[Soc-2008-dev] Week 3 Report

[Nicholas Bishop]

a) Previous week:
My current goal is getting limited sculpting working on the highest
multires level with the new modifier. (Dealing with sculpting on the
highest level is easier because the displacements don't need to be
subdivided.) I'm getting close to having this working, algorithmically
speaking, but there's still work to be done dealing with
DerivedMeshes.

b) Next week:
I plan to continue working on sculpting on the highest level. The next
step here is to write a proper DerivedMesh for multires (so far I've
just been piggybacking on cddm.) I hope to get this working fairly
quickly, so the next thing I'll work on will be subdividing
displacements.

c) Other:
--

d) Schedule:
Things are looking good; taking advantage of CCGSubsurf is helping
things to move along more quickly.

-Nicholas

On Fri, Jun 13, 2008 at 6:25 PM, Maxime Curioni
<maxime.curioni at gmail.com> wrote:
> Week 3 Report
> =============
>
> This Week
> ------------------
> Like I explained last week, due to the end of the school term, I haven't had
> much time to work on the project. Nevertheless, I still managed to make a
> few changes.
>
> Freestyle now renders the object in the Blender scene (see Suzanne example:
> http://maximecurioni.com/gsoc2008/?p=21). The scene export is done using
> Blender's internal 3ds Python script.  Since the script cannot be executed
> directly, I copied it locally in the local Freestyle tree and modified it so
> that it is executed programatically without UI input.
>
> I also removed some static dependencies, that were formerly fixed at
> compilation.  As of now, the Freestyle runtime uses the environment variable
> FREESTYLE_BLENDER_DIR to determine the different files needed for execution
> (style module, Python 3ds export script, paper texture....). Freestyle can
> now be run without any local modification and on any platform, as long as
> the FREESTYLE_BLENDER_DIR environment variable correctly points to the
> directory source/blender/freestyle and the SWIG wrapper module is compiled
> and linked correctly.
>
> Finally, I modified the view parameters so that the output screen size and
> ratio are correct. Currently, the scene is observed from the point (0,0,1)
> and points at the origin (0,0,0). Also, to make rendering efficient, I added
> a Freestyle initialization function, taking care of instanciating the view
> and the controller just once.
>
>
> Next Week
> ---------
> To finish the first phase, I need to change the view camera parameters to
> reflect those of the actual Blender camera (position and orientation) and
> allow the user to select different style modules from the interface.
>
> I intend to start the second phase the upcoming week. Part of that phase has
> been covered already (utilizing Blender's Python environment). I will need
> to do some code study, to understand how the Python API works in detail (to
> replace SWIG) and how render layers are organized and coded in the render
> pipeline. I'll need to figure out how to render OpenGL commands "within" a
> render layer.
>
> With the amount of code I am going to have to study, I don't expect to code
> extensively. I still hope to have some basic render layer functionality
> working.
>
>
> Issues
> ------
> None
>
>
> Schedule
> --------
> According to schedule, I am a week behind, as I was supposed to complete
> phase 1 last Sunday. If things go properly this week-end, I should have it
> finished for this week-end's meeting.
>
> _______________________________________________
> Soc-2008-dev mailing list
> Soc-2008-dev at blender.org
> http://lists.blender.org/mailman/listinfo/soc-2008-dev
>
>