Hi Bas,

>  - shifted screen: still there.

We had a shifting problem caused by a gcc optimizer bug once, but it was
consistant on all screens. Hmm... does this screen look right on DOS?

>  - golden ball: I now get a segfault at themoment I enter this screeen
>    (2 east, 1 north from start):
>      Program received signal SIGSEGV, Segmentation fault.
>      [Switching to Thread 1024 (LWP 23884)]
>      0x403169cf in memcpy () from /lib/libc.so.6
>      (gdb) backtrace
>      #0  0x403169cf in memcpy () from /lib/libc.so.6
>      #1  0x809d55f in gfxr_interpreter_calculate_pic (state=0x83defd8, 
>scaled_pic=0x8622758, unscaled_pic=0x8641b80, flags=1, default_palette=0, nr=21) at 
>sci_resmgr.c:114
>      #2  0x8094ee5 in gfxr_get_pic (state=0x83defd8, nr=21, maps=1, flags=1, 
>default_palette=0, scaled=1) at resmgr.c:306 #3  0x8093a2b in gfxop_new_pic 
>(state=0x80dc6c0, nr=21, flags=1, default_palette=0) at operations.c:1738
>      #4  0x80710ca in kDrawPic (s=0x840c078, funct_nr=8, argc=2, argp=29760) at 
>kgraphics.c:1139
>      #5  0x8052acf in run_vm (s=0x840c078, restoring=0) at vm.c:770
>      #6  0x8055d79 in _game_run (s=0x840c078, restoring=0) at vm.c:1554
>      #7  0x80560ab in game_run (_s=0x80dc79c) at vm.c:1620
>      #8  0x804eb4d in main (argc=2, argv=0xbffffc74) at main.c:1066
>      #9  0x402bc38b in __libc_start_main () from /lib/libc.so.6
>     
>  - the cleaning up in the dwarf house does work now, although Rosella
>    does walk a bit weirdly (even when she moves to the left or right,
>    her pixmap still looks like when she's walking towards the viewer).
>    In Sierra SCIV, her movements look normal.

Weird. Maybe it's some foreign use of DirLoop() or something...
It's .231 now.

>  - the crash when entering the left room in the spooky house is still
>    there, although the error message is somewhat different:
>      Warning: Unaligned read from a16f
>      Send: No object at a177!
[...]

Humh... weird.

What does a stack trace of that look like? (the FreeSCI debugger's stack
trace command is 'bt').
This could be any kind of script problem or memory corruption. The early
SCI versions still aren't well-supported... :-/

Thanks for your bug reports,

llap,
 Christoph


Reply via email to