- Missing deallocation of queued frames in SceneManager
- Leaking functor on each call to entity setup functions
- Add SoundEntry destructor & better Entity cleanup
as reported by digitall
svn-id: r54196
Based on the FFmpeg decoder. Only the 16bpp version has been implemented (and all that should be needed). The videos I have tried work fine with the codec. The audio does not yet play in these videos, but I hope to work on DK3 IMA ADPCM soon.
svn-id: r54194
- Reading past the end of file in SaveLoad::init
- Missing deallocation of cache in SoundManager
- Missing deallocation of coords in SceneHotspot
- Missing deallocation of event handlers in LastExpressEngine::setEventHandlers
- Missing deallocation of InSaveFile in SaveLoad::isSavegameValid
- Missing deallocation of frames in Menu
as reported by digitall
svn-id: r54193
The PSP HW wasn't able to calculate the proper stretching when given the whole image size on a huge image. This is also a better way to do it because we're not overwriting tiles of the texture.
svn-id: r54190
- Placed the SCI3 version detection in the proper place
- Some new types of SCI3 MT-32 patches (e.g. in the Lighthouse SCI3 demo) are
ignored, for now
svn-id: r54188
These corrections close a number of leaks in the Toon engine reported by running Valgrind with --leak-check=full option, but a few still remain.
svn-id: r54185
Since SWORD25 uses Debug Channels, this allows for the interactive setting of debugflags as well as providing a base for adding further debugging commands.
However, this is not currently usable as the command key code is commented out.
This is due to the event loop which reads keyboard input being buried 2 object layers below the Engine VM object and I am unsure how the engine development team would want this exposed / interfaced.
svn-id: r54184
Each map has to be mapped with its own specific volume (as opposed to earlier versions that had one map for all volumes); this code was passing the same map pointer for all volumes.
svn-id: r54179
- Extended the SCI2.1 kernel function signatures for SCI3, as they share
the same kernel table
- All the engine parts are now initialized in SCI3 games, apart from the VM
svn-id: r54178
These corrections close a number of leaks in the Toon engine reported by running Valgrind with --leak-check=full option, but a significant number still remain.
svn-id: r54176
Shivers 2 doesn't contain SCI scripts. The whole game logic has
been reimplemented from SCI in native code placed in DLL files.
Each room has its own DLL file, and some SCI functions have been
reimplemented/rewritten for this purpose in native code. The
game and demo have all the resources of a SCI game, apart from
the SCI scripts themselves. Thus, they cannot be directly
supported, unless their whole room logic is rewritten from
scratch, which classifies Shivers 2 as "not SCI"
svn-id: r54173
- Enabled the SCI3 game entries for testing purposes
- The resource manager is initialized fully now (with a slight hack)
- Added a hack for the demo of Shivers 2 (which seemingly has no
scripts or vocabularies)
- The engine will stop before parsing any game scripts in SCI3 games,
and opens the console for resource manager-related functionality
svn-id: r54167
These corrections close a number of leaks reported by running Valgrind with --leak-check=full option, but some still remain, mainly in the image loading and caching.
svn-id: r54152
my implementation of romfs_seek is based on lseek, not on fseek,
so it returns the offset on successful completion, not 0, corrected
the check in RomfsStream::seek().
svn-id: r54151
