2600 Rom Comparisions and Dumps

[Omegamatrix]

That's exactly what CloneSpy does.

I was talking about doing disassemblies to see the exact differences, and grouping the roms that way.

 

From the Missile Command output for Clonespy you posted can you say which hacks were built from the no-initial rom, and which were built from the initial rom? It's easy to determine that from a disassembly that can compile into each version. Ultimately though I'm trying to make a simple relationship diagram like this:

 

 

I did that for Challege about 11 years ago.

[+bfstats]

Thomas, I'd love to see CloneSpy produce graphical output that resembles Omega's flowchart.

 

Just a thought.

[Thomas Jentzsch]

I was talking about doing disassemblies to see the exact differences, and grouping the roms that way.

 

From the Missile Command output for Clonespy you posted can you say which hacks were built from the no-initial rom, and which were built from the initial rom? It's easy to determine that from a disassembly that can compile into each version. Ultimately though I'm trying to make a simple relationship diagram like this:

 

Challenge Versions.jpg

 

I did that for Challege about 11 years ago.

Of course disassemblies will always be more precise, but statistical analysis is much faster and gives pretty good results and indications too.

 

As for MC it seems that the Space Robot hacks are based on the NTSC original. Which hack came first cannot be said from the results, they are too close together.

 

BTW: ClonesSpy's result for Challenge are similar to your finding. But I had no HES ROM (update see below). The Zellers version definitely sticks out and the other 3 are a close bunch.

Edited April 14, 2019 by Thomas Jentzsch

[Thomas Jentzsch]

Thomas, I'd love to see CloneSpy produce graphical output that resembles Omega's flowchart.

 

Just a thought.

My spreadsheet is not structured that way. You have to read it more carefully. But you can take the results in the .txt or .csv-file and create a different output, e.g. a tree.

 

I don't come to exactly the same results as Omegamatrix. Zeller seems independent, but the other 4 ones are all together. There are subtle differences, e.g. the "208 in 1 PAL" and "Mein Weg - My Way" are minimally closer to Zellers than "Unknown" and especially "HES", but here only a disassembly can provide more exact results.

 

BTW: The HES version is a split of the 8K ROM, because the 1st 4K contain a different game.

Challenge.zip

Edited April 14, 2019 by Thomas Jentzsch

[+bfstats]

My Excel sheet is not structured that way. You have to read it more carefully. But you can take the results in the .txt or .csv-file and create a different output, e.g. a tree.

 

 

It was a half-serious half-kidding suggestion, but thanks for the response. Your and Omega's contributions are inestimable, and I know I speak for the community when I say to both of you PLEASE keep up the good work!

[alex_79]

There are new dumps of a 2in1 multicart by "Rentacom" from brasil in this thread:

They seem to use the same bankswitching scheme of these other two roms posted here in 2014:

 

 

Just like the "Fathom" rom posted in the old thread, the banks are in reverse order compared to the original F8 versions. Anyway, with these new dumps we have confirmation that they are actually arranged in that way in the eprom (as they were dumped directly from the chip).

So $2A0 is the hotspot for the lower 4k of the rom, while $2c0 is the one for the upper one.

The "Mickey"(sorcerer's apprentice) rom in the old post has the banks in the same order as the F8 version, and the hotspots are swapped. This might mean that it was dumped incorrectly (by setting the hotspots in the wrong order to dump each bank) or that both variations of this scheme exist (which is likely: it's just  a matter of swapping two address lines to invert the behaviour).

 

 

[Rom Hunter]

Thanks, Alex.

So is there also an original (merged) 16K original ROM dump of Fathom and Mickey?

Like the Time Pilot - Vanguard one?


 

[alex_79]

I have no idea. The bankswitching I was referring to is the one used in each game, and it's not related to them being in a multicart (The two games are selected purely by hardware using a physical switch visible in the pic posted by coderbr, and there isn't any change in the roms required for that to work)

 

The original poster only mentioned he found them in a brasilian website, and there's no way to tell if they were part of a multicart or not just by looking at the roms

[Tempest]

Anyone see any differences in this Pole Position proto?  It tests differently than the final, but I can't see any changes.

newpp.bin

[alex_79]

It's the same as the released version. The rom only differs in two byte corresponding to the hotspots, which are zeroed in the known dump.

It's a just the result of a different dumper used to obtain the rom (like when two dumps differ in areas where RAM is mapped).

 

[Tempest]

Ah ok.  Thanks for confirming.

[Omegamatrix]

On 7/26/2019 at 12:35 PM, alex_79 said:

Just like the "Fathom" rom posted in the old thread, the banks are in reverse order compared to the original F8 versions. Anyway, with these new dumps we have confirmation that they are actually arranged in that way in the eprom (as they were dumped directly from the chip).

So $2A0 is the hotspot for the lower 4k of the rom, while $2c0 is the one for the upper one.

The "Mickey"(sorcerer's apprentice) rom in the old post has the banks in the same order as the F8 version, and the hotspots are swapped. This might mean that it was dumped incorrectly (by setting the hotspots in the wrong order to dump each bank) or that both variations of this scheme exist (which is likely: it's just  a matter of swapping two address lines to invert the behaviour).

 

 

Nice! So it is a new bankswitching then.

