Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 4, 2019 1:11:44 GMT
List of games with issues with the US System Card 3.0 / Turbo Duo "BIOS":
'A' Ressha de Iko III (no fix)
The developer copied a chunk of the japanese System Card 3.0 code for no discernible reason -- subroutines addresses are different in the US Syscard. Altered Beast (fix - commercial)
Unknown issue which makes it work only on the 1.0 System Card - Rare reprint fixes bug and is compatible with all system cards
? ? ?
Bonanza Bros. (fix - www.interlog.com/~daves/pce_dev/translt.html)US System Card reports itself as ver. 3.1, Japanese System Card reports as 3.0. Game checks for the exact version "3.0".
|
|
|
Post by dshadoff on Mar 4, 2019 1:51:59 GMT
Well, I'm the guy who found/fixed Bonanza Brothers and Alshark...
I wasn't aware that A.III had a problem with the US card, and until recently, I hadn't heard of any real evidence that Altered Beast was ever fixed (I now have an image, but I haven't had a chance to actually try it...). But I didn't hear that it had problems with the USA card either.
I would suspect that whatever problems that exist, are of the same variety as those in Bonaze Brothers and Alshark (although that was so long ago that I don't recall what was wrong/how I fixed them).
I'd be curious and interested to help out (in-between my many other projects).
Dave
|
|
fragmare
Punkic Cyborg
Posts: 116
Homebrew skills: Graphics, Music, Level Design, Annoying Programmers
|
Post by fragmare on Mar 4, 2019 2:47:48 GMT
A.III does seem to have a problem with the syscard3 fix, i can vouch for that. At least the version I have. I believe Godzilla had some sort of issue too, iirc
|
|
|
Post by elmer on Mar 4, 2019 3:07:33 GMT
Tenshi no Uta I & II both seem to check a hardware bit and try to enable some debugging stuff. I'm not sure if it causes a problem on an actual TurboGrafx, but it certainly screws up when you run a TurboGrafx Super System Card in Mednafen. Tenshi I just completely craps out, but Tenshi II does timeout from whatever it's trying to do and recovers and runs normally. A.III does seem to have a problem with the syscard3 fix, i can vouch for that. At least the version I have. I believe Godzilla had some sort of issue too, iirc The syscard3 fix? What do you mean by that? Do you mean my patch for the TED2?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 4, 2019 3:52:42 GMT
Well, I'm the guy who found/fixed Bonanza Brothers and Alshark... I wasn't aware that A.III had a problem with the US card, and until recently, I hadn't heard of any real evidence that Altered Beast was ever fixed (I now have an image, but I haven't had a chance to actually try it...). But I didn't hear that it had problems with the USA card either. I would suspect that whatever problems that exist, are of the same variety as those in Bonaze Brothers and Alshark (although that was so long ago that I don't recall what was wrong/how I fixed them). I'd be curious and interested to help out (in-between my many other projects). Dave Thanks Dave, and yes I checked out your Bonanza Bros. patch and that's what inspired me to check out other games. To be quite honest I was mostly worried about BIOS compatibility and hidden pitfalls for current homebrew projects, since I only have a US Duo it's hard for me to test across different devices outside of emulation.
I'm not sure if AIII will have a patch as "easy" as yours (NOP'ing the sub-version number compare) but I hope it's nothing too complex (aka just NOP NOP NOP overwrite in some weird routine jump that uses the region bit as a check for unexistant hardware as Elmer hinted at). The game does not seem to check the Y register after calling EX_VERCHK or whatever that's called, so it's not just a copy paste of your other work.
