framework/uifw/evas.git
14 years agowarning--
raster [Sun, 19 Sep 2010 02:04:12 +0000 (02:04 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52444 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sun, 19 Sep 2010 02:02:09 +0000 (02:02 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52443 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarnings--
raster [Sun, 19 Sep 2010 01:57:19 +0000 (01:57 +0000)]
warnings--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52442 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sun, 19 Sep 2010 01:46:13 +0000 (01:46 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52441 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sun, 19 Sep 2010 01:42:45 +0000 (01:42 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52440 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sun, 19 Sep 2010 01:41:46 +0000 (01:41 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52439 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sun, 19 Sep 2010 01:33:51 +0000 (01:33 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52438 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agomaybe actually fix things correctly eh? -> images load again.
raster [Sun, 19 Sep 2010 01:14:56 +0000 (01:14 +0000)]
maybe actually fix things correctly eh? -> images load again.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52435 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agofix gl stride
raster [Sun, 19 Sep 2010 00:28:58 +0000 (00:28 +0000)]
fix gl stride

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52432 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sat, 18 Sep 2010 23:45:42 +0000 (23:45 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52431 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowarning--
raster [Sat, 18 Sep 2010 23:39:30 +0000 (23:39 +0000)]
warning--

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52430 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agofree the main region (leak--)
caro [Sat, 18 Sep 2010 23:17:27 +0000 (23:17 +0000)]
free the main region (leak--)

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52429 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocleanup: internal image information is now unsigned.
barbieri [Sat, 18 Sep 2010 23:16:25 +0000 (23:16 +0000)]
cleanup: internal image information is now unsigned.

There is no meaning in negative values for image loading, marking as
dirty or size, so image internals (cache, entry) were changed to
unsigned, reducing possible errors, particularly with overflow.

engines were converted to the new way, but any 3rd party modules will
still work as they should be using values >= 0 only anyway.

please review.

new cases introduced by "comparison between signed and unsigned" were
fixed in the modules that used cache_entry or Image_Entry dimensions.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52428 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocleanup: fb uses unsigned to match linux/fb.h
barbieri [Sat, 18 Sep 2010 23:11:01 +0000 (23:11 +0000)]
cleanup: fb uses unsigned to match linux/fb.h

this removes warnings and keep consistency between system and evas
value signs.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52427 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocleanup: simple clean of "comparison between signed and unsigned errors"
barbieri [Sat, 18 Sep 2010 23:07:31 +0000 (23:07 +0000)]
cleanup: simple clean of "comparison between signed and unsigned errors"

not much to see here, please review but these are simple changes.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52426 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocleanup: evas_macros.h - less comparison between signed and unsigned errors
barbieri [Sat, 18 Sep 2010 23:06:22 +0000 (23:06 +0000)]
cleanup: evas_macros.h - less comparison between signed and unsigned errors

update macro to do fine with unsigned integers (avoids width and
height getting below 0) and force casts.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52425 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocleanup: fix some "unused" errors from -Wextra.
barbieri [Sat, 18 Sep 2010 19:17:41 +0000 (19:17 +0000)]
cleanup: fix some "unused" errors from -Wextra.

As we're heading for a release we better remove as much errors as
possible and as the first step I'm removing warnings due unused
parameters, variables and functions. These tend to pollute real errors
spotted by -Wall and clang/llvm.

This does not fixes all, just the clear that could be set to
__UNUSED__, particularly to do (and I'd like some help from the
authors):

 * src/lib/engines/common/evas_font_{draw,query}.c (tasn):
   intl_props is just used while doing BIDI, but also used in other
   #ifdef blocks :-/

 * evas_map_* (raster):
   huge amount of warnings, code is quite confusing and thus I'm not
   touching it. I have no idea whenever the commented blocks or extra
   parameters are intended to be used or no.

 * src/modules/engines/fbevas_fb_main.c (raster?):
   is fb_setvt() to be used? If not do you mind removing it?

 * src/modules/engines/gl_{common,x11} (raster):
   huge amount of warnings, code is quite nested and full of #ifdefs
   that does not help to give a clear picture of what's going on.

 * src/bin/evas_cserve_main.c (raster):
   I could have ignored most of the errors, but is the code correct? I
   mean, there is no unload of images being applied. If you confirm
   none of those warnings are harmful I can flag them as unused.

