Page 2 of 2

Posted: Tue May 12, 2009 11:07 am
by tepples
tokumaru wrote:
tepples wrote:[Mapper 0 with CHR RAM is] not invalid if you treat mapper 0 with CHR RAM as an undersize BNROM, just like ordinary NROM is an undersize CNROM, and CNROM is an undersize GNROM.
I always thought that it'd make more sense to treat mapper 0 with CHR-RAM as an undersize UNROM.
UNROM has 16 KiB bankswitching. $C000 is the last bank, as one would expect, but $8000 is a random bank. I recommended BNROM because of BNROM's 32 KiB bankswitching.
I believe Nestopia said my ROM was corrupted when I tried that though, it wouldn't take less than 128KB or PRG-ROM.
Did you make sure to switch in bank 0 before reading $8000-$BFFF? It's perfectly possible to have UNROM with four banks (64 KiB PRG ROM); it's just that no commercial game was released in that configuration.

Anyway, I can assume that we agree that either a new mapper or perhaps a new iNES 2.0 flag would be needed to provide the correct pattern table mirroring.

Posted: Tue May 12, 2009 11:56 am
by tokumaru
tepples wrote:UNROM has 16 KiB bankswitching. $C000 is the last bank, as one would expect, but $8000 is a random bank. I recommended BNROM because of BNROM's 32 KiB bankswitching.
I see. UNROM is still perfectly good if you have 16KB of PRG-ROM or less.
Did you make sure to switch in bank 0 before reading $8000-$BFFF?
It didn't matter what I did (in fact I think I only had 16KB of PRG), it refused to load the ROM. Maybe it was an old version of the emulator? I'm too lazy to test the current one, I don't have that ROM anymore.
It's perfectly possible to have UNROM with four banks (64 KiB PRG ROM); it's just that no commercial game was released in that configuration.
Yeah, I know it should be valid. But are you sure no games were released like this? I remember looking at the Zanac ROM in an hex editor and both halves looked identical. Then I cut the ROM in half and it seemed to work fine. Even in Nestopia (I just tested, it says the board is "non-standard").
Anyway, I can assume that we agree that either a new mapper or perhaps a new iNES 2.0 flag would be needed to provide the correct pattern table mirroring.
Yeah, I agree.

Re:

Posted: Sun Apr 28, 2019 5:06 pm
by tepples
Mapper 218 was allocated to this configuration. Discussion continues in Single Chip Cartridge.