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: 2001-01-18 to 2001-01-09 (Commits 3895-3944 of 12744)
Fixed Martin's stricmp() in cfgfile.cpp. Note that on linux stricmp
does not exist. strcasecmp is the portable name to use.
3 lines of code changed in:
added a new module for configuration files with another format than INI. This
format uses a flat list of config files with pseudo-hierarchy instead of the
sections in INI files. The new system is intended to replace the current
system soon.
927 lines of code changed in:
- Extended iMeshObject interface with SetVisibleCallback() and
GetVisibleCallback(). This function will essentially provide
the same functionality as SetDrawCallback2() does for csSprite3D.
i.e. callback the application when the object is really (or almost
really) visible. Note that you are not allowed to significantly
change the layout, position or anything else that might affect
visibility in this callback (for obvious reasons).
- Modified all the current mesh object plugins to support this new
callback.
- Added SetDrawCallback() and GetDrawCallback() to csMeshWrapper.
This has the same functionality as the SetDrawCallback() in
csSprite3D. i.e. this callback is called when the object is
considered for drawing. This means that there is a chance it
is visible (but not guaranteed). It is allowed to do minor
changes to geometry here as long as you don't change the sectors
where this object lives.
- Added support for frames and actions to iSprite3DFactoryState.
- Fixed Mazed for my iSpriteFrame changes yesterday.
96 lines of code changed in:
Update.
7 lines of code changed in:
removed current support for threading from the wave-out driver because
it produced strange gaps in sound output. It was a bug in the concept. This
does not mean that threading won't be supported at all, but not the way
it was done till now. Wave-Out is now based on function callback mode,
which is also modified to remove the sound gaps.
84 lines of code changed in:
Eric Sunshine changed the following classes so that they embed particular
SCF interfaces, rather than inheriting them, in order to be
NextStep-friendly:
csSkeleton -> iSkeleton
csSkeletonState -> iSkeletonState
csSkeletonConnection -> iSkeletonConnection
csSkeletonConnectionState -> iSkeletonConnectionState
68 lines of code changed in:
Eric Sunshine made the following changes to document.txi:
-*- Fixed a couple factual errors.
-*- Fixed several Texinfo mark-up problems.
-*- Eliminated all of the over- and underfull `hbox' warnings reported
by TeX.
-*- Fixed the example @example block so that the Texinfo directives are
actually displayed in the output rather than being eaten by the
conversion process.
-*- Minor rewriting and enhancement of various topics.
104 lines of code changed in:
Fixed a syntax error in irix.mak.
1 lines of code changed in:
Fixed a syntax error.
1 lines of code changed in:
- Renamed iskeleton.h to iskel.h to make it work on DOS.
- Extended iskel.h with several new interface to represent the
skeleton structure of a 3D sprite: iSkeletonLimb,
iSkeletonConnection, iSkeleton, iSkeletonLimbState,
iSkeletonConnectionState, and iSkeletonState. iSkeletonBone
is retained and moved to iskelbon.h. It is used by the motion
manager and independent as it could be implemented by other
bone systems.
- The skeleton state classes no longer extend csObject. This is
not needed and it is not useful because skeleton state classes
are going to be moved outside the engine anyway. However, some
applications (like WalkTest) require the ability to attach
data to state objects so I extended the SCF interfaces with
that ability.
- Modifed WalkTest to use the new user-data instead of the
csObject system.
- Extended the iSprite3DFactoryState interface with
EnableSkeletalAnimation() and GetSkeleton(). Also extended
iSprite3DState with GetSkeletonState().
- Also added GetFactory() to iSprite3DState.
- Extended the iSprite3DFactoryState with various functions to access
the vertices, texels, normals, and triangles.
- Added iSpriteFrame SCF interface and let csFrame implement this.
- Added iSpriteAction SCF interface and let csSpriteAction implement
this.
358 lines of code changed in:
Added more error messages to openglwin32 canvas,
5 lines of code changed in:
added some hints to the docs about how to write docs
106 lines of code changed in:
Changed the image-copying rule in docs.mak so that it should once again
work on Windows/DOS. (It used to work, but the command-line for copying
images became too long, at some point, for the DOS command-line input
buffer.)
5 lines of code changed in:
Fixed compile warnings on mingw32 for opengl canvas.
