/gimpparasitelist.c
* app/core/gimppattern.c
* app/core/gimpundo.c
* app/core/gimpundostack.c
* app/core/gimpviewable.c
* app/text/gimptextlayer.c
* app/vectors/gimpstroke.c
* app/vectors/gimpvectors.c: changed get_memsize() implementations
accordingly.
* app/display/gimpdisplayshell-title.c
* app/gui/debug-commands.c
* app/widgets/gimppreview.c: changed callers accordingly.
* app/core/gimpimage-undo-push.c: changed layer, channel, vectors
and layer_mask undo steps to add/subtract the size of the
resp. objects whenever they take/drop ownership of them. Ignore
the objects' "gui_size" to get identical sizes on
adding/subtracting. Fixes bug #120429.
2003-02-05 Sven Neumann <sven@gimp.org>
* libgimpbase/Makefile.am
* libgimpbase/gimputils.[ch]: added new files that hold the new
gimp_utf8_strtrim() routine; it might be useful in more places.
* libgimpbase/gimpdatafiles.c (gimp_datafiles_read_directories):
silently ignore directories in the path that can't be opened.
* app/core/gimpobject.c (gimp_object_set_name_safe): use
gimp_utf8_strtrim().
* app/widgets/gimpwidgets-utils.[ch]
* app/tools/gimptextoptions.c: try to make the text tool options
look more like all other tool options. Still needs work; I'll
leave this up to Mitch ...
This byte --> <-- is the millionth in this file!
2003-02-05 Sven Neumann <sven@gimp.org>
* app/core/gimpobject.c (gimp_object_set_name_safe): replaced with
an UTF-8 safe rewrite.
* app/widgets/gimpwidgets-utils.c (gimp_table_attach_stock): hacked
to allow being used with non-existent stock_ids (for prototyping).
* app/widgets/gimpenummenu.c: set the radio button mode correctly.
* app/widgets/gimpfontselection.c: tweaked spacing.
* app/tools/gimptextoptions.c: added controls for justification and
indentation. Removed letter-spacing control for now.
* app/text/gimptextlayer.[ch]: rerender the text layer from a low
priority idle function.
2003-02-05 Sven Neumann <sven@gimp.org>
* app/core/gimpobject.[ch]
* app/text/gimptextlayer.c: improved the code that trims the text
before setting it as layer name and moved it to GimpObject as
gimp_object_set_name_safe().
* app/text/gimptext.[ch]: removed fixed_height and gravity again.
* app/text/gimptextlayout.c: much simpler positioning that has the
advantage that it actually works.
* libgimpproxy/gimpobject.[ch]: this crap was regenerated.
2003-01-05 Manish Singh <yosh@gimp.org>
* app/core/gimpobject.c: cast memsize to a guint instead of format
change. Problem perhaps if get_size really is > 4 GB?
2003-01-05 Manish Singh <yosh@gimp.org>
* many files in app, modules and libgimp*: cleanup, removed unecessary
G_OBJECT() casts. Should do the same for plug-ins, when more of them
get undeprecated.
2002-04-03 Manish Singh <yosh@gimp.org>
* app/core/gimpobject.c: use gobject marshaller names directly so
libgimpproxy doesn't need gimpmarshal.* around
2002-02-06 Michael Natterer <mitch@gimp.org>
* app/core/gimpobject.c: made the gimp_object_get_memsize()
debugging output configurable by a global "gimp_debug_memsize"
boolean.
* app/display/gimpdisplay.c: removed duplicated prototype.
* app/display/gimpdisplayshell.[ch]: renamed the various cursor
functions to be more consistent and shorter. Compress window title
updates by adding a "gboolean title_dirty" and updating the title
in gimp_display_shell_flush(). Added "%m" (memory size) to the
possible title string substitutions.
* app/display/gimpdisplay-foreach.c
* app/display/gimpdisplayshell-handlers.c
* app/tools/gimpfuzzyselecttool.c
* app/tools/gimptool.c: changed accordingly.
* app/display/gimpdisplayshell-callbacks.c: forgot to grab the
pointer when dragging guides from the rulers. Coincidentially,
this also fixes the buggy offset between guide and mouse
pointer...
Cleaned up the main tool event callback a but more.
* app/widgets/gimppreview.c
* app/gui/commands.c: set the new global "gimp_debug_memsize"
toggle to TRUE while calling gimp_object_get_memsize().
* app/gui/preferences-dialog.c: added a image title example
containing the new "%m" feature.
* docs/gimprc-1.3.5.in: document "%m" in the manpage.
* app/tools/gimpbezierselecttool.c: reordered some statements.
* app/tools/gimpdrawtool.[ch]: store the GimpDisplay passed to
gimp_draw_tool_start() in draw_tool->gdisp and use it for
coordinate transfomration. This way we can paint on a display
which is not tool->gdisp.
* app/tools/gimppainttool.c: changed the gimp_draw_tool_foo()
calls needed to make the straight_line preview work in a way
that does not interfere with paint_tool subclasses which want
to do their own drawing (like the clone tool).
Also changed the paint_tools PRETRACE_PAINT and POSTTRACE_PAINT
flags usage in a way that subclasses can use them without major
hackery: don't simply wrap gimp_display_flush_now() with
PRETRACE/POSTTRACE calls, but wrap the actual painting calls, so
subclasses are able to do useful things with paint_tool->*_coords.
* app/tools/gimpclonetool.c: removed poking around in draw_tool
internals and simply suspend()/resume() it in
PRETRACE_PAINT/POSTTRACE_PAINT to get the clone_src indicator
drawn correctly.