|
|
fragmare
Punkic Cyborg
Posts: 116
Homebrew skills: Graphics, Music, Level Design, Annoying Programmers
|
Post by fragmare on Mar 4, 2019 4:06:10 GMT
Tenshi no Uta I & II both seem to check a hardware bit and try to enable some debugging stuff. I'm not sure if it causes a problem on an actual TurboGrafx, but it certainly screws up when you run a TurboGrafx Super System Card in Mednafen. Tenshi I just completely craps out, but Tenshi II does timeout from whatever it's trying to do and recovers and runs normally. A.III does seem to have a problem with the syscard3 fix, i can vouch for that. At least the version I have. I believe Godzilla had some sort of issue too, iirc The syscard3 fix? What do you mean by that? Do you mean my patch for the TED2? I believe Punch is referring to the "US" region System Card 3.0 ROM that's been modified. Iirc, there's no verified dump of the US System Card 3.0, so a long time ago somebody (dshadoff maybe?) modified the existing dump to work... but it's not perfect. A few games have always had problems, so you must switch back to the Jp region card for them to work. I used to own the actual TTI released US version of the Super System Card 3.0... wish I still had it. I'd loan it out to have the BIOS re-dumped. But it's gone...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 4, 2019 4:09:06 GMT
Ok now you lost me.
|
|
fragmare
Punkic Cyborg
Posts: 116
Homebrew skills: Graphics, Music, Level Design, Annoying Programmers
|
Post by fragmare on Mar 4, 2019 4:19:50 GMT
The ROM dump of the Jp region System Card 3.0 BIOS is verified, which means somebody bothered to check it against the real hardware and it should be a perfect dump of the Jp region BIOS. The story I've always heard about the "US" version of the BIOS dump going around, is that the ROM dump was imperfect in some way, and was later modified (post dumping) to work... which explains why some games still have problems with it. Unless something has changed, and there's now a verified version of the US syscard 3 BIOS ROM that's been dumped and is now floating around... if there is, please disregard my ramblings. lol
|
|
|
Post by elmer on Mar 4, 2019 4:27:16 GMT
To be quite honest I was mostly worried about BIOS compatibility and hidden pitfalls for current homebrew projects, since I only have a US Duo it's hard for me to test across different devices outside of emulation. You really shouldn't have any problems unless you rely on undocumented behavior, or jump directly into the System Card ROM. If you're worried, you might want to get a Turbo Everdrive 2 for testing, and then you can run all of the different System Card variations. I believe Punch is referring to the "US" region System Card 3.0 ROM that's been modified. Iirc, there's no verified dump of the US System Card 3.0, so a long time ago somebody (dshadoff maybe?) modified the existing dump to work... but it's not perfect. A few games have always had problems, so you must switch back to the Jp region card for them to work. Huh? I have no idea what you're talking about. AFAIK, the US Super System Card was dumped years ago ... the ROM files are out there, and consistent. The reason that TurboGrafx Super System Card image doesn't run some Japanese games isn't anything to do with a bad or nonexistant dump, it's because switching from Japanese to English text caused the addresses of most of the code in the ROM to change by a few bytes. Badly-written games that jump to specific ROM locations, rather than using the official vectors at the beginning of the ROM, are going to fail. That's what is going on with AIII. Some half-witted idiot at Artdink made a copy of the IRQ1 handler in the Japanese System Card 3, including all of that specific System Card's baked-in subroutine addresses. If that idiot had just changed those addresses to use the officially-documented vectors, then everything would have worked fine on the US Super System Card.
|
|
|
Post by dshadoff on Mar 4, 2019 4:35:03 GMT
In case anybody was not yet convinced, I recently re-dumped all my cards. My US ver 3.0 card matches the GoodPCE dump (after bit-flippling).
The only differences I saw were on other cards: - some US images I compared against were modified to circumvent the protection (no big deal) - I found one byte difference in BeBall, and couldn't corroborate with a second card, because it was brutally corrupt - some of my cards were starting to lose bits/get corrupted
I could find a match for everything outside those parameters... although whether your ROMs match is a different question.