3 lines of code changed in:
Fixed #redefinition errors by explicitly testing for each macro before
#defining it. This helps opengl canvas compilation on mingw32.
31 lines of code changed in:
Removed the broken and unuspported cygwin.mak file.
3 lines of code changed in:
Added the first and incomplete include/imspr3d.h which contains
the iSprite3DState and iSprite3DFactoryState SCF interfaces that
will be used by to access the sprite3d plugin (when it is ready).
11 lines of code changed in:
- Fixed a bug in WalkTest. It would not release the reference
to all created particle systems. This caused memory leaks.
- Fixed another bug in WalkTest where the particle system was
created with the wrong position. Actually, it is the right position
but set through the wrong method. Now correctly uses
movable.SetPosition() to set the position.
18 lines of code changed in:
Fixed a serious bug in csMeshWrapper::SetMeshObject(). This
function would do a DecRef() on the given mesh (possibly
deleting it).
6 lines of code changed in:
Update.
24 lines of code changed in:
- explicitly set default dispdriver to NULL
1 lines of code changed in:
- enabled the perspective patch in DrawPolygonFX (although its a slowdown i'd rather trade that for a correctly mapped texture)
- perspective patch expected 1/z instead of z as the 3rd coo-value. since this seemed not consistent with software renderer i changed it to expect a plain z-value
4 lines of code changed in:
changed default DISPDRIVER to NULL, so none is loaded by default:x
1 lines of code changed in:
slightly changed errormsg when not being able to read from opengl.cfg
1 lines of code changed in:
removing empty and svga displaydrivers for MESA
0 lines of code changed in:
-*- Fixed vfs.cfg so that it works with MacOS/X Server again following
Andrew's change to the meaning of CS_PLATFORM_NAME.
-*- Eric Sunshine added several missing /lev/* entries to vfs.cfg.
33 lines of code changed in:
Added a method to csmeshwrapper to get a pointer to the mesh back again.
2 lines of code changed in:
made similar changes in the make install tareget as the changes in docs
from htm to html. The make install target should work once more.
6 lines of code changed in:
Fixed the makefiles for vsh and vfstest so that they link with the VFS
plugin when built `statically'. These broke for static builds when VFS
was moved into its own plug-in module.
3 lines of code changed in:
Addded backwards compatible to dx3 define. It should compile fine now
for dx8. (Also it still compiles fine with older dx versions)
2 lines of code changed in:
Fixed several bugs in soundraw.cpp where it was applying `delete[]' to
a void*.
9 lines of code changed in:
Moved selection of the software renderer for PLUGINS out of the
platform-specific makefiles (CS/libs/cssys/*.mak) and into CS/mk/user.mak
since it is not specific to any particular platform.
18 lines of code changed in:
Eric Sunshine made the following changes to the project:
-*- Renamed the following project directories:
CS/plugins/meshldr --> CS/plugins/mesh/loader
CS/plugins/meshobj --> CS/plugins/mesh/object
-*- Renamed the following SCF class names, prefixes, and suffixes:
crystalspace.meshldr --> crystalspace.mesh.loader
crystalspace.meshldr.ZZZfact --> crystalspace.mesh.loader.factory.ZZZ
crystalspace.meshobj --> crystalspace.mesh.object
crystalspace...explo --> crystalspace...explosion
crystalspace...fount --> crystalspace...fountain
crystalspace...spr2d --> crystalspace...sprite.2d
-*- Fixed all makefiles, source files, header files, documentation, and
`world' files to account for the directory and SCF class name changes.
112 lines of code changed in:
Update.
14 lines of code changed in:
added better explanation of the Print() function of the 2d and 3d gfx driver
10 lines of code changed in:
added some information to the procedural texture docs
21 lines of code changed in:
-*- Changed the file extension on the HTML files in the docs/html and
docs/pubapi directories from .htm to .html at Andrew's insistence. :-)
-*- Lowered the verbosity level of the build process for the HTML version
of the User's Manual (make htmldoc).
125 lines of code changed in:
Fixed a typo in win32.txi and made several other minor changes.
22 lines of code changed in:
Finally managed to write qsqrt() and qisqrt() correctly, so that both
binutils 2.9.1 and 2.9.5 can grok it.
4 lines of code changed in:
Automated Texinfo @node and @menu repair.
