Zach

Thanks, that means a lot.

It was a good turn-out for the scale of the event. There were lots of systems to play and lots of people to play them. One of the highlights was a talk by Paul Cubbage from the Atari Program Exchange. I found out that the programmer of Caverns of Mars was only 17 or 18. I brought a sample of my homebrew collection (plus Pitfall 2) for people to try out. Unfortunately I didn't get my Krokodile Cart set up, so I didn't get a chance to show K.O. Cruiser. Several people played homebrews, perhaps for the first time. Next time, I'm bringing more multi-player games. Other games that people enjoyed included Tempest for the Jaguar, and Joust for the 8-bit, projected onto a screen. BTW, for you Linux users out there, the organizer of the event was Bill Kendrick, who is the author of TuxPaint.

Overall the event was good. You can see a bunch of photos here: http://www.newbreedsoftware.com/atariparty/2009/photos. I'll post my own photos and a report soon.

I'm going to give a short presentation on Homebrew 2600 programming. http://www.newbreedsoftware.com/atariparty/2009/speakers.php

I'm sure they're not against 5200's. Email the organizer, and ask him if you can bring one.

Just found out about it today. I'm not involved in organizing it, but I'll bring Four-Play and KO-Cruiser! http://www.newbreedsoftware.com/atariparty/

"Funny... He doesn't really look like me..." That one took me a moment.

Thanks for the tip, Darrell. I am getting used to switching between the control and command keys, even though I use the same keyboard for both systems. I'm still thinking about whether to install Firefox, even though Safari is pretty good. If I stick with Safari, I'll need to change the home page. It's no fun being taken to Apple's web site, where they advertise the new Mac Mini.

Yep, this is a real sweeps week plot.

Looks like you've got quite a storyline planned out.

Ouch - just missed the new mini. Zach e-mailed me that he'd bought one, and I wrote back something like, "Of course, within days now you just know they'll release a new model with better specs." Ouch indeed. First thing I thought was, Nathan wasn't kidding. I bought the Mac a couple weeks ago, so it's not as if I just missed the new model by a day.

I am a long time Windows user who just set up a new computer on Saturday, a Mac Mini running on OS X Leopard. I didn't get the Mac to replace my PC, and I plan to use both for a long time to come. In fact, they share the same keyboard, monitor, and mouse. I haven't even used it for 12 hours yet, so here are my first impressions. Pros: Overall easy to navigate operating system. Calculator with programmer mode. Graphing calculator! Built-in dictionary. What took so long? Weather panel. Cons: No built-in simple paint program equivalent to mspaint. Um, didn't Apple pioneer this kind of application? I spent a while looking over free Linux programs before I found Paintbrush. Having to use a different key for keyboard shortcuts. Hopefully there's a way to change this without going too deep in the code. Font in Safari is harder to read than Firefox on my PC. I'm using the same monitor, but I'm getting headaches more easily on the Mac. I'm sure there's a way to change the fonts, but it was nice when Firefox looked good from the start. Safari doesn't display URL's of links before you click them? Maybe I'll install Firefox on the Apple as well. Again, I haven't been using it that long, and these are just first impressions. I haven't even tried iTunes and some of the other software yet.

Would customers have higher expectations for a 32k cart? Especially those who are not following this thread?

Thanks, Thomas. The only compression algorithm I've studied is LZW. What do you like about Huffman encoding?

Thanks Zach! Honestly this project started out as a sprite looking for a game, but as a gamer I am very hardcore when it comes to play mechanics. I think I'm going to pretend the graphics don't even exist until I get the automata up and running. Cheers, Jarod. After you work out the automata, I hope you take a look at the Programming for Newbies guide by Andrew Davie. Programming in assembly and basic are really the same process: moving numbers around, comparing them, and occasionally doing math operations on them. The difference is just that assembly involves paying attention to more details.

