Page 1 of 1

Help for testing on real hardware

Posted: Mon Apr 04, 2016 9:55 pm
by tryphon
This is a little project far from finished, but when I wanted to test it on real hardware (with a MegaEverdrive), I wasn't able to control the chars (maybe the game crashed since nothing moved).

Though, it runs flawlessly with Gens.

Could someone try it on real hatdware so I can know if the problem is related to my coding or to my hardware ?

Here's the link : http://s000.tinyupload.com/index.php?fi ... 1459679385

Re: Help for testing on real hardware

Posted: Mon Apr 04, 2016 10:36 pm
by Moon-Watcher
Everything works for me.

Re: Help for testing on real hardware

Posted: Tue Apr 05, 2016 5:01 am
by tryphon
Thank you :)

It's weird, on sega16, the game was reported crashing on a MD2.

Can I ask what's your model ? Also did you use a 3-buttons pad or a 6-buttons one ?

Note : since it's still WIP, there are some KDebugs inside. Could it be the cause ? (though I already used succesfully KDebuggy code on real hw). Fusion says the checksum is incorrect, but I'd be surprised SGDK actually tests checksum ?

Re: Help for testing on real hardware

Posted: Tue Apr 05, 2016 11:56 am
by Moon-Watcher
It's a PAL model 1 modded as american NTSC with Everdirve and a 6 button wireless pad
https://twitter.com/MoonWatcherMD/statu ... 0184608768

Edit: Also works in PAL mode

Re: Help for testing on real hardware

Posted: Tue Apr 05, 2016 3:15 pm
by Stef
SGDK does not test checksum of course :p

Re: Help for testing on real hardware

Posted: Tue Apr 05, 2016 3:43 pm
by tryphon
Stef wrote:SGDK does not test checksum of course :p
Have you got an idea about what could cause a game to run on some MD and not on others ? I'm totally clueless.

Here, ComradeOj tested unsuccessfully the game against a variety of machines. As for me, the game starts but I can't move anything. I'll add a timer of something to see if the game has crasher or if the controller is just not read properly...

Re: Help for testing on real hardware

Posted: Tue Apr 05, 2016 8:21 pm
by KanedaFr
tryphon wrote: Note : since it's still WIP, there are some KDebugs inside. Could it be the cause ? (though I already used succesfully KDebuggy code on real hw).

KDebug uses a non existing register, like writing to an reserved memory address.
So perhaps some hardware doesn't like it.
I remember I was able to crash Fusion with a call to debug at the right(ie wrong) place

Re: Help for testing on real hardware

Posted: Tue Apr 05, 2016 10:14 pm
by tryphon
I never had problems with running KDebugged code on my MD. I admit I didn't run many codes :)

Here's a version without KDebugs functions : http://s000.tinyupload.com/index.php?fi ... 4616709608

Then again, I'm interested in reports :)

Re: Help for testing on real hardware

Posted: Wed Apr 06, 2016 4:29 pm
by tryphon
I found it :mrgreen: (at least I think)

Re: Help for testing on real hardware

Posted: Wed Apr 06, 2016 6:22 pm
by Stef
\o/