Rom Hunter's V16 Old Era ROMs Collection Now Online!

[Rom Hunter]

Merry Christmas and a Happy & Healthy New Year!

Be sure to check this one out, because it contains a lot of new beta .bins!

Many thanks to Matt Reichert, Scott Stilphen and all other contributors.

If you want to add or correct any .bin file info, please do this with 100% verified documentation and I will be more than happy to include this into V17.

Enjoy:
http://www.atarimania.com/rom_collection_archive_atari_2600_roms.html

Rom

Edited December 22, 2019 by Rom Hunter

[5x7]

Thank you!!!

[Thomas Jentzsch]

Do you have a delta list somewhere?

[Rom Hunter]

5 hours ago, Thomas Jentzsch said:

Do you have a delta list somewhere?

Nope.

[Ballblaɀer]

Attached is a simple .htm report that I generated with WinMerge to show what files are new/different between v15 and v16.

 

Not all files listed are entirely new. Just as one example, Confrontation is included on the list because new information was added to the filename. A different example: it looks like the review/first-run copy of Gangster Alley that I dumped was given the old, "original" filename, while the regular release has now been given the [fixed] notation.

 

So yeah, this isn't meant to be comprehensive but will at least provide some guidance as to what has changed.

compare.htm

[Shannon]

Yep... definitely a lot of prototypes!  Thanks for this!

 

[Tempest]

A lot of those prototypes came from the Joe Gaucher archive and unfortunately only differ from the released versions of the games by a few bytes.  If there's a prototype in this collection that it appears I missed, that's probably why.

[Thomas Jentzsch]

At least we got the originals.

 

BTW: Some Amiga games may not play in current released version of Stella (it will if you create your own from the code). They use a new bankswitching which I called "FC".

[Tempest]

There were two new prototypes in this batch that I hadn't seen before.  Smurf and Q*Berts Qubes.  Since they were both done by Wickstead Design I assume they came from the same source, but I'm not sure who that is.

[Thomas Jentzsch]

Don't forget the new found title screen for SAC.

[Tempest]

1 hour ago, Thomas Jentzsch said:

Don't forget the new found title screen for SAC.

Is the PAL rom the same other than that title screen?  No matter what I do with Stella the colors seem wrong (pink sky and all that).

 

The title screen song changes when you select the other option.  For the air version it plays the Air Force song, for the water version it plays the Navy song.  That's cool!

[edladdin]

Thank you so much! Merry Christmas to you too. 

[Tempest]

I'm starting to wonder if the colors for the PAL version are actually correct.  If set to NTSC the colors look really wrong and everything scrolls.  Maybe they decided to change the game to take place towards dusk for some reason rather than daylight?  The PAL version seems to be a little more polished than the NTSC version and is a bit easier to play too.

 

NTSC

 

PAL

[+stephena]

On 12/22/2019 at 11:53 AM, Rom Hunter said:

Merry Christmas and a Happy & Healthy New Year!

Be sure to check this one out, because it contains a lot of new beta .bins!

Currently working on getting the changes integrated for Stella 6.1.

[Thomas Jentzsch]

The PAL colors are most likely wrong. If you switch Stella to NTSC 50, you can see that the color codes have not been changed for PAL.

Edited December 23, 2019 by Thomas Jentzsch

[Tempest]

1 hour ago, Thomas Jentzsch said:

The PAL colors are most likely wrong. If you switch Stella to NTSC 50, you can see that the color codes have not been changed for PAL.

So the scanlines are PAL but not the colors?  Weird.

[+stephena]

Changes are now in for Stella 6.1.

[Rom Hunter]

5 hours ago, stephena said:

Changes are now in for Stella 6.1.

Many thanks, Stephena.

Please know that Mickey (AKA Sorcerer's Apprentice) (Digivision) only works with UASW (8K UA Swapped Banks) and Super Soccer (AKA RealSports Soccer) (Digivision) only with F8 (8K Atari)

These two ROMs were most probably not dumped correctly.

Now what I like to know, is which specific differences the [a] ROMs contain, so I can put that in between the [ ] instead of just [a1], [a2], etc.

Unfortunately I don't have the knowledge to look into that, so help with that would be much appreciated.

Edited December 24, 2019 by Rom Hunter

[Thomas Jentzsch]

Small correction: Super Soccer was incompletely patched (most likely after dumping) to F8 bankswitching. So it contains remains of the original FC bankswitching, which trigger Stella's auto detection. 

