Summary Period: 2002-03-28 to 2002-10-02
36 (0.0%)
3179 (0.1%)
Directory | Changes | Lines of code | Lines per change |
---|---|---|---|
Totals | 36 (100.0%) | 3179 (100.0%) | 88.3 |
CS/plugins/video/renderer/opengl/ | 18 (50.0%) | 2778 (87.4%) | 154.3 |
CS/libs/csengine/ | 1 (2.8%) | 168 (5.3%) | 168.0 |
CS/libs/csengine/objects/ | 2 (5.6%) | 101 (3.2%) | 50.5 |
CS/include/csengine/ | 4 (11.1%) | 70 (2.2%) | 17.5 |
CS/include/ivideo/ | 2 (5.6%) | 15 (0.5%) | 7.5 |
CS/include/imesh/thing/ | 1 (2.8%) | 12 (0.4%) | 12.0 |
CS/include/iengine/ | 1 (2.8%) | 11 (0.3%) | 11.0 |
CS/plugins/video/renderer/common/ | 3 (8.3%) | 8 (0.3%) | 2.6 |
CS/plugins/mesh/thing/persist/standard/ | 1 (2.8%) | 6 (0.2%) | 6.0 |
CS/plugins/engine/iso/ | 1 (2.8%) | 5 (0.2%) | 5.0 |
CS/include/csgeom/ | 1 (2.8%) | 4 (0.1%) | 4.0 |
CS/apps/examples/simplevp/ | 1 (2.8%) | 1 (0.0%) | 1.0 |
Small Fix.
Added return false if vfs can't be loaded.
1 lines of code changed in:
Oriol Esteve has done the patch to add Smoothing fot things... to enable it just set the <smooth /> in the params section of the thing. It takes a bit more to recalc the lights, but i think it worths.
This code is stable but experimental, but feel free to play.
172 lines of code changed in:
Little optimization for DPM.
With previous implementation all the lightmap in a FASTMESH superlightmap were recalculated (because a dynamic light could affect the mesh)
Now the mesh polybuffer is marked as lightmaps dirty and only when the mesh has dirty lightmaps recalculates the supelightmap.
88 lines of code changed in:
Oriol Esteve has done some hacks to try to avoid the opengl shadow bug.
This code is experimental though. It has been tested in some levels and
seems to place the lightmaps correctly and doesn't crash (and well, that's important :).
It seems stable, but is still under observation.
Greetings,
Toni
15 lines of code changed in:
Fixed a little bug:
IsZero method compared the components of the vector with 0. Now it has a default float parameter named "precision".
Its default value is SMALL_EPSILON (0.000001f), but other values can be passed.
As a precondition precision value should be > 0, the test done is
ABS(x) < precision && ABS(y) < precision && ABS(z) < precision, so if
precision value is negative this will return always true.
It could be solved typing ABS(precision) but this is no needed if we're careful.
Enjoy
Toni
4 lines of code changed in:
Modifications to allow DrawPolygonMesh render non-lighted meshes
712 lines of code changed in:
Modifications to optimized DrawPolygonMesh.
To use DrawPolygonMesh just type FASTMESH() inside
PARAMS section.
This optimizations are good for High polygon models (a few
thousand faces) and work even better if the number of materials
is keeped low (1 is perfect, after that performance goes down)
Issues:
1) Keep in mind this optimizations have high memory requirements.
There are a lot of buffers, so use it carefully.
2) First time the mesh is viewed it has to calculate the poly buffer,
so it takes 1-2 secs to draw. I think polybuffers can be precalc
before the mesh has to be drawn for 1st time, so perhaps that could
be a good way to avoid the "frozen seconds"
Known Bugs:
- The mesh have to be lighted, otherwise it crashes
- Dunno exactly why, but for high poly models fog doesn't
draws correctly, but it works perfect for my 1000 faces sphere :)
Known Bugs 2: The one that will make Jorrit kills me :)
I haven't updated the flush transforms, so it still
flushes the lightmap queues for every draw :|, but well
i had to meet the 0.94 deadline, so...
Unknown Bugs:
- As the title says, they're unknown... so i don't have
any idea, but i'm pretty sure they're there, waiting...
1981 lines of code changed in:
difications to optimized DrawPolygonMesh.
To use DrawPolygonMesh just type FASTMESH() inside
PARAMS section.
This optimizations are good for High polygon models (a few
thousand faces) and work even better if the number of materials
is keeped low (1 is perfect, after that performance goes down)
Issues:
1) Keep in mind this optimization have high memory requierements.
There are a lot of buffers, so use it carefully.
2) First time the mesh is drawn it has to calculate the buffers, so
it takes 1 second or 2
Known Bugs:
- The mesh have to be lighted (to have lightmaps) otherwise it crashes.
- Dunno exactly why, but for high poly models fog doesn't
draws correctly, but it works perfect for my 1000 faces sphere :)
Known Bugs 2: The one that will make Jorrit kills me :)
I haven't updated the flush transforms, so it still
flushes the lightmap queues for every draw :|, but well
i had to meet the 0.94 deadline, so...
Unknown Bugs:
- As the title says, they're unknown... so i don't have
any idea, but i'm pretty sure they're there, waiting...
206 lines of code changed in:
Generated by StatCvs v0.2-dev