7800 Impossible Mission Memo

[+Mitch]

Can digitized voice even be done with a 7800/2600?

 

Check out 7800 Jinks.

 

Mitch

[krewat]

Check out 7800 Jinks.

 

Is that done with the standard sound, or a POKEY?

[Shawn]

Check out 7800 Jinks.

 

Is that done with the standard sound, or a POKEY?

 

 

Standard TIA can do voice, pokey not needed.

[BrianC]

Check out 7800 Jinks.

 

Is that done with the standard sound, or a POKEY?

 

Standard sound. The only officially released games with the POKEY were Ballblazer and Commando.

[DracIsBack]

Can digitized voice even be done with a 7800/2600?

 

Check out this little homebrew demo - done with a 7800, a 512K cart and TIA.

 

[krewat]

Check out this little homebrew demo - done with a 7800, a 512K cart and TIA.

 

 

HOLY freakin' crap!

 

In the face of overwhelming proof, I guess I need to do some real work

[+rdemming]

I was never involved in "fixing" the PAL version, and I don't think they put together a PC and 2500SD X6502 compiler(assembler) to actually recompile the code, so it was definitely a binary hack. Either way, I still want to get out a fixed version of IM, with other embellishments. Not sure what those will be, but we'll see what I can accomplish.

 

 

Actually, I was the guy that made a fixed NTSC version a few years back (the famous "Possible Mission").

I just converted the PAL stuff to NTSC. It's 100% playable and beatable. I believe the ROM is on Mitch's site.

 

As krewat mentioned, it was a binary hack; not source code.

So, it was just twisting some color tables, and a few DLL lengths, which were not too hard to find.

 

 

I think Krewat was saying that Atari's PAL version was a binary hack of the NTSC version as he was not involved in the PAL version. So the "Possible Mission" NTSC was a binary hack of a binary hack

That makes me wonder how they could make the (fixed) PAL version from the NTSC binary without recompiling the source. I assume Atari had the source but was Atari really unable to compile it because they didn't have the X6502 system Krewat used?

 

Robert

[krewat]

I think Krewat was saying that Atari's PAL version was a binary hack of the NTSC version as he was not involved in the PAL version. So the "Possible Mission" NTSC was a binary hack of a binary hack

That makes me wonder how they could make the (fixed) PAL version from the NTSC binary without recompiling the source. I assume Atari had the source but was Atari really unable to compile it because they didn't have the X6502 system Krewat used?

 

Robert

 

Well, I definitely was NOT involved in the PAL version. Haven't looked at it in detail neither, so I can't say if it's identical to the NTSC code except for graphics.

 

But I will say, that the "fixed" NTSC version from Atari (that someone posted here, after finding the ROMs in the garbage, I think) involves a couple of NOPs, and a JSR to a different location, obviously being a binary hack. The rest of the code is identical down to the last bit, except for the hack itself. So, there's almost no way they did anything with the source code.

 

And, if I recall correctly, I don't think we ever sent the source code to Atari. They only wanted the final binaries to burn to EPROM for testing.

[Lynxpro]

I really have to document this stuff, I keep saying things that are false

You are probably getting it confused with the Atari fixed version of NTSC Impossible Mission that introduced a new bug. It is unknown if this version ever made it into production.

John's version is definitely fully working.

Mitch

 

 

I'm surprised that there wasn't a class action lawsuit/voluntary recall on the cart. Then again, I didn't even hear about the bug until after I had already moved onto the Lynx and had given up on the 7800 for the most part...

 

 

 

And, if I recall correctly, I don't think we ever sent the source code to Atari. They only wanted the final binaries to burn to EPROM for testing.

 

 

I should add this tidbit to my Blundering Tramiels list...

 

Why the hell wouldn't a video game company not want source code???

 

 

Can digitized voice even be done with a 7800/2600?

Check out this little homebrew demo - done with a 7800, a 512K cart and TIA.

 

 

That is incredible. Are you sure the TIA is doing all the work or is the 6502c responsible? And is that 512Kb or KB?

Edited March 22, 2010 by Lynxpro

[DracIsBack]

That is incredible. Are you sure the TIA is doing all the work or is the 6502c responsible? And is that 512Kb or KB?

 

I didn't put it together. I only know it's 512K and no POKEY is necessary in my Cuttle Cart for it to run.

[krewat]

 

I'm surprised that there wasn't a class action lawsuit/voluntary recall on the cart. Then again, I didn't even hear about the bug until after I had already moved onto the Lynx and had given up on the 7800 for the most part...

 

SNIP

 

I should add this tidbit to my Blundering Tramiels list...

 

Why the hell wouldn't a video game company not want source code???

 

If I remember correctly, and that's a BIG IF, when Computer Magic asked me about fixing the bug, I said "pay me". They wouldn't. Even for the few hours it would take to fix it.

 

I've thought long and hard about this whole thing, and I can clearly recall FINISHING the 7800 version of the game. Either I was very lucky and none of the puzzle pieces wound up hidden in unsearchable terminals, or ... something happened that I reengineered the code to save RAM and introduced the bug. I need to go back and look at the C64 version and see how it does puzzle pieces.

 

As for the Blundering Tramiels, well, I don't think it was so much the Tramiels. It was the people who worked at Atari not really giving a crap how they got something finished, just that it was. Which is probably the Tramiels' fault, ultimately, because everyone seemed to be running around like scared rabbits.

[gdement]

That is incredible. Are you sure the TIA is doing all the work or is the 6502c responsible? And is that 512Kb or KB?

The TIA can't do all the work, it just emits whatever sound is indicated in it's registers at that moment. Whenever the CPU changes those values, the sound will change. If it happens fast enough you can get enough of a sampling rate to do speech.

 

The TIA limits the size of the samples (5-bits I think). The CPU speed limits how much sampling rate you can achieve.

Edited March 22, 2010 by gdement

[commodork]

 

I'm surprised that there wasn't a class action lawsuit/voluntary recall on the cart. Then again, I didn't even hear about the bug until after I had already moved onto the Lynx and had given up on the 7800 for the most part...

 

SNIP

 

I should add this tidbit to my Blundering Tramiels list...

 

Why the hell wouldn't a video game company not want source code???

 

If I remember correctly, and that's a BIG IF, when Computer Magic asked me about fixing the bug, I said "pay me". They wouldn't. Even for the few hours it would take to fix it.

 

I've thought long and hard about this whole thing, and I can clearly recall FINISHING the 7800 version of the game. Either I was very lucky and none of the puzzle pieces wound up hidden in unsearchable terminals, or ... something happened that I reengineered the code to save RAM and introduced the bug. I need to go back and look at the C64 version and see how it does puzzle pieces.

 

As for the Blundering Tramiels, well, I don't think it was so much the Tramiels. It was the people who worked at Atari not really giving a crap how they got something finished, just that it was. Which is probably the Tramiels' fault, ultimately, because everyone seemed to be running around like scared rabbits.

 

It could be that the NTSC 7800's architecture is screwing with how the program code is executed.

Edited July 30, 2010 by commodork