[Wine]Re: Myst Masterpiece Edition installer

[Duane Clark]

Jani Kärkkäinen wrote:
> Jani Kärkkäinen wrote:
> 
> 
>>crashes. Ran it with console, and at first there's lots of clicking 
>>and fixme:dsound messages and it throws the debugger just after this 
>>line:
>>
>>err:int21:INT21_Ioctl_Block int21: unknown/not implemented parameters:
>>int21: AX 440d, BX 0004, CX 0848, DX b61e, SI 0000, DI 0000, DS 0000, 
>>ES 0000
>>
>>Any relation?  http://www.ctyme.com/intr/rb-2896.htm
>>If so, congrats to Red Orb for programming by guidelines. :)
> 
> 
> Hmm. It seems int 21 function 440d with cx 0x0848 is simply a media lock 
> function. It shouldn't really crash wine, should it? Maybe I'm barking 
> up the wrong tree here.

Yep, that error is ok. It is not crashing Myst.

> With a bit more inspection, it seems that 
> QuickTime is crashing.
> 
> ...
> Backtrace:
> =>1 0x6687a887 in quicktime.qts (+0x7a887) (0x406cb430)
>   2 0x66911ee9 in quicktime.qts (+0x111ee9) (0x406cb478)
>   3 0x6686be5a in quicktime.qts (+0x6be5a) (0x0001002e)
>   4 0x00000000 (0x00000000)
> 0x6687a887: call        *0x60(%edx)
> 
> And this crash happens soon after the wine-display is blank and the cd 
> has spun up.

I am running the "win95" version of Myst, but the crashing behavior 
appears to be the same. The location of the backtrace varies quite a bit 
for me, depending on the compiled Wine version. Repeated runs with the 
same compiled version of Wine generally shows the same backtrace. I 
think that the backtraces are junk, but I could be wrong here.

For me, the crashing is always the same place in the game, though. The 
intro (a long Quictime sequence) works fine... or at least it did until 
this patch:
http://www.winehq.org/hypermail/wine-cvs/2004/10/0222.html
Backing out that patch causes the intro to work again for me. I am not 
convinced the patch is wrong, however. Once that change is made, Myst 
always crashes for me upon entering the game. It doesn't matter whether 
I enter the game early, by clicking during the intro, or I allow the 
intro to finish.

I should put the "always" above in quotes, because I have an old machine 
running RH7.3. I can compile Wine there, with the above mentioned patch 
backed out, and Myst works just fine. It is only on my RH9 machines (two 
of them) that Myst does not work. There appears to be something about 
the newer kernels that breaks it. For what it is worth, I tried versions 
of Wine all the way back to Aug 03, and Myst crashes the same way on all 
of them under RH9, but works under RH7.3.