SofaROM bug report thread

Page 13/20
6 | 7 | 8 | 9 | 10 | 11 | 12 | | 14 | 15 | 16 | 17 | 18

Par sd_snatcher

Prophet (3517)

Portrait de sd_snatcher

10-09-2017, 02:40

As an effort to help to try to isolate the problem, these are the checksums of the ROMs I tested. According to the openMSX softwaredb.xml they're the originals:

SHA1(GALIOUS.ROM)= 4d51d3c5036311392b173a576bc7d91dc9fed6cb
SHA1(USAS.ROM)= 632cb0834441626d20cc113cfad2e3a7ec0e0dc7

Could those who are having problems check if these are the same checksums of the ROMs they're using?

Par gdx

Enlighted (5595)

Portrait de gdx

10-09-2017, 10:16

sd_snatcher, we have already verified the checksums of ROM and we have already done several tests that show that it works on some Turbo R and on others it does not work. The problem that remains to be found is the cause.

I tested with following Roms.

MoG SHA1:
4d51d3c5036311392b173a576bc7d91dc9fed6cb

USAS SHA1:
632cb0834441626d20cc113cfad2e3a7ec0e0dc7
4ff2aad8371e382e203c7f29b665612a8c9d937c

I exectuted TestMap v4.2 under DOS1 and it says “Mapper oke” after checking the 64 pages.

Under DOS2 it found following error.

Bit error on address 8000, block 3C
Output : FF / Input : 00
TestMap can’t continue, push space-bar.....

TestMap not seem to be supported by DOS2.

Par Louthrax

Prophet (2419)

Portrait de Louthrax

10-09-2017, 11:32

gdx wrote:

I exectuted TestMap v4.2 under DOS1 and it says “Mapper oke” after checking the 64 pages.

Under DOS2 it found following error.

Bit error on address 8000, block 3C
Output : FF / Input : 00
TestMap can’t continue, push space-bar.....

TestMap not seem to be supported by DOS2.

It does not work in R800 DRAM mode (the last pages are "ROM" in this mode, that's why you have this error.

It works if you type "CPU 0" or "CPU 1" before launching it (but MSX-DOS 1 should be in Z80 mode anyway).

Par gdx

Enlighted (5595)

Portrait de gdx

10-09-2017, 12:25

OK, "This Mapper is oke" also under DOS2 in R800 ROM mode. My RAM looks OK.
To be 100% sure, we would need an equivalent software that checks the RAM using the Panasonic Mapper.

PS: The default values for the Panasonic Mapper may differ from one Turbo R to another. If it was the case can this affect SROM?

Par Grauw

Ascended (10605)

Portrait de Grauw

10-09-2017, 13:20

My memory is original.

Par sd_snatcher

Prophet (3517)

Portrait de sd_snatcher

01-10-2017, 02:17

Here's a bug report: The newly released game Sword of Iana shows corrupted graphics for the main character when either the MemoryMapper or the Turbo-R slot3-3 MegaROM is used.

BTW, one feature suggestion: given the problems with the Turbo-R slot3-3 MegaROM, maybe the default device for ASCII16 ROMs should be the MemoryMapper? It not only will have less problems, but will also be faster since the page size is the same.

Par AxelStone

Prophet (3121)

Portrait de AxelStone

01-10-2017, 12:25

sd_snatcher wrote:

Here's a bug report: The newly released game Sword of Iana shows corrupted graphics for the main character when either the MemoryMapper or the Turbo-R slot3-3 MegaROM is used.

BTW, one feature suggestion: given the problems with the Turbo-R slot3-3 MegaROM, maybe the default device for ASCII16 ROMs should be the MemoryMapper? It not only will have less problems, but will also be faster since the page size is the same.

Confirmed, same problem here, corrupted graphics.

Par Louthrax

Prophet (2419)

Portrait de Louthrax

02-10-2017, 13:59

sd_snatcher wrote:

Here's a bug report: The newly released game Sword of Iana shows corrupted graphics for the main character when either the MemoryMapper or the Turbo-R slot3-3 MegaROM is used.

Problem identified, that's caused by the way the game transfers the character's GFX from ROM to video RAM. An RLD instruction is used (to get data in 4 bits chunks), that works well on a real ROM device, but corrupts the game when used on a RAM area. There's no quick fix for that, just buy the game (or use a real ROM flash device).

sd_snatcher wrote:

BTW, one feature suggestion: given the problems with the Turbo-R slot3-3 MegaROM, maybe the default device for ASCII16 ROMs should be the MemoryMapper? It not only will have less problems, but will also be faster since the page size is the same.

Good idea, I'll change the devices priority depending on the ROM mapper type.

Par AxelStone

Prophet (3121)

Portrait de AxelStone

02-10-2017, 14:33

Louthrax wrote:

There's no quick fix for that, just buy the game (or use a real ROM flash device).

Anyway is there a fix, althought not quick? Thanks Smile

Par Louthrax

Prophet (2419)

Portrait de Louthrax

02-10-2017, 15:48

AxelStone wrote:
Louthrax wrote:

There's no quick fix for that, just buy the game (or use a real ROM flash device).

Anyway is there a fix, althought not quick? Thanks Smile

Sure, but the whole transfer routine needs to be re-written, it will certainly be slower, and I'd prefer the game author's to do it, if they want to (releasing that might have an impact on the cartridge sales, so I prefer not to do it).

Page 13/20
6 | 7 | 8 | 9 | 10 | 11 | 12 | | 14 | 15 | 16 | 17 | 18