[Soc-2013-dev] Weekly Report #12 BGE LoD and Bug Fixing

[Daniel Stokes]

== Progress ==
This week I spent some time updating my Wiki page with information
regarding my GSoC progress:
http://wiki.blender.org/index.php/User:Kupoman/GSoC_2013

I plan to record a video high lighting some of the progress. I am new to
doing screen captures and videos of that kind, so I did spend some time
looking at resources (mostly software). I will try to get a video up
sometime this weekend.

As far as bugs go, the following 2 bugs have fixes in Trunk:
VSMs destabalize with transparent receivers (reported on IRC by SamCameron)
[#32360] Standalone Player switches not
working<https://projects.blender.org/tracker/index.php?func=detail&aid=32360&group_id=9&atid=306>(some
more back and forth with the submitter is required before closing)

I have a fix for [#35692] Vertex colors fail when mesh contains quads and
tris<https://projects.blender.org/tracker/index.php?func=detail&aid=35692&group>,
but I want to figure out why it fixes the problem before committing to
Trunk. The fix is to remove a glColor call from the vertex array drawing
code when it is using a color array. ATI cards may have problems with
defining color in two ways.

I have closed the following two reports for reasons described in the
reports:
[#28743] Regression test file webskategirl_bullet.blend act different on
svn 2.59<https://projects.blender.org/tracker/index.php?func=detail&aid=28743&group_id=9&atid=306>[#32793]
strange behaviour of runtime command line options (windows
afaik)<https://projects.blender.org/tracker/index.php?func=detail&aid=32793&group_id=9&atid=306>

I have also investigated a few other bugs:
[#28888] BGE Material - Alpha Blend: Add - not respecting object.color alfa
value<https://projects.blender.org/tracker/index.php?func=detail&aid=28888&group_id=9&atid=306>
This does not look like a bug to me, but rather how Add alpha works.
Waiting for response from submitter.
[#35950] Standalone crashes when "start game from a
file"<https://projects.blender.org/tracker/index.php?func=detail&aid=35950&group_id=9&atid=306>
There does not seem to be much luck reproducing this problem. Would be
great if we could get an OSX user to try reproducing it. I still need to
give it a try on my laptop (Linux with ATI graphics).
[#33024] Masked physics objects collide with one another on
startup<https://projects.blender.org/tracker/index.php?func=detail&aid=33024&group_id=9&atid=306>
I found a fix for this bug, but later discovered a better fix can be found
in the following patch: [#33710] Expose collision mask and group
functionality to the Python
API<https://projects.blender.org/tracker/index.php?func=detail&aid=33710&group_id=9&atid=127>.
Once that patch makes it to Trunk, we should be able to close the bug
report.

== Next Week ==
Continue working on bugs and updating my wiki page. With less stuff going
on next week, I think I should be able to close 7-12 reports.

== Questions/Problems ==
I am still looking for a review on my LoD work from a Blender dev. The BGE
devs appear to be fine with the changes to the BGE code base. I think it is
on Campbell's to do list, but he is also a busy guy. The code review link:
https://codereview.appspot.com/12216043/

Regards,
Daniel Stokes
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/soc-2013-dev/attachments/20130906/4734a13e/attachment.htm