 * src/lib/engines/common_8 (dottedmag):
   lots of unused functions that were acquired from common_16, they
   are unused and if they will not, then they should be removed.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52421 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoAdd missing evas_common prefix, avoid symbol redefinition.
barbieri [Sat, 18 Sep 2010 17:43:13 +0000 (17:43 +0000)]
Add missing evas_common prefix, avoid symbol redefinition.

soft16 was written as a single engine, thus it was all static/global
and had no EAPI in its functions, but then it was moved into
"src/lib/common_16" and got that, but got no prefix! That could cause
clash with other libraries, so adding such prefix.

soft8 was a copy of 16, thus had the same problems.

the engines were all based on software_x11, thus they defined the same
methods to deal with Xlib, however if you link them all in the same
binary (--enable-MODULE=static), the symbol would be redefined. Rename
symbols according to their module.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52420 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoremove debug messages
caro [Sat, 18 Sep 2010 17:31:00 +0000 (17:31 +0000)]
remove debug messages

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52419 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoFix shaped windows on Windows. It uses regions.
caro [Sat, 18 Sep 2010 17:19:11 +0000 (17:19 +0000)]
Fix shaped windows on Windows. It uses regions.

Advantages:
 * it is the simplest method to implement

Disadvantages:
 * it's slow
 * it does not take into account transparency
 * it does not work with the composite manager (Windows >= Vista)

Layered windows should be used (all the disadvantaged above are
fixed), but i've never succeeded in making them work.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52416 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agofix actually supporting clip for maps in gl.
raster [Sat, 18 Sep 2010 14:23:20 +0000 (14:23 +0000)]
fix actually supporting clip for maps in gl.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52415 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agobe less agressive performance-wise but clear up some gl sync issues
raster [Sat, 18 Sep 2010 13:49:56 +0000 (13:49 +0000)]
be less agressive performance-wise but clear up some gl sync issues
that have been lurking.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52414 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoFix function call after previous change
lucas [Sat, 18 Sep 2010 06:41:04 +0000 (06:41 +0000)]
Fix function call after previous change
Forgot one function call.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52408 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoMake _evas_event_havemap_adjust() recursive
lucas [Sat, 18 Sep 2010 06:31:41 +0000 (06:31 +0000)]
Make _evas_event_havemap_adjust() recursive
Make this function recursive, so it can adjust the coords for all
parent objects. It starts with the grand-grand-grand-...-parent and goes
down until the same object.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52407 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoUse also map of current object
lucas [Sat, 18 Sep 2010 06:31:12 +0000 (06:31 +0000)]
Use also map of current object
Do not use only the map of the parent. Apply the obj's map first and
then go to the parent.

+  simple formatting

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52406 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoRemove havemap_parent shortcut
lucas [Sat, 18 Sep 2010 06:31:05 +0000 (06:31 +0000)]
Remove havemap_parent shortcut

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52405 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoClean unneeded check
lucas [Sat, 18 Sep 2010 06:30:59 +0000 (06:30 +0000)]
Clean unneeded check

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52404 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoDon't set parmap for the other objects
lucas [Sat, 18 Sep 2010 06:30:51 +0000 (06:30 +0000)]
Don't set parmap for the other objects
We are iterating EINA_INLIST_REVERSE_FOREACH(list, obj) in a recursive
function. Don't mark the other objects as havemap_parent if the first
in the list has it.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52403 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoformatting
lucas [Sat, 18 Sep 2010 06:30:45 +0000 (06:30 +0000)]
formatting

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52402 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoclean a bit more
lucas [Sat, 18 Sep 2010 06:30:37 +0000 (06:30 +0000)]
clean a bit more

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52401 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoClean it a bit to be easier to understand
lucas [Sat, 18 Sep 2010 06:30:30 +0000 (06:30 +0000)]
Clean it a bit to be easier to understand

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52400 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoRevert r52345
lucas [Sat, 18 Sep 2010 06:29:47 +0000 (06:29 +0000)]
Revert r52345

This function is not needed at all, and as of now it's borken. Coming patches
will properly fix coords on events.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52398 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoAdd function to check if point is inside an object
lucas [Thu, 16 Sep 2010 13:59:46 +0000 (13:59 +0000)]
Add function to check if point is inside an object

Add evas_object_inside_get() to check if a certain (x,y) point is inside
an evas_object. This is needed because there's no simple way to
determine it outside of evas when map transformations are used.