3 lines of code changed in:
changed the win32 docs: grouped msvc5 stuff in one section and added a FAQ entry
68 lines of code changed in:
fixed some minor naming and comment flaws in the idungeon interface file
8 lines of code changed in:
committed a patch to win32.cpp from Ed Halley
43 lines of code changed in:
-*- Fixed a bug in the Octree caching code where, in one circumstance, the
`write' routine would emit more data than the `read' routine was
expecting; with the result that the data stream would become
unsynchronized.
-*- Made csPolygonTree's ReadString(), ReadBox3(), ReadVector3(),
ReadPlane3(), ReadLong(), ReadUShort(), ReadByte(), and ReadBool()
methods more robust by checking for VFS read errors and returning a
somewhat sensible values rather than garbage values.
14 lines of code changed in:
Fixed a long-standing bug in archive.cpp where
csArchive::ArchiveEntry::FreeBuffer() was clobbering the compressed-size
value even though that value may have been used later.
3 lines of code changed in:
Fixed a bug in resizing X canvases, where SHM was used. If for some reason
SHM allocation would be denied during resizing, x2d will now switch to non
shared memory mode.
16 lines of code changed in:
Small folk unite!
7 lines of code changed in:
Repaired several compilation problems following Andrew's recent changes to
MazeD.
66 lines of code changed in:
-*- Started implementing "hints" at startup. Actually just wrote a number
of them in MazeD.zip::hints.cfg
-*- I decided to change a little the ideology of MazeD. The ex-"Draft View"
will be split now into two different kinds of windows: Model Editors and
World Editors. In Model Editor you will edit models (rooms, things, sprites)
and in the World Editor you will place them (without being able to change
the models itselves). mzDraftView component is the common ancestor now for
both mzModelEditor and mzWorldEditor components, and can draw the coordinate
axis and grid, nothing mode. The split between DraftView and ModelEditor
is finished.
-*- Created the mzWorldEditor window. It doesn't do much yet, but already has
basic functionality. As the world is loaded, the rooms are "centered" around
origin, and a respective mzModelInstance with the respective translation
transform is set up. This allow to edit every room as if it is located
in the origin, which is easier.
-*- Fixed a bug recently introduced by somebody. For software renderer
iTextureHandle::Prepare() did no remap_texture() which results in NULL
palette->global palette conversion table. This means you cannot use texture
after you do Prepare, but can if you do PrepareAllTextures(). Or
alternatively you can call Prepare() then SetGamma() :-) In short, a mess.
-*- Fixed a lot of bugs in fullscreen switching logic for all three X11 canvas
plug-ins (x2d, linex2d, glx2d):
- Moved all the code relevant to fullscreen switching to two files,
x2dfs.h and x2dfs.inc for easier maintainance. All three X11 plugins
include them from the relevant .h and .cpp files.
- Fixed a bug: when a fullscreen mode not matching current canvas size
is not available, the canvas may be cleared, which is not desirable.
- Bug: when current XFree86 videomode is the "best match" videomode for
the current canvas size, the canvas driver would hang (loop forever
in XWaitEvent).
- Bug: If videomode does not match canvas size, and canvas resizing is
allowed, the canvas size will be adjusted to match videomode. It
was supposed to work before but it didn't. If canvas resizing is not
allowed, the canvas is not resized, but centered instead.
- Bug: If an unexpected ConfigureNotify would arrive to our window,
canvas may be suddenly resized to an erroneous size. I've added a
check for ConfigureNotify to be related to our window.
-*- Re-implemented "model instances" in MazeD to allow arbitrary hierarchical
structures. For now the top level of the hierarchy can contain only rooms,
although this is not a MazeD limit, it is a engine limit. Each instance
has its move vector and space warp matrix, which are relative to parent
object. Moving/rotating parent object will move/rotate child objects as
well. Model instance position and warp are stored either as a MOVE(),
or a HARDMOVE() statement (for rooms and for instances that have the
"LockedPos" flag set). Alas, for now engine doesn't allow rooms at any
level other than top level, maybe it makes sense to remove this limitation.
For example, a castle (which would be a room) would contain a hall
(child room of castle) which contains something even smaller
(child of hall). Also for now we don't have sector templates, which
is bad again (if you duplicate a room multiple times, you'll get it
multiple times in output, although it could be avoided).
199 lines of code changed in:
fix my local mod committing
0 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