On 25/04/14 09:03, Stefan Schmidt wrote:
> On Fri, 2014-04-25 at 00:18, Taekyun Kim wrote:
>> hermet pushed a commit to branch master.
>>
>> http://git.enlightenment.org/core/efl.git/commit/?id=8fda63173063e03e1d48c8b026ecf8b94298162c
>>
>> commit 8fda63173063e03e1d48c8b026ecf8b94298162c
>> Author: Taekyun Kim <[email protected]>
>> Date:   Fri Dec 27 16:56:30 2013 +0900
>>
>>      Evas: 3D: Introducing 3D scene rendering features
>>
>>      Enable 3D features using --enable-evas-3d=yes when configuring.
>>      APIs are exposed through Evas_3D.h.
>>      Currently, evas-3d is being supported only on gl_x11 engine.
>
>
> Missing the @feature tag here for such a big thing.
>
>>      Conflicts:
>>
>>              src/lib/evas/Evas_Eo.h
>> ---
>>   configure.ac                                       |   17 +
>>   src/Makefile_Evas.am                               |   26 +
>>   src/lib/evas/Evas.h                                |    1 -
>>   src/lib/evas/Evas_3D.h                             |  289 ++++
>>   src/lib/evas/canvas/evas_3d_camera.c               |  159 ++
>>   src/lib/evas/canvas/evas_3d_light.c                |  273 ++++
>>   src/lib/evas/canvas/evas_3d_material.c             |  221 +++
>>   src/lib/evas/canvas/evas_3d_mesh.c                 |  828 +++++++++++
>>   src/lib/evas/canvas/evas_3d_mesh_loader_md2.c      |  440 ++++++
>>   src/lib/evas/canvas/evas_3d_node.c                 | 1162 +++++++++++++++
>>   src/lib/evas/canvas/evas_3d_object.c               |  129 ++
>>   src/lib/evas/canvas/evas_3d_scene.c                |  622 ++++++++
>>   src/lib/evas/canvas/evas_3d_texture.c              |  440 ++++++
>>   src/lib/evas/canvas/evas_image.eo                  |   27 +
>>   src/lib/evas/canvas/evas_object_image.c            |  194 ++-
>>   src/lib/evas/canvas/evas_object_main.c             |   21 +-
>>   src/lib/evas/include/evas_3d_private.h             |  377 +++++
>>   src/lib/evas/include/evas_3d_utils.h               | 1526 
>> ++++++++++++++++++++
>>   src/lib/evas/include/evas_private.h                |   32 +
>>   src/modules/evas/engines/gl_common/evas_gl_3d.c    | 1332 +++++++++++++++++
>>   .../evas/engines/gl_common/evas_gl_3d_common.h     |   37 +
>>   .../evas/engines/gl_common/evas_gl_3d_private.h    |  145 ++
>>   .../evas/engines/gl_common/evas_gl_3d_renderer.c   |  284 ++++
>>   .../evas/engines/gl_common/evas_gl_3d_renderer.h   |    6 +
>>   .../evas/engines/gl_common/evas_gl_3d_shader.c     | 1482 
>> +++++++++++++++++++
>>   .../evas/engines/gl_common/evas_gl_common.h        |    2 +
>>   src/modules/evas/engines/gl_x11/evas_engine.c      |  202 +++
>>   src/modules/evas/engines/gl_x11/evas_engine.h      |   19 +
>>   src/modules/evas/engines/gl_x11/evas_x_main.c      |   90 ++
>>   .../evas/engines/software_generic/evas_engine.c    |   16 +
>>   30 files changed, 10394 insertions(+), 5 deletions(-)
>
>
> Puh, this thing is HUGE. A new feature with over 10K lines of code in
> a core component like evas without any tests 50h before closing of the
> merge window. I was really hoping to avoid such things.
>
> It also just failed on jenkins mingw build. Funny enough because it
> should only build when enabled during configure, no? But this is so
> new nobody has enabled it which means it builds unconditional right
> now and fails at least on mingw...
>
> https://build.enlightenment.org/job/changely_efl_mingw_x86_64/lastFailedBuild/console
>
> Next question is what we do about all this new API's. I stopped
> counting at 40 new function calls with EAPI. They don't seem to be
> hidden behind any BETA configure flag. Thus they are fully supported
> and can't go away until 2.x. I can only hope they have been reviewed
> internally before. Is that the case?
>
> I have to say I'm nervous about this big change coming in so late. We
> basically only have three weeks now to find the biggest problems.
> Taekyun/Hermet I expect you two spending some time on this code during
> stabilization. Look at the next coverity reports for issues on this
> code. With our current ration of 0.5 issues on 1000 lines of code it
> should have at least 5. But that is for old debugged code not for
> something new. Guess we will see more.
>
> I would also be way less nervous about this if it would have come with
> tests.

One more thing about this change which I haven't really reviewed, just 
wanted to provide a reminder given the name. You have to remember that 
because of autogenerated c++ bindings (well, actually, because of c++, 
and many other language's bindings in general), starting a function name 
with a number is not allowed.

Because evas_object_image_3d_foo() will become obj->3d_food(), you have 
to be extra careful with naming.

--
Tom.



------------------------------------------------------------------------------
Start Your Social Network Today - Download eXo Platform
Build your Enterprise Intranet with eXo Platform Software
Java Based Open Source Intranet - Social, Extensible, Cloud Ready
Get Started Now And Turn Your Intranet Into A Collaboration Platform
http://p.sf.net/sfu/ExoPlatform
_______________________________________________
enlightenment-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Reply via email to