For instance, edje uses evas_object_geometry_get() and checks if point is
whithin the rectangle. This is wrong because the object might be
rotated, scaled. Below is a test program:

/**
 * Simple Evas test
 *
 * Compile with: gcc -O0 -g  -o evas_hello_world evas_hello_world.c $(pkg-config --cflags --libs eina evas ecore ecore-evas)
 */
 #include <Eina.h>
 #include <Evas.h>
 #include <Ecore_Evas.h>
 #include <Ecore.h>
 #include <stdio.h>

 #define WIDTH (320)
 #define HEIGHT (240)

Eina_Bool main_signal_exit(void *data, int ev_type, void *ev)
{
    ecore_main_loop_quit();
    return EINA_FALSE;
}

static void
_cb(void *data, Evas *e, Evas_Object *obj, void *event_info)
{
    Evas_Event_Mouse_Up *ev = event_info;
    Eina_Bool b;

    b = evas_object_inside_get(obj, ev->canvas.x, ev->canvas.y);
    fprintf(stderr, "mouse_up: x=%d, y=%d inside=%d\n", ev->canvas.x,
            ev->canvas.y, b);
}

static void
_cb_move(void *data, Evas *e, Evas_Object *obj, void *event_info)
{
    Evas_Event_Mouse_Move *ev = event_info;
    const char *id = data;
    Eina_Bool b;

    b = evas_object_inside_get(obj, ev->cur.canvas.x, ev->cur.canvas.y);
    fprintf(stderr, "[%s] mouse_move: x=%d, y=%d inside=%d\n", id,
            ev->cur.canvas.x, ev->cur.canvas.y, b);
}

int main(void)
{
   Evas *evas;
   Ecore_Evas *window;
   Evas_Object *bg, *r1, *r2;
   Evas_Map *m;

   evas_init();
   ecore_init();
   ecore_evas_init();

   window = ecore_evas_new(NULL, 0, 0, WIDTH, HEIGHT, NULL);
   if (!evas)
     return -1;

   evas = ecore_evas_get(window);

   bg = evas_object_rectangle_add(evas);
   evas_object_color_set(bg, 255, 255, 255, 255); // white bg
   evas_object_move(bg, 0, 0);                    // at origin
   evas_object_resize(bg, WIDTH, HEIGHT);         // covers full evas
   evas_object_show(bg);

   r1 = evas_object_rectangle_add(evas);
   evas_object_color_set(r1, 255, 0, 0, 255); // 100% opaque red
   evas_object_move(r1, 50, 50);
   evas_object_resize(r1, 100, 100);

   m = evas_map_new(4);
   evas_map_util_points_populate_from_object(m, r1);
   evas_map_util_rotate(m, 45.0, 100, 100);
   evas_map_alpha_set(m, 0);
   evas_map_smooth_set(m, 1);

   evas_object_map_set(r1, m);
   evas_object_map_enable_set(r1, 1);
   evas_map_free(m);
   evas_object_show(r1);
   evas_object_event_callback_add(r1, EVAS_CALLBACK_MOUSE_UP, _cb, NULL);
   evas_object_event_callback_add(r1, EVAS_CALLBACK_MOUSE_MOVE, _cb_move, "r1");

   r2 = evas_object_rectangle_add(evas);
   evas_object_color_set(r2, 0, 255, 0, 255);
   evas_object_move(r2, 210, 150);
   evas_object_resize(r2, 50, 50);
   evas_object_show(r2);
   evas_object_event_callback_add(r2, EVAS_CALLBACK_MOUSE_MOVE, _cb_move, "r2");

   ecore_event_handler_add(ECORE_EVENT_SIGNAL_EXIT, main_signal_exit, evas);

   ecore_evas_show(window);
   ecore_main_loop_begin();

   ecore_evas_free(window);

   ecore_evas_shutdown();
   ecore_shutdown();
   evas_shutdown();

   return 0;
}

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52345 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Added a lot of fixes to evas_textblock_cursor_geometry_get in the...
tasn [Thu, 16 Sep 2010 09:59:39 +0000 (09:59 +0000)]
Evas textblock: Added a lot of fixes to evas_textblock_cursor_geometry_get in the case where ctype = EVAS_TEXTBLOCK_CURSOR_BEFORE

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52341 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed evas_textblock_range_text_get that didn't work on textblocks...
tasn [Thu, 16 Sep 2010 07:12:23 +0000 (07:12 +0000)]
Evas textblock: Fixed evas_textblock_range_text_get that didn't work on textblocks with no formats at all (not even newlines or tabs).

