Summary Period: 1999-11-17 to 2003-04-08
[root]/CS/scripts
aws
(2 files, 883 lines)
blender
(0 files, 0 lines)
blend2cs
(22 files, 915 lines)
exporter_editor
(2 files, 489 lines)
unmaint
(6 files, 949 lines)
cs-config
(8 files, 655 lines)
deb
(2 files, 143 lines)
core
(0 files, 0 lines)
debian
(24 files, 1199 lines)
xpm
(4 files, 205 lines)
data
(0 files, 0 lines)
debian
(15 files, 402 lines)
hammer
(1 files, 237 lines)
jamtemplate
(9 files, 512 lines)
max
(6 files, 6189 lines)
perl5
(1 files, 15845 lines)
python
(19 files, 17273 lines)
glext
(18 files, 316 lines)
rpm
(1 files, 121 lines)
xml
(1 files, 924 lines)
Total Lines Of Code:
3 (2003-07-11 20:29)
Author | Changes | Lines of code | Lines per change |
---|---|---|---|
Totals | 39 (100.0%) | 1270 (100.0%) | 32.5 |
azverkan | 35 (89.7%) | 1249 (98.3%) | 35.6 |
andyz | 1 (2.6%) | 15 (1.2%) | 15.0 |
sunshine | 2 (5.1%) | 3 (0.2%) | 1.5 |
matzebraun | 1 (2.6%) | 3 (0.2%) | 3.0 |
added the Jamfiles
3 lines of code changed in:
Moved python scripts to scripts/python
0 lines of code changed in:
-*- Changed perfstat plugin to compute fps once per 1/2 second rather than once
per 10 frames. This will compute more accurate values in high-fps
environments and also makes fps numbers more steady. Also
iPerfStat::SetResolution sets the interval in msecs rather than in frames.
-*- Changed FPS display in walktest to display just two numbers after dot.
I don't think you really care whenever you have 20.111 or 20.112 fps :)
-*- Created the iEvent interface. It is really a very special interface since
its the first time (as far as I know) an SCF interface with member
variables is used. However, this is supposed to work on all platforms.
Be warned though.
-*- csEvent is now just a particular implementation of iEvent. You can create
your own objects implementing iEvent interface. PLEASE always use
QueryInterface (or rather QUERY_INTERFACE macro) to get access to
extended fields; NEVER do direct typecasts.
-*- Added a new halo generation routine - csutil/halogen2.cpp. It is not used
anywhere yet though. The algorithm has been borrowed from GIMP (`SuperNova'
plugin).
-*- Started to work on MazeD material support. Not finished though (even not
compilable).
-*- Improved material support. The MATERIALS() statement in world file now
looks like this:
MATERIALS
(
MATERIAL 'material name'
(
TEXTURE ('texture name')
COLOR (float, float, float) ; R,G,B = 0..1
DIFFUSE (float) ; 0..1
AMBIENT (float) ; 0..1
REFLECTION (float) ; 0..1
)
...
)
I have added the GetFlatColor() method to iMaterial and iMaterialHandle.
For now the flat_color_X members in iGraphics3D polygon structures are
still present; however I hope one day they will be gone and a material
handle will be always present, since materials also support flat color.
For now only software renderer supports materials with flat colors;
I hope with time other renderers will be also updated for this.
-*- Renamed RGBcolor -> csRGBcolor and RGBpixel -> csRGBpixel.
In general, we should think where to put all library-independent header
files such as rgbpixel.h and cspixmap.h. I would say they could live in
main include directory since those classes are absolutely independent
and completely inline.
-*- Changed the way engine defines and uses texturing. There are four polygon
texturing types now: POLYTXT_NONE (no texturing, useful when using
materials without textures -- it defines no texturing parameters at all
thus saves memory), POLYTXT_FLAT (flat-shading, only the angle between
light and polygon normal is considered if CS_POLY_LIGHTING flag is set,
and entire polygon is painted with one lighting value; every vertex has
an U/V pair associated), POLYTXT_GOURAUD (every polygon vertex has a color
and those colors are interpolated across scanlines) and finally
POLYTXT_LIGHTMAP (a polygon which has an associated lightmap and a texture
plane). Now it is possible to define and use objects which have flat-color
but still receive light with Gouraud as well as without Gouraud
interpolation; in general this open a whole new area for experimenting.
-*- New keyword in world file inside the POLYGON () context:
TEXTURING (NONE | FLAT | GOURAUD | LIGHTMAP)
This defines one of the texturing modes above.
-*- MIXMODE() keyword is applicable to all texturing types except LIGHTMAP.
Alpha transparency for such polygons can be defined in two ways:
POLYGON (... ALPHA (50) MIXMODE (MULTIPLY2))
or
POLYGON (... MIXMODE (MULTIPLY2 ALPHA (0.5)))
To be honest, I don't like having two different ways to specify alpha.
I don't like the different scale either (0..1 and 0..100).
Polygon always keeps alpha as a value 0..255 (and iGraphics3D always
receives alpha in this range, the old 0..100 scale is gone).
In general, I would say that giving colors and alpha as floating-point
value looks like a bad idea. I would store both them as 0..255 value,
or as a floating-point 0..1 value. Maybe it makes sense to change this
in the new loader?
-*- DECLARE_GROWING_ARRAY does not have a reference counter anymore. If you
need a reference counter, use DECLARE_GROWING_ARRAY_REF. This gives
slight less memory usage in some places. I have changed thing and
polygon template objects to use growing arrays for vertices. This does
not change memory footprint and gives better readable code (and maybe
a little less binary size).
-*- Simplified polygon template class a lot by moving most state variable
into a csFlags variable. Added support for COLORS() keyword in polygon
templates. GOURAUD keyword is now obsolete in both normal polygons and
polygon templates; use TEXTURING (GOURAUD) instead.
Overall results for flarge.zip:
Was: World geometry is using 1526938 bytes.
Now: World geometry is using 1465962 bytes.
----
Gain: 60976 bytes (flarge contains just 5 thing templates)
-*- In addition to TEXNR() everywhere it appears MATTERIAL is accepted as
a synonym; one day TEXNR should die.
15 lines of code changed in:
Yes Jorrit, I am restarting my Crystal Script work :)
New layout is interface based, which allows for more dynamicism
and lower runtime memory requirements.
385 lines of code changed in:
Fixed csPython so that it can once again compile following recent changes
to SCF.
3 lines of code changed in:
Added iSCF support for use with plugin dlls
38 lines of code changed in:
Fixed for console mode
8 lines of code changed in:
Small Updates, waiting for Andrew for rest
32 lines of code changed in:
Added Just In Time Debugger for Python
34 lines of code changed in:
Map displays
14 lines of code changed in:
Removed CSG_Add
3 lines of code changed in:
The Sector loads from file
121 lines of code changed in:
Fixed for pickle
3 lines of code changed in:
Added support for non C pickle module
10 lines of code changed in:
Unreal Map loading runs, bu geometry is messed up
36 lines of code changed in:
almost got unreal working
357 lines of code changed in:
Starting Python Makefile & SWIG combine
11 lines of code changed in:
Vector3 running
47 lines of code changed in:
Added util script
3 lines of code changed in:
Small fixes
68 lines of code changed in:
(2 more)
Generated by StatCvs v0.2-dev