Zach

Although A* will be faster, a brute force complete search will be simplest to program. Since you only have to solve each puzzle once, it may not be that bad. You might even be able to recruit computing time from fellow AA'ers.

Sounds cool. Although I couldn't make it to Austin, some of my work did. I'd love to have a copy of the program. BTW, Tank AI is slowly being resurrected. I'll post an update on my blog soon

Hey Al, your programs don't by any chance have a list of the demos you showed at the expo, do they?

Then cheat and write small program which finds the optimal solutions for you. Then you have the numbers. 896893[/snapback] Coincidentally, there is a common search technique you could use called A* (A star).

I was wondering about that. I figured if anyone knew, you would, Batari.

I went to this event in 2001 and 2002. Had a great time.

That was a fun puzzle. Nice job. It didn't take too long to solve the first level. I trust you will be adding graphics for when Pac-Man faces up or down, or will you do it Todd Frye style?

I still don't get how RESP works. From what I read, the object is moved to the cycle where data is written to RESPx, but when I set the ball to go to cycle 28 (marked by a PF color change), it appears about four clocks later, as shown in the code and screenshot below. I'm sure I could do empirical tests to find out where objects land on each cycle, but I'd rather understand what's going on. Can anyone explain how to determine where an object will be moved using RESP? sta WSYNC ;-------------------------------- sleep 25 sta RESBL; BL set at cycle 28 sta WSYNC lda #2 sta ENABL sta WSYNC lda #%10101010 sta PF1 lda #%01010101 sta PF0 sta PF2 sta WSYNC ;------------------------------- lda #%11111111 sta PF1 lda #%11111111 sta PF0 sta PF2 sta WSYNC sleep 25 stx COLUPF; PF color changes at cycle 28 sta WSYNC

I use a boombox with my Supercharger. It's an old Sony CFD-770 with two tape decks and a CD player. I'm using the CD player with a copy of Stella Gets a New Brain.

Wow! I hadn't noticed Robert's 13 char display before. Robert, if you're reading this, I applaud you. I didn't even realize that resetting the ball took effect immediately. I thought the multi-sprite trick only worked for objects that could be repeated with NUSIZx. That information will be useful for future kernel designs. There is some similarity to Robert's kernel in that I use masking. Here is how the 8 stone kernel works. PF is set to |0111|01110111|01110111| and is repeated on the right side. At the beginning of the line COLUPF is gray and is then reset to the color of each stone. COLUBK is set to black, and each gap in the PF is covered with a gray mask. The ball is on the far left, followed by six player sprites set to 11100111. So far this creates six masks to round off the stones, and six columns. The remaining three masks are done with missiles. M1 is 8 clocks wide and M0 is on top at 2 clocks wide. There are just enough cycles in between all the STx COLUPF's to reset M0 to black. At the end, COLUBK is set to gray to close off the last gap. To round off the stones, asimple change is made to priority. At the upper and lower parts, the sprites are in front. In the middle PF is in front. I'll keep adding updates as I make progress.

Today I figured out how to implement my original design for the kernel. This design has evenly spaced squares, and the grid can be any color. I had posted a mockup in this thread, and the actual kernel is very close to what I had planned. As I said it is very easy to adapt the code to an 8x8 board as well. Before I explain how it works, I'll see if anyone wants to try to figure it out. At least half the fun of Atari programming is trying to figure out what the machine can do, and how do to it.

Thanks to everyone who tested on hardware. Today I finished up the Tank version. Assuming no problems are found, the plane version will be posted soon. When that is done, I will release a tool for making your own playfields. Updates to this version include: Removed airplane levels. Made game reset from attract mode possible with left joystick button. Made game variations selectable with left joystick. Reduced picture size from Combat's original 263 scanlines to the standard 262. Sprite data stored more efficiently. Fixed a subtle bug that was only apparent when rotating a tank in slow motion. at_tank_jul9.zip

Well done indeed. I look forward to seeing what games the community makes with this.

I'll look up all the games on KLOV before voting, but I'm already leaning towards Qix, which only ought to be done with extra RAM and flickering, like Stell-A-Sketch or Video Life.

Thanks Cybergoth. I just figured out that Planet Bob is yours. Had your blog been called Planet Manuel, I would have recognized it more quickly. I'm taking a look now...

Greetings Mr. M-U-N-D-S-C-H-A-U Good luck with your ideas. That'd be cool to get a 1K game in the competition. Four player Indy 500 sounds especially interesting, and could lead the way to a four player Combat.

Hey! Ataxx is pretty cool! I'm surprised it isn't better known. I made a mock-up that should be feasible. I can think of a few other games that could be played on this board, such as Vertical Four in a Row, Reversi, and Hip.

Hi Neoscott! Congratulations on your find. It's nice to see another Sacramentan around here. I wonder if the cartridges came from Sirius Software? Anyway, Let me know if you'd like to use my Atari to take screenshots. I live near Sac State. Zach

You are right about the use of sprites and balls. It is a complicated kernel, but here's how it works briefly. The stones and grid are drawn in the playfield, and the corners of the stones are masked by sprites and missiles. There are three different routines to draw the go board: a) The scanlines between stones b) The scanlines where a horizontal line of the board is drawn c) The scanlines where stones are drawn above and below the horizontal lines Routine a is the simplest. The playfield and ball registers are set and left alone for several scanlines while variables used for b and c are calculated. Routine b is an asymetrical playfield where COLUPF is black and COLUBK is white (sometimes vice-versa). There are carefully timed color changes at both edges of the board. Routine c is a combination of an asymetrical playfield and quick color changes to the PF using STX and STY. The trick is to run a small part of this code in RAM so that STX can be used for white and STY can be used for black. Without running code in RAM, I don't think this kernel could have worked (not in 4K anyway). This explanation is a bit of a simplification, but it gets across the main points. Here's the source code if you want to look into it more. Once again, it's complicated. go9x9_mar30.zip

Well, eBay spotlights have been on the front page for a long time, and so have ads for the AtariAge store. So the question is not whether there should be ads, but what is an acceptable level of advertising? I'm glad to hear that Al is against the most egregious kinds. Albert, you do an amazing job with this site. I've caught a glimpse of how much work is involved from the few projects where I've helped out. You deserve far more than breaking even for the contribution you make to the Atari community. If a few non-intrusive ads help generate revenue, then I'm all for it.

Rarest title: Crash Dive (5) Rarest label variation: Demon Attack Blue (6) Rarest Homebrew: Qb, AA edition Rarest Prototype reproduction: NTSC Klax

Cool artwork. I refuse to be the one to spoil the mood by pointing out that a kangaroo's mammary glands are in her pouch.

I was wondering when someone was going to bring this up. It's an interesting topic, a debate about which is worth its own thread. @Andrew. I don't know much about Boulderdash, but I'm certainly impressed by your videos. Everyone seems to love BD, so I'll do my homework someday to see what the fuss is about.

When do we find out who programmed Caverns of Mars and Lunar Lander?

I just successfully loaded the new Combat on a Supercharger. It is becoming evident that the hardware issue has something to do with the Krokodile Cart. (It's still a great piece of equipment, though.)