The proper US system Card 3.0 image should be 262144 bytes without header, with a crc32 of 2b5b75fe (in PCE bit-sequence), or 9759a20d (in TG16 bit-sequence).
|
|
fragmare
Punkic Cyborg
Posts: 116
Homebrew skills: Graphics, Music, Level Design, Annoying Programmers
|
Post by fragmare on Mar 4, 2019 4:42:12 GMT
To be quite honest I was mostly worried about BIOS compatibility and hidden pitfalls for current homebrew projects, since I only have a US Duo it's hard for me to test across different devices outside of emulation. You really shouldn't have any problems unless you rely on undocumented behavior, or jump directly into the System Card ROM. If you're worried, you might want to get a Turbo Everdrive 2 for testing, and then you can run all of the different System Card variations. I believe Punch is referring to the "US" region System Card 3.0 ROM that's been modified. Iirc, there's no verified dump of the US System Card 3.0, so a long time ago somebody (dshadoff maybe?) modified the existing dump to work... but it's not perfect. A few games have always had problems, so you must switch back to the Jp region card for them to work. Huh? I have no idea what you're talking about. AFAIK, the US Super System Card was dumped years ago ... the ROM files are out there, and consistent. The reason that TurboGrafx Super System Card image doesn't run some Japanese games isn't anything to do with a bad or nonexistant dump, it's because switching from Japanese to English text caused the addresses of most of the code in the ROM to change by a few bytes. Badly-written games that jump to specific ROM locations, rather than using the official vectors at the beginning of the ROM, are going to fail. That's what is going on with AIII. Some half-witted idiot at Artdink made a copy of the IRQ1 handler in the Japanese System Card 3, including all of that specific System Card's baked-in subroutine addresses. If that idiot had just changed those addresses to use the officially-documented vectors, then everything would have worked fine on the US Super System Card. I see... well, then, I guess the problems are inherent to the US card itself. I just know, from my days on the high seas, that there didn't used to be a verified [!] dump of the US ROM. The US System Card ROM image was always listed as {h1}, which meant the ROM had been modified in some way. The Jp version, however, was verified. My point was, if you were trying to get these problematic games running on the {h1} US version of the ROM, perhaps the issue was the ROM dump of the system card image and not inherent to the US region card itself....
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 4, 2019 4:51:02 GMT
Badly-written games that jump to specific ROM locations, rather than using the official vectors at the beginning of the ROM, are going to fail. That's what is going on with AIII. Some half-witted idiot at Artdink made a copy of the IRQ1 handler in the Japanese System Card 3, including all of that specific System Card's baked-in subroutine addresses. If that idiot had just changed those addresses to use the officially-documented vectors, then everything would have worked fine on the US Super System Card.
The one at $4a44 on boot? LMAO so that's what that was! "it works on my machine bro". This is hilarious...
I'd patch it by simply writing a JMP immediately, in the first instruction of the subroutine, but the IRQ1 handler has some custom code interleaved in it. Can't be that hard to do, right? Unless it has more than one copy of the IRQ handler on the CDROM that gets swapped from time to time.
EDIT: lol are you fucking kidding me, did the programmer at Artdink really copy the whole IRQ1 routine from the syscard just to INCREMENT A VARIABLE AND COMMENT OUT THE JOYPAD READING ROUTINE???
is this real
|
|
fragmare
Punkic Cyborg
Posts: 116
Homebrew skills: Graphics, Music, Level Design, Annoying Programmers
|
Post by fragmare on Mar 4, 2019 5:20:58 GMT
- some US images I compared against were modified to circumvent the protection (no big deal) Elmer: This, right here. ^^ I wanted to make sure you guys were checking against a physical US card or a verified ROM dump thereof. For years, the *only* ROM dump of the US card going around was spoopy. Apparently, from Dave's findings, it doesn't matter, though... so never mind. lol
|
|
gilbot
Punkic Cyborg
Posts: 137
|
Post by gilbot on Mar 4, 2019 5:25:15 GMT
I remember reading somewhere (probably from magazines, but couldn't find this now) BiTD about Pioneer announcing an official list of games not compatible with the LaserActive. Couldn't remember detailed but it was a short list (possibly fewer than 10 entries) and I'm sure that Altered Beast and A Train III were definitely on it, quite possibly expecting certain functions to be at specific addresses as mentioned.
|
|
|
Post by feltempreparatio on Mar 4, 2019 5:27:07 GMT
A.III does seem to have a problem with the syscard3 fix, i can vouch for that. At least the version I have. I believe Godzilla had some sort of issue too, iirc JP Godzilla runs fine on my US Duo, so unless there's some difference between that and a US 3.0 card, I think we can cross that off the list.
|
|