Generated by StatCvs v0.2-dev
Pages: << Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 Next >>
Summary Period: 2001-05-06 to 2001-05-01 (Commits 4745-4794 of 12744)
- Moved G3D->SetPerspectiveAspect() to init in csEngine::Draw()
instead of everywhere else.
- Removed the obsolete inv_aspect field from the structures
sent to iGraphics3D.
- Fixed the python plugin again.
- Started work on improved DrawTriangleMesh for OpenGL. OpenGL
renderer will now keep a 3D frustum in camera space and recalc
this when needed (not finished yet).
56 lines of code changed in:
Update.
5 lines of code changed in:
Added Martins strserv plugin to global build process.
Fixed one of my own typo's :-)
1 lines of code changed in:
Beginings of adding Martins new strserv plugin
to CS proper and MSVC workspace.
1 lines of code changed in:
added string server plugin and new pseudo-RTTI system. Both are still unused.
0 lines of code changed in:
- Added HardTransform() to iMeshFactoryWrapper and
csMeshFactoryWrapper.
- Received a new shuttle from Thomas Hieber. It is now used
as the main shuttle in the demo.
- Made softer transitions in the demo between camera switches.
43 lines of code changed in:
Update.
7 lines of code changed in:
- Fixed a bug in the new HEIGHTGEN keyword. Added SIZE and
PARTSIZE to make it easier.
- Did a quick hack in the terrain engine so that it will correctly
align for textures of size 64! Need to make it more general so
that it works for all texture sizes though.
5 lines of code changed in:
Codewarrior is very picky with the floating point constants passed to templated routines. Making sure they all are floats.
14 lines of code changed in:
- Fixed a bug with 3D sprites which would crash on lighting if
the sprite contained vertices that were not used in triangles.
The function to calculate normals now sets the normal to 1,0,0
in such case (previously it was undefined which could cause
crashes).
- If a sprite has only one frame then tweening will be disabled
by default. This should increase performance a little for them.
7 lines of code changed in:
Fixed the makefiles for several apps for the new dependency on CSFX.
3 lines of code changed in:
- Added iTerrFuncState::SetHeightMap() to set a heightmap instead
of a function. Any iImage can be used. It will average the
red, green, and blue components. Size of the image is also not
important (except for terrain accuracy of course). Linear
interpolation is used to get inbetween values.
- Added HEIGHTMAP keyword to the terrfunc loader.
- Added new HEIGHTGEN keyword to the loader. With this keyword it
is now possible to use the new class from Wouter to dynamically
generate textures in combination with a heightmap.
17 lines of code changed in:
Fix
34 lines of code changed in:
Fixed a typo in the simple tutorial in the docs.
A const char const* argv[] was missing the first const.
This caused Initialize not to override SysSystemdriver Initialize,
and this caused people to go crazy at the broken example. :-)
4 lines of code changed in:
Update.
4 lines of code changed in:
Performed cast to signed on "-dt" and changed to "-(signed)dt".
This eliminates MSVC warnings and seemingly has no side affects etc.
4 lines of code changed in:
Minor fullscreen fix
1 lines of code changed in:
Fixed compile errors. Removed old GetClipper ()
and SetClipper () functions as are now auto inlined
within "null_g3d.h", with todays changes.
0 lines of code changed in:
Small change.
4 lines of code changed in:
Update.
17 lines of code changed in:
- Changed the do_clip parameter in G3DTriangleMesh to clip_portal
and clip_plane.
There are two kinds of clipping: portal clipping (clipping to the
boundaries of the last portal we went through) and portalplane
clipping (clipping to the plane of the portal itself). Both are
handled seperatelly.
There are three cases for both portal clipping as portalplane
clipping (these are set in the fields clip_portal and clip_plane):
- No clipping required (CS_CLIP_NOT): in this case the
engine is sure that the mesh will not need any clipping.
- Clipping needed (CS_CLIP_NEEDED): in this case the
engine thinks that clipping may be needed. Depending on
the type of clipper (see later) this clipping is optional
or not.
- Clipping for toplevel (CS_CLIP_TOPLEVEL): in this case the
engine knows that no clipping is needed to the current
clipper but clipping may be needed for the toplevel
clipper (or Z=SMALL_Z plane if used for plane clipping).
The engine will give a clipper to G3D. There are three cases:
- Optional clipper (CS_CLIPPER_OPTIONAL): the clipper is
optional. This means that the z-buffer contents can
be trusted to clip the object. But a renderer for which
overdraw is slow can choose to do a lazy clip or exact
clip if it wants.
- Toplevel clipper (CS_CLIPPER_TOPLEVEL): the clipper is
the toplevel clipper (i.e. the outermost portal, usually
close to the size of the window or screen). In this
case clipping is required but the renderer may have
some special ways to do this (like initializing the
Z-buffer around the borders of the portal).
- Required clipper (CS_CLIPPER_REQUIRED): in this case the
clipper MUST be used (if CS_CLIP_NEEDED is set in the
mesh) and clipping must be exact (i.e. triangles may
not overlap the clip area).
- Did the following changes to iGraphics3D:
- Changed SetClipper() and GetClipper() to work with iClipper2D.
This avoids a copy of the clipper.
- Added 'cliptype' parameter to SetClipper() which will be
one of the flags CS_CLIPPER_?
- Added GetClipType() function.
- Added SetNearPlane(), GetNearPlane(), ResetNearPlane(), and
HasNearPlane() to set the near clipping plane.
- Fixed all 3D renderers except glide and Direct3D. These are now
totally outdated. Note that the handling in the renderers is not
yet smart. This is the framework. Now I need to do some work
to use the framework for very efficient mesh handling.