It has been over a year since I've posted anything to my blog. Unfortunately I do not have much time for homebrewing these days, but I still hope someday to get back to some of my unfinished projects. A while ago, I took on the challenge of making a chess kernel without flickering or Venetian blinds. After many aggravating attempts, I finally got it with some convoluted code. I know a chess homebrew is low priority compared to the other projects I've started, but I've been thinking about how better chess AI could work on the 2600. There was a brief discussion in this thread about making a better chess game by allowing more RAM and ROM. However, by employing more efficient data storage, it should be possible to work with the standard 128 bytes of RAM. The idea is to encode a chess ply (a move by a single player) in one byte. You need a way to encode a chess move in such a way that can be undone. The most obvious way to encode a ply is to save the chessman's starting square, ending square, and the identity of a captured piece, if any. Normally that would take three bytes, and this encoding would fill up RAM quickly as the computer looks ahead. For a long time, I didn't think it would be possible to reduce a ply to one byte, but I recently figured out how, in conjunction with a 64 byte array to store the board. The idea takes advantage of the fact that pieces don't move to just any square on the board. Each piece is limited to a few squares where it can move at any time. The piece that can move to the most squares is the queen, and the maximum number of places where she can go is 27. -- -- -- 04 -- -- -- 18 24 -- -- 03 -- -- 17 -- -- 23 -- 02 -- 16 -- -- -- -- 22 01 15 -- -- -- 12 13 14 -Q 08 09 10 11 -- -- 21 05 25 -- -- -- -- 20 -- 06 -- 26 -- -- 19 -- -- 07 -- -- 27 -- So, the queen's move can be limited to 5 bits. The queen's starting point does not really need to be saved because she can be found on the board. I do not need to be concerned with the whether the black or white queen is moving, because her color is implied by the position in the search tree. Thanks to pawn promotions, there can be more than one queen, so I need 2 bits to keep track of which queen is moving. (That leads to a maximum of four queens. Even though more queens are legally possible, it is not likely to happen in a practical game. A reasonable compromise, I think.) Finally I need one bit to indicate that the piece moving is a queen rather than some other piece. And so the coding for a queen move would look like this: 1|XX|XXXXX 1 bit: Indicates queen move 2 bits: One of up to four queens 5 bits: One of 28 possible moves. Each of 1-7 squares to the right, up, up-right, and up-left (with wrapping). Similarly, rooks and bishops can move to 14 places, so a rook/bishop move would look like: 0|1|XX|XXXX 1 bit: Not a queen move 1 bit: A rook or bishop move 2 bits: One of two rooks or two bishops (Unfortunately, this scheme does not allow for three or more rooks/bishops, which is legally possible with pawn promotions.) 4 bits: One of 15 possible moves. Whether the moves are read as rook or bishop depends on the value of the previous two bits. There are actually 14 standard moves, and 1 special move, castling. Whether the castling is king's side or queen's side can be ascertained from the rook's position. Next are king/knight moves 0|0|1|XX|XXX 1 bit: Not a queen move 1 bit: Not a rook/bishop move 1 bit: A king/knight move 2 bits: Which king/knight. Up to three knights are possible. 3 bits: The move. Both kings and knights can go to up to eight squares. Finally: Pawn moves: 0|0|0|XXX|XX 1 bit: Not a queen move 1 bit: Not a rook/bishop move 1 bit: Not a king/knight move 3 bits: Which pawn 2 bits: The move. Either 2 steps forward, 1 step forward, or a leftward/rightward capture. The question is now how to store captures. At first I thought I'd need an additional byte to record them, until I realized that there are plenty of empty squares in the board array. Whenever a piece captures, it swaps positions with its captive, and a flag is set to indicate the captured piece is no longer in play. Of course another piece may need to move to that square, so pieces will have to swap with every move, whether a capture or not. Some mechanism will be needed to indicate when an undo returns a captured piece to the board. What I'm thinking now is that three bits can indicate the current depth of the search, and during an undo when it reaches 0 it comes back into play. That allows for a 14-ply tree, which is probably deeper than the Atari can handle in a reasonable amount of time. Undoing en-passant captures is a little tricky, and I haven't ironed out all the details yet, but I'm sure it can be done. Again, this whole article is just a description of what is possible, not a plan to get back to work on chess. If I do find time for homebrewing again, Chetiry will be my first priority.

Sorority event right after soccer practice. No time to wash hair.

Nice game, jrok. In addition to the obvious eye-candy, you've got some innovative gameplay. If I had more free time, I'd volunteer to make a flicker-free kernel for you.

Nice job!

I've uploaded a couple Four-Play clips to YouTube so that you can see the game without an emulator: Title Screen Human (blue) vs. Computer (red) I've been meaning to put clips on YouTube for a while, and your post finally motivated me to get around to it.

Thanks for trying Four-Play. I just tried the copy from the link you provided, and it works on my emulators, Stella 2.2 and z26 2.13 for Windows.

psst, Thomas, Thrust and Jammed are mentioned on page 142, according to the index. Congratulations! I'm glad to see a reference to homebrewing in the book.

Thanks, Albert. And lucifershalo, thank you for your interest in Four-Play.

That was my reason for opting for the iPod touch instead.

If you're getting into Atari because the yacht and the private plane are getting boring we can work out a deal. But for that price you'd be better off getting the original creator to make you a new one.