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52339 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agotrivial: spelling in documentation
lucas [Wed, 15 Sep 2010 20:40:51 +0000 (20:40 +0000)]
trivial: spelling in documentation

Some misspellings found in doxy.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52326 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoCheck there are callbacks before trying to copy them.
sachiel [Wed, 15 Sep 2010 19:51:08 +0000 (19:51 +0000)]
Check there are callbacks before trying to copy them.

Patch by Otávio Pontes

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52323 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years ago * evas: we don't need all the stuff from pthreads to do preload.
cedric [Wed, 15 Sep 2010 17:20:37 +0000 (17:20 +0000)]
* evas: we don't need all the stuff from pthreads to do preload.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52311 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: fix deleting the first tab in 'a<TAB><TAB>a' that caused weird behavior.
tasn [Wed, 15 Sep 2010 14:07:09 +0000 (14:07 +0000)]
Evas textblock: fix deleting the first tab in 'a<TAB><TAB>a' that caused weird behavior.
The solution is that we only delete invisible standalones now, not visible ones, this is correct intuitively and of course fixes the bug.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52302 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: we should also take width of tabs into account when calculating line...
tasn [Wed, 15 Sep 2010 12:45:06 +0000 (12:45 +0000)]
Evas textblock: we should also take width of tabs into account when calculating line width.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52297 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fix and simplify _find_layout_item_line_match which is an helper...
tasn [Wed, 15 Sep 2010 10:24:07 +0000 (10:24 +0000)]
Evas textblock: Fix and simplify _find_layout_item_line_match which is an helper function used in many parts of textblock.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52295 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas font: removed old (already removed a long time ago) functions from evas_font.h.
tasn [Wed, 15 Sep 2010 09:11:19 +0000 (09:11 +0000)]
Evas font: removed old (already removed a long time ago) functions from evas_font.h.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52293 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed the bug with disappearing text with many tabs and text.
tasn [Tue, 14 Sep 2010 13:57:26 +0000 (13:57 +0000)]
Evas textblock: Fixed the bug with disappearing text with many tabs and text.
I removed a function that caused the issue and made no sense at all, honestly, it didn't make any sense.
I did a lot of testing trying to see if there are any new bugs after the fix, and nothing, so I guess my instincts were correct.
Please if you can, check out the removed function (_layout_walk_back_to_item_word_redo) and see if it makes any sense to you, if it does, please let me know.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52243 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoPending patch from glima, who's on vacations.
sachiel [Mon, 13 Sep 2010 18:04:23 +0000 (18:04 +0000)]
Pending patch from glima, who's on vacations.

Add two new canvas level callbacks: OBJECT_FOCUS_IN/OUT
As we already had callbacks for objects gaining or losing focus, then
two more for the Canvas, now we can have the entire Evas be notified when
any object changes its focused status. The object in question is passed
as the event_info for the callback.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52196 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoAssert cache->usage never becomes negative
lucas [Mon, 13 Sep 2010 13:58:34 +0000 (13:58 +0000)]
Assert cache->usage never becomes negative
Assert cache->usage never becomes negative as was occurring before the
fix in r52149.

Just in time, the fix in r52149 was made by Ulisses, not me.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52190 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoFix accounting of memory usage in image cache
lucas [Fri, 10 Sep 2010 23:00:26 +0000 (23:00 +0000)]
Fix accounting of memory usage in image cache
Memory usage was not accounted right because
cache->func.mem_size_get(ie) returns 0 when called after
cache->func.destructor(ie). Thus the total memory used, kept on
cache->usage, is never decremented in _evas_cache_image_remove_activ()

This implies that cache->usage will keep growing and eventually it will
overflow, becoming negative. So evas_cache_image_flush() will not do its
job because cache->limit (assumed to be positive) will not be less than
cache->usage anymore. So the total memory allocated will start to grow
and the limit won't be respected.

Strictly speaking, it's not a leak, since all the memory will be
eventually freed when evas shutdown is called, but the program might be
killed by over allocating memory. This is caught by valgrind with the
massif tool. The graphic below shows that in the end a huge memory amount
is allocated. This is the moment when cache->usage became negative.

