Age | Commit message (Collapse) | Author | |
---|---|---|---|
2010-06-24 | Cleanup uninitialized read workaround(s). | Matthew Hoops | |
svn-id: r50225 | |||
2010-06-24 | SCI: added uninitialized read for room 720 in lsl1 | Martin Kiewitz | |
svn-id: r50222 | |||
2010-06-24 | Added more information concerning the Dr. Brain uninitialized read | Filippos Karapetis | |
svn-id: r50219 | |||
2010-06-24 | SCI: removing warning for known uninitialized reads | Martin Kiewitz | |
svn-id: r50217 | |||
2010-06-24 | SCI: change fix for island to make it work | Martin Kiewitz | |
svn-id: r50216 | |||
2010-06-24 | SCI: fixed workaround table from r50214 to have a valid terminator, also ↵ | Martin Kiewitz | |
added another fix for pharkas svn-id: r50215 | |||
2010-06-24 | SCI: created a workaround table for uninitialized reads, removing old ↵ | Martin Kiewitz | |
commented out workarounds for pharkas and laura bow 2 svn-id: r50214 | |||
2010-06-24 | SCI: added currently known object/methodnames in comments for uninitialized ↵ | Martin Kiewitz | |
reads svn-id: r50213 | |||
2010-06-24 | SCI: added objectname and method name printed out when uninitialized temps ↵ | Martin Kiewitz | |
are read svn-id: r50212 | |||
2010-06-24 | SCI: implement detection for uninitialized temp variable reads and implement ↵ | Martin Kiewitz | |
direct fixes for them, i couldnt figure out a valid replacement for island, pharkas and laurabow2 work that way. this should also get changed to a replacement table svn-id: r50211 | |||
2010-06-24 | Turned two hacks inside validate_property() into errors till we find out ↵ | Filippos Karapetis | |
where and why they occur (they shouldn't occur). Also turned an error inside send_selector() when argc > 1 into a warning - these are usually script bugs and aren't fatal svn-id: r50208 | |||
2010-06-24 | Fixed script 140 in the Island of Dr. Brain by adding a workaround for it ↵ | Filippos Karapetis | |
inside op_link svn-id: r50203 | |||
2010-06-22 | Fixed 2 crashes in QFG1VGA, when dying: | Filippos Karapetis | |
- Ignoring a non-game breaking script bug (modifying the "longSong" object with argc = 2) - Don't pop the callk stack frame if the stack is empty - fixes a crash when restoring from the death screen svn-id: r50157 | |||
2010-06-21 | When loading or restarting, stop script processing if the game is being ↵ | Filippos Karapetis | |
loaded or is restarting and the stack position has changed (it's always changed when loading) - fixes the case where the VM is trying to parse an invalid script svn-id: r50121 | |||
2010-06-21 | Add a workaround for a MUMG script bug, MUMG now works again. | Matthew Hoops | |
svn-id: r50117 | |||
2010-06-18 | Added hints inside validate_arithmetic(), together with some code useful for ↵ | Filippos Karapetis | |
debugging. Some cleanup svn-id: r50020 | |||
2010-06-17 | Strict mode: Changed several warnings due to logic bugs into errors. If an ↵ | Filippos Karapetis | |
error pops up from these ones, please add the game where it occurred and the steps to reproduce the error before turning it into a warning svn-id: r49973 | |||
2010-06-17 | SCI: Remove hack related to compiled kernel signatures. | Max Horn | |
Also change some things to comply to our code formatting conventions. svn-id: r49967 | |||
2010-06-17 | Don't show warnings for a known script bug in SQ4CD with the ↵ | Filippos Karapetis | |
Sq4GlobalNarrator object svn-id: r49950 | |||
2010-06-13 | Disabled two cases of findObjectByName() in SCI0 early games | Filippos Karapetis | |
svn-id: r49639 | |||
2010-06-10 | camelCase changes | Filippos Karapetis | |
svn-id: r49570 | |||
2010-06-10 | Resolved a FIXME with getSciLanguage(), by creating a separate setter. Also, ↵ | Filippos Karapetis | |
some camelCase changes svn-id: r49568 | |||
2010-06-10 | Use the SELECTOR() macro for readability | Filippos Karapetis | |
svn-id: r49563 | |||
2010-06-10 | Moved all of the game init/run/exit logic inside the SciEngine class | Filippos Karapetis | |
svn-id: r49559 | |||
2010-06-09 | Merged script_init_engine() and game_init() and cleaned up SciEngine::run() ↵ | Filippos Karapetis | |
a bit svn-id: r49537 | |||
2010-06-09 | Globals from script 0 are now initialized in script_init_engine(), and are ↵ | Filippos Karapetis | |
accessed from the relevant variables pointer. Removed direct reference to script 0 from the engine state svn-id: r49536 | |||
2010-06-09 | Merged restAdjust and restAdjustCur, as we don't save the restAdjust ↵ | Filippos Karapetis | |
modifier inside saved games (rightfully so). Also, the segment manager is now reset inside the main loop, when the game is restarted, not in game_exit() svn-id: r49533 | |||