 

Would be nice to get hands at the original, non patched dump. 

[+H.E.R.O.]

ty, Mr. Rom Hunter, sir.  Been grabbing these sets since 9.0. Exemplary work as always.

[Tempest]

6 hours ago, Rom Hunter said:

Now what I like to know, is which specific differences the [a] ROMs contain, so I can put that in between the [ ] instead of just [a1], [a2], etc.

Unfortunately I don't have the knowledge to look into that, so help with that would be much appreciated.

My biggest regret is that I can't look through the code of these games and see what the changes are. All I can do is play a protoype side by side with a released version and look for differences.  I really need to team up with a coder to look through these roms and find all the hidden stuff in the code. 

[Thomas Jentzsch]

@Tempest
There is a mistake on your website for S.A.C. Alert: "Set your emulators to PAL50 NTSC50 to see the correct colors."

[Thomas Jentzsch]

On 12/24/2019 at 8:47 AM, Rom Hunter said:

Now what I like to know, is which specific differences the [a] ROMs contain, so I can put that in between the [ ] instead of just [a1], [a2], etc.

Since we cannot analyze each code, CloneSpy should be able to help us here. Usually a game extends over the time, new code and data are being added much more frequently than getting removed. Therefore CloneSpy always gives you two percentage numbers (how much code of A can be found in B and how much code of B can be found in A). If more code of A can be found in B than vice versa, the B is most likely a later version. Additionally CloneSpy tries to order the versions accordingly in the CSV output. Also further apart versions usually have less code in common than closely related ones. This is also reflected in the ordering.

 

Here is an example:

		A	B	C	D	E	F	G
A-Team, The (AKA Saboteur) (05-08-1984) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26133) (Prototype)	A	.	84	82	69	52	38	-
A-Team, The (AKA Saboteur) (03-30-1984) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26133) (Prototype)	B	83	.	89	73	55	41	39!
Saboteur (Sabotage) (12-20-1983) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26119) (Prototype) ~	C	79	88	.	80	60	44	43!
Saboteur (Sabotage) (09-02-1983) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26119) (Prototype)	D	69	75	82	.	66	48	45!
Saboteur (Sabotage) (06-15-1983) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26119) (Prototype)	E	47	50	55	59	.	60	49!
Saboteur (Sabotage) (06-09-1983) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26119) (Prototype)	F	31	34	37	39	56	.	65
Saboteur (Sabotage) (05-20-1983) (Atari, Jerome Domurat, Howard Scott Warshaw) (CX26119) (Prototype)	G	-	28!	31!	32!	39!	57	.

 

Here ClonesSpy has already ordered the games correctly (it is not using the dates!). The top version (A) is the latest one. Why? If you compare the values at column and row A, you will find that the values in row A are always larger than or identical to the values in column A. So A contains more code of B..G than vice versa. If you do the same with B..G, the results will be (about) the same (there is are no values for A vs. G, because both values are below the threshold of 33%).
(BTW: Ignore the '!', they are just a hint into the same direction, but less accurate and detailed)

 

There are some exceptions, no development is perfectly forward only. E.g. code which is not needed anymore is getting removed, old code will be refactored to save space etc. But on average the values match.   

 

Finally the numbers together with the known dates can give you an idea about the development cycles. E.g. the gap between E and F is quite large for just 6 days, meaning that quite intensive development was going on during that period.

 

So that should help answering your question.  

 

Note: This is a solely statistical approach, so individual results may be completely wrong. But I found it overall very reliable and informative.

Edited December 25, 2019 by Thomas Jentzsch

[Tempest]

4 hours ago, Thomas Jentzsch said:

@Tempest
There is a mistake on your website for S.A.C. Alert: "Set your emulators to PAL50 NTSC50 to see the correct colors."

Doh!  I'll fix that.  Thanks.

[Thomas Jentzsch]

1 hour ago, Tempest said:

Doh!  I'll fix that.  Thanks.

NP

 

There is another one for S.A.C. Alert: "...South America where they use a standard called PAL-50 60". And because of this, your theory makes little sense. Colors are wrong and frame rate too for PAL-60. No idea why anyone would develop for NTSC-50.

Edited December 25, 2019 by Thomas Jentzsch