Age | Commit message (Collapse) | Author |
|
don't allow the engine to change the directory where saved games will be placed
svn-id: r44044
|
|
svn-id: r44043
|
|
svn-id: r44042
|
|
BREAKPOINT() was used to an error
svn-id: r44041
|
|
getObjectName, respectively)
- Renamed alloc_List -> allocateList, alloc_Node->allocateNode, alloc_hunk_entry->allocateHunkEntry, free_hunk_entry->freeHunkEntry, for consistency
svn-id: r44039
|
|
svn-id: r44017
|
|
svn-id: r44015
|
|
multiply one dimension with an integer multiplier which is different to the multiplier of the other dimension, otherwise we'll end up with funny looking and squashed resolutions like 640x200 or 320x400. Also, removed the now-unused pixelFormat member variable of the graphics driver struct
svn-id: r44003
|
|
svn-id: r44000
|
|
svn-id: r43999
|
|
svn-id: r43998
|
|
svn-id: r43996
|
|
svn-id: r43995
|
|
colors, so hi-color isn't really used anywhere, and it only makes the overall code more complex for no reason
svn-id: r43994
|
|
messages, introducing a new debug level
svn-id: r43993
|
|
bool, and rename setExportWidth() to setExportAreWide()
svn-id: r43988
|
|
Add new methods init, allocateObject and getObject to class Script, and use them instead of directly accessing Script::obj_indices. Replace RAW_GET_CLASS_INDEX with Script::getObject()
svn-id: r43987
|
|
Common::HashMap<int,int>
This simplifies the code considerably. Also changed the savegame format
accordingly, which required me to bump the format version to 10. Old
saves should still load fine.
svn-id: r43986
|
|
svn-id: r43985
|
|
svn-id: r43984
|
|
svn-id: r43982
|
|
- rename segGet and getSegment to getScriptSegment; the two can be
distinguished by the parameter count.
- rename type SCRIPT_GET to ScriptLoadType to conform with our code
formatting conventions
- rename get_class_address to getClassAddress
- some cleanup
svn-id: r43981
|
|
svn-id: r43980
|
|
in two cases on the stack, for simplicity
svn-id: r43979
|
|
svn-id: r43978
|
|
svn-id: r43943
|
|
- Changed the message shown when the game tries to move the cursor off the screen bounds from a warning to a debug message, to avoid spam in games that do this behavior, e.g. the Camelot demo
svn-id: r43942
|
|
corrupted, we should error out instead of piggy-banking an error flag to another function
- Do not try and set the position of the mouse cursor when it's hidden
svn-id: r43941
|
|
svn-id: r43940
|
|
are now always 0 for transparent, up to 255 for opaque
svn-id: r43939
|
|
unfiltered scaler in (to be replaced by ScummVM's scaler code). We really don't need bilinear and trilinear filtered scaling in the engine, as ScummVM's filters already perform bilinear and trilinear filtered scaling, if requested
svn-id: r43938
|
|
svn-id: r43937
|
|
changes, and it's used in one place only)
svn-id: r43936
|
|
to allocateClone()
svn-id: r43935
|
|
svn-id: r43916
|
|
the 3 dithering modes as an enum (and reorganized them a bit)
- Added 3 config options for dithering in the ini file - mainly for purists (config option "dither_mode"): 0 - disabled, 1 - 16 color dithering (Sierra style) and 2 - 256 color dithering
- Dithering is now always applied depending on the screen scale (removed the relevant parameter)
- Removed the background picture buffering option, used to speed-up room changes. Room changing is quite fast (instantaneous), and the extra memory allocated is not worth the possible speed increase in small devices. Plus.... there is no reliable value to set this option to, so there is no point in letting the user decide what value to put there using guesswork
svn-id: r43915
|
|
svn-id: r43913
|
|
left the algorithm of updating the screen with multiple small rectangles, like we do in all the other engines that support dirty rectangle screen updates
svn-id: r43909
|
|
getresourceManager -> getResourceManger
resourceManager -> resMan
segmentManager ->segMan
svn-id: r43908
|
|
svn-id: r43907
|
|
* removed kNOP
* renamed k_Unknown to kUnknown
* added FIXME to kfunct_mappers table
* more cleanup
svn-id: r43906
|
|
removed eventually)
svn-id: r43905
|
|
svn-id: r43903
|
|
make sense to dither, as Sierra games use up to 256 colors (even the later SVGA ones), and it messes up the result of the FreeSCI graphics filters. Plus, the resulting image looks worse than the original one
svn-id: r43901
|
|
(is dithering even used anywhere?)
svn-id: r43897
|
|
filtering algorithm options from the ScummVM ini file (options "cursor_filter", "view_filter", "pic_filter" and "text_filter", 0 - unfiltered, 1 - bilinear filtering, 2 - trilinear filtering)
svn-id: r43894
|
|
for VGA remakes
- The extras field is now filled by the fallback detector for VGA games
svn-id: r43893
|
|
- Added the MD5's of the French versions of GK2 and SQ6 (bug reports #2846842 and #2846515)
- Added a different game id for QFG1 VGA (SCI1.1), to distinguish it from the older EGA version (SCI0) and added detection for the VGA version in the fallback detector
- Changed some VGA SCI game IDs: lsl1sci -> lsl1vga, pq1sci -> pq1vga, sq1sci -> sq1vga, to distinguish SCI VGA remakes from SCI EGA remakes (like with QFG1VGA)
svn-id: r43891
|
|
room with the pier in KQ6.
svn-id: r43888
|
|
game-specific workarounds are tested against ScummVM IDs
- Added detection for the EcoQuest 1 demo in the fallback detector
- Partially reverted my previous "fix" for EcoQuest 1 CD, and turned it into a script-specific hack for that game, for now
- Added handling of kDisposeScript calls made with 2 parameters, e.g. in KQ5CD and others (still untested, haven't found a test case)
svn-id: r43887
|