DanBoris

First question, are you putting the correct header information on the image once you have dumped it? The most important bytes are 53 and 54 which determine which type of cart it is. The looping of the Atari logo indicates that the cart isn't passing the validation test. If it was passing the validation, but crashing when the cart started you probably wouldn't get the looping. "Since I am getting the title screen and that beginning screen I feel the rom is probably okay but wanted " This is not necessarily true. The bad sections of the ROM may be in the gameplay area not in the section of the code that displays the title page, etc. Since the real cart is getting through the validation test, at least the first bank of the ROM must be ok, but the validation check doesn't check anything beyond the first bank. Dan

Mspacman definitly uses P/Ms for the player and ghosts. When you get 2 ghosts and a player in the same horizontal region you will notice the flicker start up. So they are basically doing sprite re-use when possible, then switching to page flipping of the PMs when necessary. Galaxian is pretty interesting. It uses mode E for the graphics, so all the invaders are drawn as bitmapped graphics. If you look at the display list, VSCROLL and HSCROLL are enabled on every line and are probably used to move the alien formation around without having to redraw the whole thing every frame. This should leave enough time to paint the attacking aliens in thier appropriate places. Dan

Ah, ok, I don't think that document has ever surfaced. Atari-History.com has managed to come up with Hardware Manuals for a couple other Atari chips, to may someday they will run across this one. Dan

Just goto the link above, scroll down to "Atari 2600 Technical Files" and click on "Stella Programmers Guide". Dan

You can get it from my page: http://atarihq.com/danb/a2600.html Dan

Ok, a few suggestions on this.. - You DO have to write the color values to the shadow registers. As Cafeman pointed out the hardware registers get overwritten by the shadows every v-blank. - Trig0 is not shadowed so you have to read the hardware registers. - Remeber the trig inputs only read the top triggers on the 5200 control. The bottom triggers are read via the keyboard scanning along with the rest of the keypad buttons. Other then this, the code should work as written. You might want to check to be sure the color register isn't getting reset elsewhere in the code. Dan

These are excellent steps to follow to fix most old computer or video game consoles. I have fixed 90% of non-working/twitchy systems using this procedure. Dan Boris

The limit on the length of variables and labels is determined by the specific assembler you are using. In most cases the limits are probably so large (128 character or more) it really would not become an issue. Dan

"Was this related to the fact that all of these games have onboard RAM?" Yeah, I was writing the cart RAM data to the wrong memory address. Really stupid bug, can't imagine how I missed it this long. Dan

I submitted a bug fix to MESS that actually gets a couple more games working. Winter/Summer games and Impossible Mission will be working now. Tower Toppler also works much better now, but it still has some graphics problems. This bug fix should be in the next MESS release. Dan

The .bin file needed it's .A78 header added to it, as well as the validation key to get it working in MESS. I have sent the modified ROM to AtariAge so it should appear here soon. I check the game out in MESS and it seems the same as normal Asteroids, but there are some differences in the rom image so there must be something different about it. Of course both this version and the normal version of Asteroids play almost twice as fast as they should in MESS, still haven't figured out why this happends. Dan

The 7800 was also known at the 3600 for a time, there are some tech docs for the 7800 floating around that call it 3600. Dan

This can be done to some extent on the 2600 with a Starpath Supercharger. You can play most 2K and 4K games with it, but none of the larger bank switched games. Dan

Don't feel bad about having a hard time getting MESS to run. When I started back to work on the 7800 driver a couple weeks ago, it even took me a while to get it going! There was a discussion recently on the MESS development mailing list about the poor quality of the docs. I believe work is going on to fix this. Dan

Do you have the appropriate bit set in the PRIOR register? Also are you trying the sprite overlap on a real 5200 or on an emulator? It's a fairly obscure feature so some emulators may not support it. Dan

There was never a direct port of Rally X, maybe there where games similar to it, but I can't think of any. Rally X would actually be a really good game to port to the 2600, you could probably do a pretty faithful 2600 version of it. Also, on the subject of programming languages. If someone knows Basic, learning C won't help them to much with the transition to ASM programming. Dan

