Hi Bas,
[KQ4]
> - Warning "Unknown SDL keysym: 0136 (0)" when pressing Meta key (which
> I use for my windows manager short cuts). This key should probably
> just be ignored.
Agreed.
> - Warning whenever meeting the unicorn:
> FSCI: kNOP: Warning (kernel.c L622): Warning: Kernel function 0x55
> invoked: NOP
As has been pointed out, this means that a kernel function (DoAvoider(), a
no-op in all but the earliest SCI interpreters, in this case) that hasn't
been implemented yet is invoked. It's on the TODO list, and it's also
responsible for the unicorn's moonwalking.
> - the screen 4 east from the starting point is shifted by about 20 pixels
> to the left (savegame #1 `displaced screen').
That'd be the dwarvish mines, right?
(I can't use your savegames with my version of KQ4, since you have one of
the earlier ones).
> - When startling cupid at the pool area (save game #2 `cupid pool'):
> FSCI: kEmptyList: Warning (klists.c L213): Warning: EmptyList() was
> invoked with 1 parameters
We still need to verify that this function works correctly; there's
nothing else to this warning.
> - When looking under the bridge and getting the golden ball (save game #3
> `golden ball'):
> FSCI: ERROR in kfree (kernel.c L365): Error: Attempt to kfree() non-handle
> FSCI: ERROR in invoke_selector (kscripts.c L82): Selector 'delete'
> of object at b754 could not be invoked (kgraphics.c L1761)
Makes me wonder- didn't we have that problem before?
Does explicitly setting FreeSCI to emulate a very early (e.g. 0.000.395)
version of SCI help here?
> FSCI: _k_view_list_dispose_loop: Warning (kgraphics.c L1763):
> Object at b754 requested deletion, but does not have a delete
> funcselector
> After trying to reproduce this some times, i got a failed assertion:
> sciv: heap.c:26: set_next: Assertion `next>=1000 && next<=0xffff' failed. Abort
Seems to be a heap corruption or an out-of-heap-space problem. Did you
just restore and try again to reproduce the problem?
llap,
Christoph