MB
26.04^                                                                   #
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
|                                                                   #::::
| :::::::::::::::::@@:::::::::@:::@::::::@::::::::::::::::::::::::::#::::
| : ::: ::: ::: :::@ :: ::: : @:: @:: :::@: :: ::: : : :: :: : ::: :#::::
0 +----------------------------------------------------------------------->Gi
0                                                                   54.83

This patch is a one line fix, which swaps the calls to
_evas_cache_image_remove_activ() and cache->func.destructor() making
cache->limit to be respected.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52149 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoFix priority dropping
lucas [Thu, 9 Sep 2010 12:45:39 +0000 (12:45 +0000)]
Fix priority dropping

Lowering priority was wrong. Some bugs:

1) You don't lower the priority by setting the scheduler policy to some
   of the real-time ones (SCHED_RR or SCHER_FIFO). If you do so, you are
   actually increasing the priority of the workers and your main thread
   you be preempted and stalled until the workers complete their job.
   Fortunately this will only happen if your programming is running as
   root, as normal users (without CAP_SYS_NICE) are unable to set
   priority to real-time values.

2) setpriority() and getpriority() are not part of pthread and you can't
   use the id returned by pthread. Manpage explicitly says so on
   pthread_self(3):
   "The  thread ID returned by pthread_self() is not the same thing as the
       kernel thread ID returned by a call to gettid(2)."

   Since glibc does not have a gettid, here we are using
   syscall(SYS_gettid)

This patch was tested with the program below. Compile and run:
   $ gcc p_hello2.c -o p_hello2 -lpthread
   $ ./p_hello2 10

You'll see that the main thread remains with its priority and threads
created by the main thread change their own niceness.

 #include <errno.h>
 #include <pthread.h>
 #include <sched.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <sys/resource.h>
 #include <sys/syscall.h>
 #include <sys/time.h>
 #include <sys/types.h>

/* Lower priority of current thread.
 *
 * It's used by worker threads so they use up "bg cpu" as it was really intended
 * to work. If current thread is running with real-time priority, we decrease
 * our priority by 5. This is done in a portable way.  Otherwise we are
 * running with SCHED_OTHER policy and there's no portable way to set the nice
 * level on current thread. In Linux, it does work and it's the only one that is
 * implemented.
 */
static void
_ecore_thread_pri_drop(void)
{
   struct sched_param param;
   int pol, prio, ret;
   pid_t tid;
   pthread_t pthread_id;

   pthread_id = pthread_self();
   ret = pthread_getschedparam(pthread_id, &pol, &param);
   if (ret)
     {
        fprintf(stderr, "Unable to query sched parameters\n");
        return;
     }

   if (pol == SCHED_RR || pol == SCHED_FIFO)
     {
        prio = sched_get_priority_max(pol);
        param.sched_priority += 5;
        if (prio > 0 && param.sched_priority > prio)
           param.sched_priority = prio;

        pthread_setschedparam(pthread_id, pol, &param);
     }
 #ifdef __linux__
   else
     {
        tid = syscall(SYS_gettid);
        errno = 0;
        prio = getpriority(PRIO_PROCESS, tid);
        if (errno == 0)
          {
             prio += 5;
             if (prio > 19)
                prio = 19;

             setpriority(PRIO_PROCESS, tid, prio);
          }
     }
 #endif
}

/*
 * p_hello.c -- a hello program (in pthread)
 */
 #define MAX_THREAD 1000

typedef struct {
    int id;
} parm;

void *hello(void *arg)
{
    parm *p=(parm *)arg;
    pid_t tid;
    int prio;
    tid =  syscall(SYS_gettid);
    printf("[%d] Hello from node %d\n", tid, p->id);
    pthread_yield();

    printf("[%d] HELLO!\n", tid);

    _ecore_thread_pri_drop();

    prio = getpriority(PRIO_PROCESS, tid);
    printf("[%d] New nice value: %d\n", tid, prio);
    return (NULL);
}

int main(int argc, char* argv[]) {
    int n,i;
    pthread_t *threads;
    pthread_attr_t pthread_custom_attr;
    parm *p;

    pid_t tid;
    int prio;

    if (argc != 2)
    {
        printf ("Usage: %s n\n  where n is no. of threads\n",argv[0]);
        exit(1);
    }

    n=atoi(argv[1]);

    if ((n < 1) || (n > MAX_THREAD)) {
        printf ("The no of thread should between 1 and %d.\n",MAX_THREAD);
        exit(1);
    }

    threads = (pthread_t *)malloc(n * sizeof(*threads));
    pthread_attr_init(&pthread_custom_attr);

    p = (parm *)malloc(n * sizeof(parm));
    /* Start up thread */

    tid = syscall(SYS_gettid);
    for (i=0; i<n; i++) {
        prio = getpriority(PRIO_PROCESS, tid);
        printf("[%d] root thread nice value: %d\n", tid, prio);

        p[i].id=i;
        pthread_create(&threads[i], &pthread_custom_attr, hello, (void *)(p+i));
    }

