Aley Keprt wrote: > If somebody wants to test the today's version of ASCD, > I can send it by e-mail.
Sure - why not! :-) (we're really gonna have to do something about the version numbering between of SimCoupé and ASCD - it's _so_ confusing!) > 2. The assembler-"optimised" versions are slower than C versions > optimised by C compiler. Yeah, I never gained any speed from the small sections either - it seems to hurt the compiler's register allocation too much, taking away any benefit from the asm block. I imagine there would be nice gains if larger logical blocks were re-coded (particularly with the display rendering), but for the sake of portability I didn't bother. > I have never understood why Allan did this, does anybody > else understand? I guess the compiler Allan used wasn't affected in the same way, or perhaps the optimisations weren't as good then? There must have been some benefit to them or he wouldn't have kept them! I've a vague memory one of the readme files or release notes mentioned a speed gain from them... > btw. I have big problems with Win32 version. The version > dated 9th March 2001 doesn't work. It starts correctly, but > then my computer halts or resets Win98 a few seconds after > start. Version dated 26th Februrary 2001 works fine. It's terrible! If you were having trouble, why didn't you just e-mail me about it? I've not had any reported problems with that version, despite 600 odd downloads of it in the last couple of weeks. It smells a bit like a driver issue - try changing the Surface= entry in the SimCoupe.cfg to read "Surface=0" (that's a zero), and the Sound= to read "Sound=No". If you still have problems after that, e-mail me off the list and I'll think of more things to try. Of course, since you have the source code it might be easier to locate the point Windows falls over using that (something I couldn't do with the crash I had under W2K in _your_ version!). Si
