forums.marpirc.net

so it is .. i guess it doesn't work then .. in which case i really don't know what to change sorry .....

here's some code for thought, i haven't checked the official mame source but this (in XMAME source) is a nono when it comes to srand and expecting the same result, these should either be constants passed to srand or ideally: when recording, time(0) is stored in the inp. when playsback: it can seed srand with the inp loaded time.


srand(time(NULL));
./src/unix/config.c
srand( (unsigned)time( NULL ) );
srand((unsigned)time(NULL));
./src/windowsui/Win32ui.c
srand(time(0));
./src/windows/config.c


Since we don't want to change the inp format just yet: barry can you try replacing the time(NULL)'s that are passed to srands with constant 0's. and give me a copy to see if this affects anything? that is if this is in the official source too and not just xmame. I'll try doing it with xmame and see what happens.

I tend to think that the problem with Puzz Loop is more likely to do the emulation of the inputs, as it says in the source that they were problematic in getting to work in the first place. That is only in regards to Puzz Loop and no other Super Nova game, which as far as I know, all playback fine.

you're proly right. I tried changing the srands in xmame, got a 214k panic mode and it still only played back to 20k. anyways if rand is being used sranding to time() for playback and recording isn't goign to help, but it seems to be ok for the rest of games...

mahlemiut wrote:

Haze wrote:I think somebody (Kale?) came up with an RTC patch for it .. if you try that and find it works fine I can put it in the code for the driver, if not I don't know how to fix it....

I believe it's already in 0.63.

Yes,it's already in 0.63/0.64,and it should work for Puzz Loop as well,if it's really a RTC issue and not a rand()-specific thing...

And yes it's me the guy that submitted the RTC patch.My real name is Angelo Salese...