SCI/TODO

• Known bugs while using newgui
  • King's Quest 5
    • Sometimes there is animation cels shown in talking boxes. This is caused by the script calling kDrawCel for drawing the talking boxes and it also calls kAnimate later, which then draws current active cels on top of it
    • Pathfinding in the screen that is before getting captured by the big bird isn't working correctly. At least when clicking on the place where one is supposed to go, graham goes backwards. It's possible to get him to the required position but 2-3 mouse clicks are needed
    • priority map issues, seems to be caused by floodfill (look at the anchor at the beach house)
    • text is not correctly centered when viewing the credits (see FIXME in SciGui::display)

• • Larry 5
    • Patti loop isn't correct, resulting in glitches and glitchy animation

• Known games that can be completed
  • King's Quest 5 (10/18/2009), King's Quest 6 (10/2009)

• Easy
  • rename the sfx/ dir to sound/ or audio/ (it doesn't just deal with "sound effects).
  • change the many nice existing function documentation comments to use doxygen syntax
  • Change #define lists into enums

• Medium
  • Convert stuff from "C pseudo classes" to real classes.
  • check for places that use char or unsigned char for boolean values, and change those to bool. Even nicer would be to change it to use enum types whose value names express the intention (SearchOptions.RECURSIVE, GraphicsScaleOptions.2X, etc), but just going for bool will already be a big improvement.
  • Replace various generic FreeSCI code by their ScummVM counterparts, where it make sense.
  • Check old FreeSCI-scummvm branch / repos by Jordi for change that could be salvaged, e.g. the files sfx/device/scumm-midi.cpp and sfx/softseq/scumm-adlib.cpp, and stuff.

• Difficult
  • Turn code into templates in following files (now they generate code via #defines):
    • gfx/picfill.cpp
  • make the SCI tools compilable (to avoid bitrot, while we change code that they depend on), and finally usable again.
  • rewrite the YACC parser code in said.y in pure, readable C++, so we don't need YACC/Bison.
  • use direct drawing instead of using widgets for rendering to be compatible with SSCI.
  • fix palette glitch on PIC transition (when changing rooms).

• Bugs
  • Quest for Glory
    • The FileSelector (SCI) class used in the import screens in QfG2 and QfG3 has a 'text' pointer, and 'cursor' and 'lsTop'/'topString' pointers that should point to inside 'text'. However, in FileSelector::readFiles, 'text' is freed and reallocated, causing the other pointers to sometimes become invalid.
  • Longbow
    • The sound of the bow in the opening screen is repeated a number of times
    • After leaving the cave and the conversation that follows, a list is not sane anymore (perhaps it's related to pathfinding? There's a relevant warning just before this occurs)
  • King's Quest 1
    • After typing a wrong word, the game attempts to open a file with an empty name

• Bugs observed from FreeSCI (check here)
  • Codename: Iceman displays the warning "NodeValue() on a NULL pointer attempted" at numerous points during the map table scene
  • Quest for Glory 1 displays the warning "List node at x is not sane anymore" when the dinosaur snaps at you during the intro
  • Space Quest 3 calls functions from a script which has already been unloaded when Roger climbs up the ladder from the rat cave (the place containing the reactor at the very beginning of the game) - NOTE: This should be fixed in ScummVM, to the best of my knowledge
  • PQ2 calls CanBeHere() with a disposed object in the clip list when Keith enters the car
  • LSL2 tries to allocate zero bytes of memory when entering Molto Lira.
  • QG1 accesses property 0xf02 in a couple of cases
  • SQ1 calls UnLoad() with only one argument when exiting the Ulence Flats bar
  • SQ4/Floppy will set the speed to maximum if you access the control panel without tweaking the speed slider. Either tweak the speed slider, or use function keys to access the control panel functionality