[Soc-2009-dev] Raytrace api

André Pinto andresusanopinto at gmail.com
Fri Jun 26 20:29:00 CEST 2009

===Week 5 report===
==This week==
*Improved stuff on the test-script (theeth fast-reviewed it and gave some
nice clues)
*Fixed bugs that were found last week by the test-script
*Added documentation on wiki
*Started selection on test scenes and identified a few test cases (
<= althought I still don't have the scenes to test those.
Anyone is welcome to contribute for scenes that match the described setups.

==Next Week==
*add counters/statistics on render code and look on metadata/stamp export
*get other machine ready to run tests
*gather more scenes

I am getting lost on the render pipeline (or at least losing time on reading
it), can anyone give me clues on:
*Where should I export counters from raytree build and from each raytrace
*Where does the shading code starts (and in what way can I better isolate
raytrace and raytree creation time from the other processing time)?
*Where can I merge counters from each thread?

I have spend some time trying to setup the test system on an ssh account
(that runs on Xeon3.2, a quad core.. since testing on my eeepc or my desktop
would just be too slow).

Altought I have quite a few limited space on that account, so I need to
compile on other machine and move to there. I am having some troubles with
that.. for missing libs. This would be fixed by having all libs static..
But do all libs have WITH_BF_STATICblabla options?
Including the libgettext ?
How can I create a blender with the least or even zero dependencies?

About the bug I reported on SUN lights..
there are 2 options:
 *make the new code behave wrong like old render code
 *just have it fixed
Anyway this isn't a priority atm.. so it can be address later.

I dont exactly have any show-off until now...
Appart from instance-support (there was already a bn post on that)
Speedup is still not done and very inconclusive atm since I haven't tested
any real-case.
I can hand-craft a scene to show speedup (like 13x times faster :p).. but
that doens't counts.

I am on schedule (or maybe 1 week ahead).. and by the midterm I will have
all issues sorted out and be fully ready to start the optimization part, as
the first part was:
 *raytrace recode (instance deliverable is done)
 *test-script and test-environ (close to be ready)

I must get more ahead of schedule, for the break in end of july.
I already have dates for that 20july to 31july.

*my last exam is day 29 june. At last!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/soc-2009-dev/attachments/20090626/51a8ffbf/attachment.htm 

More information about the Soc-2009-dev mailing list