    /* Synchronize the completion of each thread. */

    for (i=0; i<n; i++) {
        pthread_join(threads[i],NULL);
    }
    free(p);

    return 0;
}

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52040 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoreturn NULL after if to avoid other lku
raster [Thu, 9 Sep 2010 07:27:56 +0000 (07:27 +0000)]
return NULL after if to avoid other lku

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52033 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agothat should have been an LKU
raster [Thu, 9 Sep 2010 07:23:55 +0000 (07:23 +0000)]
that should have been an LKU

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52032 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoand some final fixes anc cleanups - tested.
raster [Thu, 9 Sep 2010 07:14:11 +0000 (07:14 +0000)]
and some final fixes anc cleanups - tested.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52031 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agolots of cleanups... actually tested! :)
raster [Thu, 9 Sep 2010 07:08:10 +0000 (07:08 +0000)]
lots of cleanups... actually tested! :)

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52030 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoactually revert that - not enough testing. need to work on it.
raster [Thu, 9 Sep 2010 06:36:01 +0000 (06:36 +0000)]
actually revert that - not enough testing. need to work on it.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52029 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agomajor clean of the preload stuff. leaks. bugs. nastinessesssss...
raster [Thu, 9 Sep 2010 06:31:29 +0000 (06:31 +0000)]
major clean of the preload stuff. leaks. bugs. nastinessesssss...

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52028 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoWhitespace --;
nash [Thu, 9 Sep 2010 05:36:56 +0000 (05:36 +0000)]
Whitespace --;

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@52024 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed paragraph char last.
tasn [Wed, 8 Sep 2010 10:22:38 +0000 (10:22 +0000)]
Evas textblock: Fixed paragraph char last.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51985 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Should not clean the props of the item.
tasn [Wed, 8 Sep 2010 10:00:47 +0000 (10:00 +0000)]
Evas textblock: Should not clean the props of the item.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51984 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed a typo.
tasn [Wed, 8 Sep 2010 07:54:00 +0000 (07:54 +0000)]
Evas textblock: Fixed a typo.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51981 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: cursor_geometry_get now also returns the direction of the cursor...
tasn [Wed, 8 Sep 2010 07:28:16 +0000 (07:28 +0000)]
Evas textblock: cursor_geometry_get now also returns the direction of the cursor: rtl/ltr/whatever.
Fixed documentation a bit.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51979 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoFix common misspellings
lucas [Wed, 8 Sep 2010 03:51:24 +0000 (03:51 +0000)]
Fix common misspellings

Following misspellings were fixed:

accomodate->accommodate
achive->achieve
beacuse->because
caluclate->calculate
cant->can't
carefull->careful
convertion->conversion
dependancy->dependency
dependant->dependent
doesnt->doesn't
existant->existent
extention->extension
fucntion->function
impliment->implement
inital->initial
lenght->length
occured->occurred
occuring->occurring
onyl->only
positon->position
possibilty->possibility
postion->position
proccessing->processing
proccess->process
propogate->propagate
recieve->receive
sucessive->successive
teh->the
ther->there
throught->through
thsi->this
wasnt->wasn't
whcih->which
wheras->whereas

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51965 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agosvn:ignore
barbieri [Tue, 7 Sep 2010 15:43:57 +0000 (15:43 +0000)]
svn:ignore

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51954 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agowell dldrawarrays doesnthave a limit according to the gl spec - but
raster [Mon, 6 Sep 2010 13:36:32 +0000 (13:36 +0000)]
well dldrawarrays doesnthave a limit according to the gl spec - but
drivers may implement one. as such add a check. but check and fix
seems to fail for some rendering. wonder why?

