Convert Japaneese Contra (Gryzor) to MMC3

Bregalad · Post by **Bregalad** » Sun Nov 13, 2005 7:54 am

Timing out ??!? What are you talking about ?

kode54 · Post by **kode54** » Sun Nov 13, 2005 10:56 am

Bregalad wrote:That's fault of translators, not mine.

Or is it?

Hmm, I wonder how much work it would be to change the palette on the rest of the letters... ( Patch author complains that nobody came forward to produce a better title, and that the one person who did stopped contacting him at some point and suggested that he was probably going to give up on the project anyway. I <3 the scene. )

Bregalad · Post by **Bregalad** » Sun Nov 13, 2005 12:07 pm

Huh ? From wich ROM did you shot this.

kode54 · Post by **kode54** » Sun Nov 13, 2005 12:29 pm

Contra Japanese, with MadHatter's contraj.ips applied. I assume you used his patch, how many others are there?

Oh, and I stab in Cowering's general direction for Gryzor naming in the database, and in the resulting pack of files that I have.

Post by **tepples** » Sun Nov 13, 2005 12:59 pm

Quietust wrote:1. The Japanese release of Contra IS NOT GRYZOR - it is 魂斗羅, read as コントラ == KONTORA == Contra.

What's next? 塊魂斗羅, read as "Clump soul volume net" == "Clump Contra"?

kyuusaku · Post by **kyuusaku** » Sun Nov 13, 2005 1:02 pm

Bregalad wrote:Timing out ??!? What are you talking about ?

Code: Select all

C:\>ping jonathan.microclub.ch

Pinging www.scdi.org [66.92.187.206] with 32 bytes of data:

Request timed out.

That's what I mean, I also tried a proxy without results.

kode54 · Post by **kode54** » Sun Nov 13, 2005 3:30 pm

What I meant to say was, the pack includes both contraj.ips, which translates it that way, and then there's gryzor.ips, which has the other names and such, and the animated title.

Bregalad · Post by **Bregalad** » Mon Nov 14, 2005 11:14 am

I'll try this patch stuff later, if you guys absolutly want the original "Contra" name.

And, if you don't have a fair conexion, I can do nothing for you :

Code: Select all

Envoi d'une requête 'ping' sur www.scdi.org [66.92.187.206] avec 32 octets de données :

Réponse de 66.92.187.206ÿ: octets=32 temps=382 ms TTL=56
Réponse de 66.92.187.206ÿ: octets=32 temps=204 ms TTL=56
Réponse de 66.92.187.206ÿ: octets=32 temps=202 ms TTL=56
Réponse de 66.92.187.206ÿ: octets=32 temps=206 ms TTL=56

Great Hierophant · Post by **Great Hierophant** » Mon Nov 14, 2005 1:22 pm

Bregalad, wonderful work on the conversion. I played it completely through with FCE Ultra and had no problems.

Post by **tepples** » Mon Nov 14, 2005 1:39 pm

Bregalad wrote:And, if you don't have a fair conexion, I can do nothing for you :

http://www.yousendit.com/ and clones perhaps?

hap · Post by **hap** » Mon Nov 14, 2005 3:32 pm

Great Hierophant wrote:Bregalad, wonderful work on the conversion. I played it completely through with FCE Ultra and had no problems.

Yes, great work. It's working nice here too. That is, after I insert a temporary hack to ignore writes to 0xa001 (or mirrors), since the game writes 0 to it, which disables WRAM, which crashes the game. Howcome this works on a real MMC3 ?

blargg · Post by **blargg** » Mon Nov 14, 2005 10:53 pm

I just checked in my emulator (including disabling the WRAM write routine completely) and it works fine. It is writing 0 to $A001 as you say. Kode54 found that Bregalad had silently updated the data files on his site; perhaps you have the old versions? Here are the checksums on the ROM I constructed from them (and unlike GoodNES, this includes the header!):

73d819d0490b9a35dbfe764c71d6d3ea
d0e831474aeeed9a55edb01512cf4cb780ee7ffa ?SHA1
e45075f7 ?CRC-32

I used these bytes for the iNES header:

4E 45 53 1A 08 10 40 00 00 00 00 00 00 00 00 00

hap · Post by **hap** » Tue Nov 15, 2005 3:05 am

Uh, yes, I had the old versions indeed. It's working ok now.

Bregalad · Post by **Bregalad** » Tue Nov 15, 2005 11:58 am

I just "forgot" to change the last of those $6000 reads/writes to nops. Actually, I replaced the first time by a code disabling MMC3's IRQs and disabling SRAM. Afterwards, I forgot to check other ones. Emulators emulate the thing with SRAM, and that will luckily let the strange $6000 read/writes work fine, but if the first version would run on a card without SRAM, the game will probably crash. The second version is okay. I tested it on hardware, the PRG is working fine but... the screen is messed up. I probably did wrong with CHR /CE rewiring, so the chip interferes with name table reading.... I'm not sure, trough. The sprites are partially good, and partially messed up.

I have to make better description about rewiring on my site, because when I wrote it I hadn't do it yet.

Bregalad · Post by **Bregalad** » Tue Nov 15, 2005 1:56 pm

My problem was that I forgot to cut pin 22 of CHR from GND, and it was still be tied to GND... so I got my output always enable. That's horrible, because I had the choise of desoldering ALL 32 pins without damage the chips, disolver all my wires, cut the track, and resolder everything, but I eventually bent up the pin to solver the wire direcly to it. I had trouble to be able to close the cart with the bent pin, and the pin is pretty much damaged, but I had no choise.
However, graphics are still messed up. I think A10 or A11 is tied to GND, because it's just an adressing problem now. I don't have any idea why, because both pins aren't getting altered by the rewiring at all. I hate soldering and dissoldering.