Identify a game

Page 3/3
1 | 2 |

By Louthrax

Prophet (2093)

Louthrax's picture

23-11-2019, 09:05

Has anyone checked if the same problem is observed on OCMs ?

By Manuel

Ascended (15815)

Manuel's picture

23-11-2019, 09:13

Alexey wrote:

I fixed the games by modifying all unused sprites in VRAM:

http://www.podrezov.com/temp/labirint_fixed.zip

But the games weren't broken! Can you help to fix the emulation by helping to write test programs to find out what is going on in detail, as described in the openMSX issue tracker?

By Alexey

Guardian (2444)

Alexey's picture

23-11-2019, 17:57

Why do we need test programs while we have 2 games that work incorrectly in the emulator? The reason for the false sprite collision detection was properly identified by Maarten.

All sprites in these games beside 0 and 1 have transparent color (because they are not defined after initialization). When the unused sprites' color is changed, the collision problem gets resolved without any side effects. What I did in the fix (it's a hack rather than a fix for those who like to play the games now) was just writing 0xFF to the 4th byte of every sprite's entry in the Sprite Attribute Table, except for 0 and 1 sprites. So to fix the problem, the emulator must start ignoring the collision of sprites with transparent color in their SAT entry.

Btw, I fixed the dangerous PSG output in "labirin2.com", similar to what we had in other games (Train, Blagger, etc.). The files have been updated on my share.

By Manuel

Ascended (15815)

Manuel's picture

03-12-2019, 21:55

We need test programs to find out what the exact behaviour of the VDP is regarding sprite collisions.

Wouter wrote a test program in BASIC and ran it on his turboR. The same test program behaved differently (and also the Labirint game) on a Philips VG 8020/40. It would be nice if some people could run it on their own machines, with TMS or even Toshiba VDP, to see whether the behaviour is the same as on V9958 or as on the 8020/40's VDP (TMS9129NL?).

The test program is still to be extended, but it might already be interesting to know what it does on your specific MSX, especially if you have an MSX with a different VDP than already tested. See https://github.com/openMSX/openMSX/issues/1198#issuecomment-... for the first version of the test program.

By Manuel

Ascended (15815)

Manuel's picture

04-12-2019, 23:10

Anyone who has an MSX setup with a known VDP who can run that test program?

By Alexey

Guardian (2444)

Alexey's picture

06-12-2019, 16:19

I will run them on my Yamahas with v9958.

By Manuel

Ascended (15815)

Manuel's picture

06-12-2019, 18:43

That's good, but we already had a test run on a turboR, so V9958 is covered. Tests on other VDP's are therefore more interesting.

Who can help?

By Alexey

Guardian (2444)

Alexey's picture

08-12-2019, 00:54

My colleague tested on 9938 with YIS-503III. Updated the thread.

By Manuel

Ascended (15815)

Manuel's picture

07-12-2019, 23:15

Seen it, thanks! Smile

By Manuel

Ascended (15815)

Manuel's picture

08-12-2019, 21:52

So, who can test it on another TMS type or even the Toshiba VDP? Would be very interesting to know the result for more accurate emulation!

Page 3/3
1 | 2 |