2010-06-08 | Cleanup | Filippos Karapetis | |
svn-id: r49521 | |||
2010-06-08 | Merged the restarting_flags, script_abort_flag, and restoring members of the ↵ | Filippos Karapetis | |
EngineState class into one variable, abortScriptProcessing. The flag kept to signify a game restart has been placed in a boolean, gameWasRestarted svn-id: r49518 | |||
2010-06-08 | Merged _game_run() inside game_run(). Some cleanup | Filippos Karapetis | |
svn-id: r49513 | |||
2010-06-07 | Fixed regression from commit #49465 - there are in fact two restAdjust ↵ | Filippos Karapetis | |
variables, the current and global one svn-id: r49486 | |||
2010-06-06 | Now that EngineState is not deleted when loading games, we can move some ↵ | Filippos Karapetis | |
more state-related variables to it, and remove several FIXME's about non-const global variables. Also, the entries in the data stack are now deleted when loading (fixes a memory leak - thanks to digitall for this). svn-id: r49465 | |||
2010-06-01 | Stop reconstructing the engine state when restoring, but reset it instead | Filippos Karapetis | |
svn-id: r49376 | |||
2010-05-31 | Fixed operator precedence warning | Filippos Karapetis | |
svn-id: r49353 | |||
2010-05-31 | The code that loads saved game from the launcher | Filippos Karapetis | |
can now load SCI2 saved games. Added an explanation why the injected call is necessary svn-id: r49349 | |||
2010-05-30 | Limit access to the _bufSize, _scriptSize and _heapSize members of the ↵ | Filippos Karapetis | |
Script class svn-id: r49327 | |||
2010-05-29 | Mass renaming of selector-related functions, and removed some defines which ↵ | Filippos Karapetis | |
were just cloaking functions with a different name - GET_SEL32 -> readSelector - GET_SEL32V -> readSelectorValue - PUT_SEL32 -> writeSelector - PUT_SEL32V -> writeSelectorValue Also, changed some selector-related function names and variables to CamelCase svn-id: r49317 | |||
2010-05-28 | Added some more info when severe script errors occur | Filippos Karapetis | |
svn-id: r49293 | |||
2010-05-26 | SCI: Making various object and script related things const | Max Horn | |
svn-id: r49246 | |||
2010-05-26 | Formatting | Filippos Karapetis | |
svn-id: r49233 | |||
2010-05-26 | Added more info to the warning thrown when argc > 1 while modifying a ↵ | Filippos Karapetis | |
variable selector in send_selector() svn-id: r49232 | |||
2010-05-26 | After running some tests, it appears that modifying argc inside ↵ | Filippos Karapetis | |
send_selector() was a bad idea. Reverting the relevant code, and putting the warning back, with some extra comments svn-id: r49231 | |||
2010-05-26 | Silence a warning inside send_selector(): argc > 0 means "write selector". ↵ | Filippos Karapetis | |
Added code do make sure that argc is always either 0 or 1 when referencing variable selectors (this isn't strictly necessary, but better safe than sorry...). Some cleanup svn-id: r49230 | |||
2010-05-26 | More work on controlling access to members of the Object class: | Filippos Karapetis | |
- Moved the code for initializing the object class, species and base object inside the Object class - Made propertyOffsetToId() a method of the Object class - Made relocateObject() a method of the Object class - The Object getVariable() method now returns a reference to the requested variable Only SegManager::reconstructScripts() is left needing direct access to the members of the Object class svn-id: r49228 | |||
2010-05-23 | Remove the unused SCI_GAME_WAS_RESTARTED_AT_LEAST_ONCE flag | Filippos Karapetis | |
svn-id: r49161 | |||
2010-05-23 | Made shrink_execution_stack() a member of EngineState | Filippos Karapetis | |
svn-id: r49159 | |||
2010-05-23 | Cleaned up the game ID code: | Filippos Karapetis | |
- The game ID is now obtained from ScummVM ID directly, not by converting Sierra's internal ID - Moved the code which reads the internal Sierra ID inside the resource manager - Moved the code which converts the internal Sierra ID to ScummVM's IDs together with the rest of the detection code svn-id: r49152 | |||
2010-05-19 | Removed the exportsAreWide variable from the segment manager and save games, ↵ | Filippos Karapetis | |
and moved validateExportFunc() in the Script class, thus resolving a TODO svn-id: r49093 | |||
2010-05-19 | Moved the breakpoint information inside the DebugState struct | Filippos Karapetis | |
svn-id: r49092 | |||
2010-05-18 | - Moved kernel_lookup_text inside the Kernel class | Filippos Karapetis | |
- Added a pointer to the segment manager from within the Kernel class, thus simplifying the calls to it svn-id: r49076 |