Age | Commit message (Collapse) | Author |
|
svn-id: r51924
|
|
Not just the 6 bits per color component used in VGA
svn-id: r51921
|
|
svn-id: r51920
|
|
To match the other VideoDecoder classes with sound support.
svn-id: r51919
|
|
Like it should have been in the first place...
Fixes Woodruff slot-VMD positions.
svn-id: r51917
|
|
svn-id: r51915
|
|
svn-id: r51912
|
|
svn-id: r51909
|
|
svn-id: r51908
|
|
svn-id: r51907
|
|
svn-id: r51906
|
|
svn-id: r51905
|
|
Change CoktelDecoder::renderBlock*() and IMDDecoder::renderFrame()
to receive a Common::Rect instead of mucking about with _dirtyRects.
svn-id: r51904
|
|
svn-id: r51903
|
|
svn-id: r51902
|
|
svn-id: r51901
|
|
svn-id: r51900
|
|
svn-id: r51899
|
|
svn-id: r51898
|
|
svn-id: r51897
|
|
svn-id: r51896
|
|
svn-id: r51892
|
|
svn-id: r51890
|
|
svn-id: r51888
|
|
svn-id: r51886
|
|
svn-id: r51884
|
|
svn-id: r51881
|
|
svn-id: r51877
|
|
Stubbing CoktelDecoder::hasEmbeddedFile() and
CoktelDecoder::getEmbeddedFile(), formerly hasExtraData/getExtraData.
svn-id: r51875
|
|
svn-id: r51874
|
|
Fix IMD playing for some fringe cases, especially when seeking
beforehand.
svn-id: r51871
|
|
In the gob engine, some videos are opened, and then played, at
a later time, one frame at a time. In this case, we need to
emulate the original's video player's behaviour, not calculating
any "lag" for videos without sound.
svn-id: r51870
|
|
svn-id: r51867
|
|
Rendering the frame video data is still stubbed out.
svn-id: r51866
|
|
svn-id: r51865
|
|
svn-id: r51862
|
|
Allows client code to overwrite the video's frame rate.
svn-id: r51860
|
|
This allows for the client code to only update the parts of the
video frames that actually changed.
svn-id: r51859
|
|
This allows for positioning the video within the video memory.
svn-id: r51858
|
|
This allows the video player to directly draw onto its own video
memory without having to blit each frame another time.
Will also be needed for proper handling of transparency in Woodruff.
svn-id: r51857
|
|
Implemented bare PreIMD decoding using the new CoktelDecoder interface.
No fancy stuff yet, only basic vieo frames.
svn-id: r51854
|
|
This creates a new CoktelDecoder class using the VideoDecoder
interface, which will eventually become the new way to decode
PreIMD, IMD and VMD videos.
Since the VideoPlayer in gob is not yet ready for this, we're
disabling all video playback in the gob engine for now.
svn-id: r51850
|
|
This way, Indeo3 is properly usable in various container VideoDecoder,
for example AVI, and not just VMD.
Since VMD is not yet ready for this, we're disabling Indeo3 in VMDs
for now.
svn-id: r51849
|
|
Now takes a pointer to a stream, instead of a reference.
The rational is that in all instances, callers have a pointer
(and dereference it to call load), and all load implementations
turn the reference back into a pointer.
svn-id: r51725
|
|
Add const qualifier to FlicDecoder::hasDirtyPalette to make it
match that if VideoDecoder::hasDirtyPalette.
svn-id: r51724
|
|
svn-id: r51692
|
|
svn-id: r51668
|
|
The bounding boxes of the glyphs use signed coordinates. We stored only
unsigned coordinates, which resulted in incorrect glyph positioning.
Conrecte example: the bounding box of the glyphs for clR6x12-iso-8859-5.bdf
used:
x y w h
0 -3 6 12
We on the other hand interpreted that as:
x y w h
0 65533 6 12
when loading the font from our font cache.
svn-id: r51586
|
|
Previous behavior was to assert
svn-id: r51517
|
|
Thanks to salty-horse for finding this. Also, use Common::Rational directly to hold the frame rate to avoid rounding.
svn-id: r51516
|