openMSX about to release: testing help wanted!

Page 6/7
1 | 2 | 3 | 4 | 5 | | 7

By hit9918

Prophet (2911)

hit9918's picture

09-11-2014, 17:39

and those little 30 bytes files I thought would cost "nothing", I guess each takes a 1k cluster and then a hundred k is used.

By diedel

Resident (48)

diedel's picture

06-12-2014, 20:27

In the game Karateka there are some corrupted graphics. I've tested the original CAS version in the Philips VG-8020 model.

http://s12.postimg.org/jzigxu5kt/Karateka_0001.png
http://s29.postimg.org/psrf9sfvr/Karateka_0002.png

Edit: In the MSX2 Philips NMS 8245 model it seems OK.

By Manuel

Ascended (18240)

Manuel's picture

06-12-2014, 21:51

Which openMSX version did you test with, diedel?

Also, have you tested it on a real MSX1?

By diedel

Resident (48)

diedel's picture

07-12-2014, 16:26

I've tested it with the openmsx 0.11.0 official release and the following Fixato builds until now.
I don't have tested it in a real MSX1, neither previous versions of openMSX. But I've tested it in blueMSX and it seems OK.

By Grauw

Ascended (10157)

Grauw's picture

07-12-2014, 16:28

blueMSX doesn’t emulate VRAM access timing limitations, so no surprise there.

By mars2000you

Enlighted (6014)

mars2000you's picture

07-12-2014, 19:00

So, the temporary solution on openMSX is to disable this emulation.

By Grauw

Ascended (10157)

Grauw's picture

07-12-2014, 19:35

Have we already established that this problem does not occur on a real MSX then?

By Manuel

Ascended (18240)

Manuel's picture

07-12-2014, 19:56

It might be a new case where the uncertain part of the timing stuff we implemented appears to have been set too tight again (like Chase H.Q.), but we don't know until it's tested on real hardware.

In the mean time I made a ticket for it: https://sourceforge.net/p/openmsx/bugs/558/

Please let us know if you find any other software showing graphics corruption on MSX1 (but not on MSX2) in openMSX.

By hit9918

Prophet (2911)

hit9918's picture

07-12-2014, 23:08

just checked my vdptest again.
things work same except 26 cycles write test fails more wild on openmsx. but still fails on real machine.
funny karate title screen, I wouldn't loosen the rules for that. warning contemporary coding I find more important.

guessing. maybe the titlescreen made less sparcles on real thing and then nobody noticed.
And I guess the issue wont pop up with game rendering code because there too fast code does too wild errors.

By hit9918

Prophet (2911)

hit9918's picture

07-12-2014, 23:33

or maybe the intro was added on MSX2.

Page 6/7
1 | 2 | 3 | 4 | 5 | | 7