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51923 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Update bidi props also when merging nodes and when deleting formats.
tasn [Mon, 6 Sep 2010 12:19:17 +0000 (12:19 +0000)]
Evas textblock: Update bidi props also when merging nodes and when deleting formats.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51922 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas bidi: Fixed docs.
tasn [Sun, 5 Sep 2010 09:45:28 +0000 (09:45 +0000)]
Evas bidi: Fixed docs.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51900 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Added evas_textblock_cursor_content_get.
tasn [Sun, 5 Sep 2010 08:28:58 +0000 (08:28 +0000)]
Evas textblock: Added evas_textblock_cursor_content_get.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51898 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed documentation typos. Patch by Jihoon Kim.
tasn [Sun, 5 Sep 2010 07:04:59 +0000 (07:04 +0000)]
Evas textblock: Fixed documentation typos. Patch by Jihoon Kim.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51896 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoand fix gl engine to actually do map for yuv images! :) fix fix.
raster [Sun, 5 Sep 2010 02:58:30 +0000 (02:58 +0000)]
and fix gl engine to actually do map for yuv images! :) fix fix.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51893 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoUpdate email address across the board.
devilhorns [Sat, 4 Sep 2010 18:32:59 +0000 (18:32 +0000)]
Update email address across the board.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51890 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agomake priority of thread drop linux only - fallback.
raster [Sat, 4 Sep 2010 14:34:23 +0000 (14:34 +0000)]
make priority of thread drop linux only - fallback.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51886 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoback to RR - hope kernel allows us to set minimum realtime pri anyway.
raster [Sat, 4 Sep 2010 05:40:09 +0000 (05:40 +0000)]
back to RR - hope kernel allows us to set minimum realtime pri anyway.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51877 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agofix async enabled evas on a single core system.
raster [Sat, 4 Sep 2010 04:12:46 +0000 (04:12 +0000)]
fix async enabled evas on a single core system.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51876 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agodont call mouse out or in if already in or out in evas. causing
raster [Sat, 4 Sep 2010 03:53:34 +0000 (03:53 +0000)]
dont call mouse out or in if already in or out in evas. causing
infinite loops in edje_viewer! bad!

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51875 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoalso rr->other
raster [Sat, 4 Sep 2010 00:54:42 +0000 (00:54 +0000)]
also rr->other

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51872 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agosched other pls. :)
raster [Sat, 4 Sep 2010 00:52:24 +0000 (00:52 +0000)]
sched other pls. :)

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51871 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agopreload in bg. low pri th.
raster [Fri, 3 Sep 2010 10:07:20 +0000 (10:07 +0000)]
preload in bg. low pri th.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51859 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agofix raols bug he found map + del obj.
raster [Fri, 3 Sep 2010 08:44:13 +0000 (08:44 +0000)]
fix raols bug he found map + del obj.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51857 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agogrrrr. slowness bug. fix. now display bug is back.
raster [Fri, 3 Sep 2010 08:23:38 +0000 (08:23 +0000)]
grrrr. slowness bug. fix. now display bug is back.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51855 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoin order to fix some horridibubble things. i had to add 4 api's. ugh.
raster [Fri, 3 Sep 2010 00:06:56 +0000 (00:06 +0000)]
in order to fix some horridibubble things. i had to add 4 api's. ugh.
done. working. now... still 1 more bug. seems to be a changed flag bug
too in dlip 2 test in scrolling list after map anim finished. smaller
bug to deal with now. ugh. bug one bug to the other

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51849 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed char coord set to work better.
tasn [Thu, 2 Sep 2010 13:53:14 +0000 (13:53 +0000)]
Evas textblock: Fixed char coord set to work better.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51842 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: width should remain the same no matter what's the position, removed...
tasn [Thu, 2 Sep 2010 12:01:23 +0000 (12:01 +0000)]
Evas textblock: width should remain the same no matter what's the position, removed that unwanted adjustment.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51838 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: updated docs.
tasn [Thu, 2 Sep 2010 11:57:52 +0000 (11:57 +0000)]
Evas textblock: updated docs.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51837 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed compilation without fribidi.
tasn [Thu, 2 Sep 2010 11:53:40 +0000 (11:53 +0000)]
Evas textblock: Fixed compilation without fribidi.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51836 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock+font engine: Implemented evas_textblock_cursor_geometry_get.
tasn [Thu, 2 Sep 2010 11:49:00 +0000 (11:49 +0000)]
Evas textblock+font engine: Implemented evas_textblock_cursor_geometry_get.
Fixed evas_common_font_char_coords to work correctly with the NULL character in RTL text.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51834 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agofix more map+clip goop.
raster [Thu, 2 Sep 2010 09:40:23 +0000 (09:40 +0000)]
fix more map+clip goop.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51830 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoevas_object_text*_get: return 0 in case of error
caro [Thu, 2 Sep 2010 05:48:23 +0000 (05:48 +0000)]
evas_object_text*_get: return 0 in case of error