I usually clean the cartridge slot by wrapping a piece of paper towel or tissue around a thin straight screwdriver and wetting that with alcohol. This does a pretty good job at cleaning the contacts in the slot. As for cleaning the top cover, if you are worried about the decal coming off, that portion of the cover actually snaps off so you can take it off before washing the rest of the cover. Dan

Ok, let me see if I can clear up some of the confusion about 5200 carts and how VSS loads them. First, the 5200 carts have two chip enable pins, one for $4000-$7FFF and one for $8000-$BFFF. Each of those ranges is 16K for a total of 32K. 5200 carts come in a number of sizes, 8K, 16K, 32K, and maybe some others. 32K carts are simple, they have 2, 16K ROM chips each using one chip enable line, and filling up the whole address range. 8K carts used a single 8K ROM chip on the $8000-$BFFF chip enable line. But since the address range is 16K and the rom is only 8K, the upper address bit is ignored. The result is that the same data will appear at $8000-$9FFF and $A000-$BFFF. 16K carts get a little trickier. These are normally implemented using 2, 8K ROMS one on each chip enable line. So the data at $4000-$5FFF will repeat at $6000-$7FFF and the data at $8000-$9FFF will repeat at $A000-$BFFF. VSS always assumes that 16K carts are setup like this, so when they are loaded the first 8K in the image file is loaded to address $4000 and the copied to $6000, and the second 8K in the image file is loaded to address $8000 and then copied to $A000. VSS loads the images strickly based on size, nothing in the cart data is used to determine where the cart starts. I am not sure how you would setup an assembler to build 16K binary images properly. OldGuru: In regards to your keypad code not working on a real 5200. Be sure you are setting bits 0 and 1 of address $E80F (SKCTL in the Pokey), this enables keyboard scanning. The emulator doesn't check for this so it will work, but it won't work on a real 5200. As an interesting side note enabling keyboard scanning also enables POT scanning, so you can read the joysticks. This point is not documented in to many places. Dan Dan

Neither is better, they should both run on 5200 emulators. The reason for this was the way 5200 carts are put together, and how they where dumped. The 32K cart was a dump of the entire cart address range, but in some cases the cart only had 16K of ROM in them. Dan

You can get the game here for use on an Atari 8-bit computer emulator: http://homepages.paradise.net.nz/atari/ You can also find out about the history of the game here: http://www.gamesoffame.com/game_design/eas...tern_front.html This is definitly one of the best war games for the Atari 8-bit. Dan

Turning the power on without a game in won't harm the system. Dan

You CAN combine players to get a third color. For this to happen you have set bit 5 of the PRIOR register (C01B) to 1. Dan

It may not even be a bad chip, just a loose one. Most of the chips in the 5200 are socketed. Diassembling the 5200 is not to hard, just a couple screws on the bottom to get the case off. Once you have the case off you will see the circuit board with a shield can covering most of it. You need to be somewhat careful about static electricity when working with circuit boards (although these older systems are not as succeptible as newer electronics). You can discharge and static electricity by touching a metal cold water faucet. Be especially careful of this in the winter when the air tends to be a lot dryer. Around the edge of the can are little metal tabs that have been twisted to hold the two halves of the shield together. With a pair of needle nose pliers straighten out the tabs. This can be a little tedious, just take your time doing it. Once you have the shield off, place the circuit board on a hard flat surface with something underneath it to pad it, a towel works good for this. Push down on each of the chips to seat them into thier sockets. You will actually hear sort of a crunching sound if they are loose. If this doesn't work, removing and re-installing each chip can also help, but this takes a lot more care so that you don't damage the pins. Hope this helps. Dan [ 08-26-2001: Message edited by: Dan Boris ]

You findings are not surprising. The 7800 has all the chips from the 2600 in it plus a few chips more thus it would pull far more power then the 2600 AC adaptor could provide. I am guessing the reason that it works in 2600 mode and not in 7800 mode is that the internal processor clock runs faster in 7800 mode then 2600 mode, and running a chip at a higher frequency generally causes it to draw more power. Dan

On disassembly: I have been using MESS to do disassembly recently. It's debugger has a good diassembler which can write the disassembly out to a file. There is also Dis6502 (http://www.atarimax.com/dis6502/index.html) which can disassemble 5200 and Atari 8-bit carts. Of course generatting the assembly code is only the tip of the iceberg, the hard part is figuring out how it works. Dan