- Changed the engine to call SetClipper() and SetNearPlane() every
time we go through a portal and at init time. This is much faster
than letting every single mesh object do this.
- Changed csRenderView and iRenderView::ClipBBox() to return the
right clip_portal and clip_plane flags depending on the given
bounding boxes (screen and camera space).
- Extended csCurve::GetScreenBoundingBox() so that it also returns
the camera bounding box.
- Fixed all mesh and terrain plugins to use the new system when
talking to DrawTriangleMesh.
- Changed terrfunc terrain plugin to be more clever about clipping
now. This also means that the bug with disappearing polygons
in front of the camera is now completely gone (touch wood!).
381 lines of code changed in:
Changed Win32 OpenGL2D driver from using a
standard popup window. to using a complete
and correct caption window with full system
menu support.
3 lines of code changed in:
* IMPORTANT *
Due to forthcoming changes in iGraphics3D. The
Win32 Direct3D and it's related Direct Draw plugin
are being temporarily removed from the Crystal Space
Win32 build process.
This change affects The MSVC and MingW ports.
Note: Old MSVC ".dsp" project files will be removed
by the CRON daemon, when it runs tonight.
Regards
Philip Wyett - Win32 Platform Maintainer
1 lines of code changed in:
Removed test for elapsed_time == 0 because this seems to give
problems for FPS counting.
0 lines of code changed in:
Update.
1 lines of code changed in:
Also enabled stencil for X.
1 lines of code changed in:
fixed some warnings
1 lines of code changed in:
Small fix with allocate of 0 bytes.
3 lines of code changed in:
Fixed bumptest to use DrawPolygon instead of DrawPolygonFX.
And the light moves faster.
10 lines of code changed in:
testcommit
5 lines of code changed in:
Disabled memory debugger that Wouter accidently turned on.
1 lines of code changed in:
Removed support for multiple textures in G3DTriangleMesh.
This is obsolete and unused and unneeded since we have
real multi-texture support in materials now.
31 lines of code changed in:
Update.
8 lines of code changed in:
Fix.
1 lines of code changed in:
- Enabled stencil buffer for Windows OpenGL canvas.
- Experimented a bit with the stencil buffer for clipping and
DrawTriangleMesh. Not enabled by default yet since I have to
experiment with it.
58 lines of code changed in:
Fixed jerkiness in walktest with my opengl driver.
Although it limits the amount of redraws to the resolution of the timer
(about 100 or 50 fps, depending on timer accuracy), the fps counter reports
the total amount. I get about 100000 fps reported :-)
4 lines of code changed in:
Fixed several cases in the parsers where a boolean was read
but put in 'int'. This doesn't work since a recent change where
ScanStr works correctly on the 'bool' type. If %b is used for
ScanStr then the parameter should be 'bool' and not 'int'.
This mostly affected the LIGHTING() flag in several of the mesh
object loaders.
6 lines of code changed in:
Fixed the color bug in OpenGL DrawTriangleMesh with multi-
texturing.
6 lines of code changed in:
Forgot to commit.
17 lines of code changed in:
- FINALLY fixed the very bad bug which completely locked up
my machine when a sprite was visible in fog using OpenGL
on an NVidia GeForce. It appears that the NVidia doesn't like
the use of glTexCoordPointer with 1 as the first parameter
(i.e. only one texture component). So fog and DrawTriangleMesh
finally works right with OpenGL!
- Also fixed fog for DrawPolygonFX itself. DPFX was using perspective
correct texture mapping but fog was not doing that.
- Added support for multi-texturing in OpenGL renderer for DrawPolygon,
DrawPolygonFX, and DrawTriangleMesh. data/flarge shows two
examples (one on the initial room and one on the sprites in the
large hall). Software renderer currently ignores the extra
textures. There is still a bug in the DrawTriangleMesh version
though. The color of gouraud shading seems to shift.
410 lines of code changed in:
fixed a lot of cases where the system driver was IncRef'ed and DecRef'ed,
which should not happen
17 lines of code changed in:
Fixed small bug in the opengl rendrerer introduced yesterday.
1 lines of code changed in:
- Extended texture layer support in the engine so that maximum
four additional layers are supported (this means one base
texture and four extra). Note that none of the 3D renderers
currently support this.
- The material manager in the 3D renderers is now capable of
holding four texture layers too. They are still not used
however.
32 lines of code changed in:
Added a (temporary?) bugfix to the mingw engineplugin compilation problem.
The engine was not being included. This fix, includes $(OUT)engine$(O) in
the linkage for all platforms. (this file is also present in libcsengine.a)
2 lines of code changed in:
Fixed MazeD for latest material changes.
1 lines of code changed in:
Update.
10 lines of code changed in:
Added CS_FX_TILING to the bumptest bump layer. It works perfectly nice in
software now!
1 lines of code changed in:
Added small fix to demo.
if(elapsed_time == 0) return; from NextFrame().
This avoids (almost 99%) of the jerkyness I experienced in opengl mode.
The reason was that high framerates and bad timer caused 0 elapsed time,
and a redraw of exactly the same frame, filling up queues inside my bad
driver, and causing a tug when real work needed to be done.
5 lines of code changed in:
- Extended the iMaterial structure in preparation of multi-texturing
support.
- Extended the loader of materials in preparation of multi-texturing
support. There is now a new LAYER keyword in the definition
of a MATERIAL. In LAYER you can specify the texture, scale, shift,
and mixmode to use for that texture layer.
72 lines of code changed in:
Fixed a bug in software renderer so that tiling really works.
25 lines of code changed in:
Pages: << Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 Next >>
Generated by StatCvs v0.2-dev