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51829 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agonotes. still a bork.
raster [Wed, 1 Sep 2010 22:51:00 +0000 (22:51 +0000)]
notes. still a bork.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51827 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoand fix up some related map+clip+smart render issues along with fixing
raster [Wed, 1 Sep 2010 22:45:30 +0000 (22:45 +0000)]
and fix up some related map+clip+smart render issues along with fixing
map boundary bugs as in trying to fix it i added some bugs.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51826 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocomment++
raster [Wed, 1 Sep 2010 21:38:34 +0000 (21:38 +0000)]
comment++

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51824 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agonow that mostly fixes this map clip issue... ugh. took a while to come up
raster [Wed, 1 Sep 2010 21:37:10 +0000 (21:37 +0000)]
now that mostly fixes this  map clip issue... ugh. took a while to come up
with an elegant solution. also it does add overhead to fix. i might be
able to improve the overhead.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51823 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed compilation without fribidi. <-- For real now. :P
tasn [Wed, 1 Sep 2010 14:26:55 +0000 (14:26 +0000)]
Evas textblock: Fixed compilation without fribidi. <-- For real now. :P

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51809 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed compilation without fribidi.
tasn [Wed, 1 Sep 2010 14:08:23 +0000 (14:08 +0000)]
Evas textblock: Fixed compilation without fribidi.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51808 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas bidi: Changed the design a bit so it'll be easy to implement async-rendering...
tasn [Wed, 1 Sep 2010 13:15:04 +0000 (13:15 +0000)]
Evas bidi: Changed the design a bit so it'll be easy to implement async-rendering with it.
Fixed a couple of issues with unwanted cleans and generally data corruption.
Cleaned up async-rendering.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51806 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years ago * evas: reenable pipe rendering almost work for me with 16 CPU.
cedric [Wed, 1 Sep 2010 12:25:25 +0000 (12:25 +0000)]
* evas: reenable pipe rendering almost work for me with 16 CPU.
Just don't forget to Spank Tasn.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51805 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoEvas textblock: Fixed updating BiDi properties when we delete.
tasn [Wed, 1 Sep 2010 08:25:19 +0000 (08:25 +0000)]
Evas textblock: Fixed updating BiDi properties when we delete.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51801 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoanity check options for option silliness.
raster [Tue, 31 Aug 2010 22:36:35 +0000 (22:36 +0000)]
anity check options for option silliness.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51789 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoaaaaaaaaaaaaaaargh! where's me rum! :(
raster [Tue, 31 Aug 2010 22:16:08 +0000 (22:16 +0000)]
aaaaaaaaaaaaaaargh! where's me rum! :(

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51788 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agocolor interpolation? never used. did nothing. not needed. go go go!
raster [Tue, 31 Aug 2010 13:29:17 +0000 (13:29 +0000)]
color interpolation? never used. did nothing. not needed. go go go!

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51781 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoevas_smart_new - deprecated api gone.
raster [Tue, 31 Aug 2010 12:01:46 +0000 (12:01 +0000)]
evas_smart_new - deprecated api gone.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51780 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years agoimaging api--. as promised.
raster [Tue, 31 Aug 2010 11:58:51 +0000 (11:58 +0000)]
imaging api--. as promised.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51779 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years ago Evas SVG loader: if both width and height scale dimensions are given, we want...
illogict [Tue, 31 Aug 2010 08:14:39 +0000 (08:14 +0000)]
   Evas SVG loader: if both width and height scale dimensions are given, we want to use the biggest computed dimensions.

   Fixes some ugly views if width and height factors are too much different.

   See r51774.

   Thank you Cédric!

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51775 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33

14 years ago Evas JPEG loader: if both width and height scale dimensions are given, we want to...
illogict [Tue, 31 Aug 2010 06:50:48 +0000 (06:50 +0000)]
 Evas JPEG loader: if both width and height scale dimensions are given, we want to use the biggest computed dimensions.

 Fixes some ugly views if width and height factors are too much different.

git-svn-id: svn+ssh://svn.enlightenment.org/var/svn/e/trunk/evas@51774 7cbeb6ba-43b4-40fd-8cce-4c39aea84d33