 

If I read your post correctly then the roms the original poster had for Fathon was correct, but Mickey needs the banks reversed?

[alex_79]

On 8/7/2019 at 9:47 PM, Omegamatrix said:

If I read your post correctly then the roms the original poster had for Fathon was correct, but Mickey needs the banks reversed?

I think we can't say for sure unless someone who has access to the actual carts can check. It's possible that both variation of the scheme existed, as I said it's just a matter of swapping two signals in the circuit.

 

The current Stella build on github supports both versions: the "UA" banking scheme has been generalized so to only consider address lines A12,A9,A6 and A5 and it works with the two new dumps as well as with the "Fathom" rom the original poster uploaded. Moreover there's an additional "UASW" (UA with swapped banks) scheme that can be selected to run the "Mickey" rom.

 

I'd say to consider the roms as they were originally posted as the official versions, unless/until further info is found.

 

 

P.S. Sorry for the late reply, I was on vacation and missed your post.

[Tempest]

Can someone do a Clonespy on that new Super Cobra demo that was found and the released version?  I'm curious if they're actually based on the same code or not.

[Thomas Jentzsch]

On 8/28/2019 at 4:21 PM, Tempest said:

Can someone do a Clonespy on that new Super Cobra demo that was found and the released version?  I'm curious if they're actually based on the same code or not.

CloneSpy does not find any relation between the two. So I suppose they started from scratch.

Edited August 30, 2019 by Thomas Jentzsch

[Tempest]

Interesting.  Thanks Thomas, that answers some questions but also brings up new ones.

[+Psionic]

Actually, I spoke with the programmer of the released version again last week and he said that his version was coded from scratch at Roklan, so that makes sense.  He had told me previously that his version was started by someone else, but he corrected himself and now states that was a different game he was thinking of.  He was never handed a WIP of Super Cobra.

The demo that was found is almost certainly a Parker Brothers creation exclusively.  According to the PB people, the guy who was originally working on the game there left the company before it was completed, so the details of the story all mesh together.  But now we know there were two separate versions and that the first one was apparently totally abandoned.  Tempest should update his review to reflect this.

[Tempest]

Yeah I'll do that, but it still seems weird that they're both so similar.  The new programmer had to have used the proto version for at least a visual idea of how it should look.  There's no way that they're so similar by coincidence. 

[+Psionic]

Graphics and code are two different things and I'm just going by what I was told.  Paul claims he doesn't recall being given an actual working demo of the game.  All he was given was some documentation, which may have included paper charts of the game graphics.  I had sent him a link to the screenshots of the demo to confirm or deny if he ever saw it but I haven't heard back.  I'll give it another week or two before I bug him again.

[Tempest]

Posted an odd hacked prototype (I think) of Pitfall over in the proto forum.  Maybe someone here can take a look at the rom and see what the heck they were trying to accomplish:

 

 

 

[bobcurtiss]

On 8/30/2019 at 3:35 PM, Psionic said:

Graphics and code are two different things ...

Generally true, but not for the 2600. Code that generates the playfield, especially in a side scroller like Super Cobra, should reveal a few clues about whether it was done from scratch or partially copied.

[+Psionic]

4 hours ago, bobcurtiss said:

Generally true, but not for the 2600. Code that generates the playfield, especially in a side scroller like Super Cobra, should reveal a few clues about whether it was done from scratch or partially copied.

Very true, I stand corrected.   To be fair, when I spoke of "graphics" I was referring to the sprites specifically and whether they may have been directly copied pixel-for-pixel (or nearly so).  As Thomas Jentzsch stated above, an analysis of the two binaries using CloneSpy apparently found no coding similarities.

This is not all that surprising since the other programmer besides yourself who worked on this game at Roklan told me he never saw the demo version (which I assume was created internally at Parker Brothers).  Based on your response, I assume you never saw it either (until now).

[bobcurtiss]

46 minutes ago, Psionic said:

Very true, I stand corrected.   To be fair, when I spoke of "graphics" I was referring to the sprites specifically and whether they may have been directly copied pixel-for-pixel (or nearly so).  As Thomas Jentzsch stated above, an analysis of the two binaries using CloneSpy apparently found no coding similarities.

This is not all that surprising since the other programmer besides yourself who worked on this game at Roklan told me he never saw the demo version (which I assume was created internally at Parker Brothers).  Based on your response, I assume you never saw it either (until now).

Would you happen to have contact info for that other programmer? Are you referring to Paul Crowley?

[+Psionic]

3 minutes ago, bobcurtiss said:

Would you happen to have contact info for that other programmer? Are you referring to Paul Crowley?

Yes, I can put you in contact with Paul.  Send me a PM.

[Rom Hunter]

These last third party PAL ROMs have never been dumped:

- Coco Nuts (Telesys)

- Plate Mania (Home Vision)
- Stargunner (Telesys)
- Stronghold (CommaVid)
- Topy (Home Vision)
- X’Mission (Home Vision)

The Home Vision ones are PAL originals.

If anyone with a dumper has the chance to .bin these, please let me know.

Also interested in a ROM dump of Air Battle:
http://www.atarimania.com/game-atari-2600-vcs-air-battle_19390.html

Edited December 30, 2019 by Rom Hunter