Special thanks to Nathan Strum for the artwork.I'm sure that experienced programmers can imagine how to do a 51 pixel bitmap without flickering. Anyone care to guess how I managed 52? EDIT: Now explained below.
I recently pulled out my Dark City DVD, which I haven't watched in years. I had forgotten how good it was. Truly brilliant and definitely underappreciated.I've also been watching Mad Max 2 (The Road Warrior) a lot lately. I noticed for the first time that the dune buggy Papagallo drives at the end was inherited from Humungus' gang. One of them drives it into the compound during the skirmish where Max delivers the rig. OK, not exactly an epiphany, but it's always fun to notice something new about
My first attempt at stabilizing the frame height while Four-Play is thinking did not go so well. From what I understand, a frame lasts for 19912 cycles (76 cycles/line * 262 lines). To set the timer to last a whole frame, you would write #18 to T0124T. Then the timer would count down to 0 for 19456 (1024*19) cycles. At this point the timer is set to 255, and counts down once per clock, so it would loop from 255 to 0 twice in the last 456 clocks of the frame.In order to write to VSYNC at the corr
I can't prove it, but I had plans to post a to-do list before Cybergoth did so. The list is fluid and likely to change.1) Four-Play (Summer 2006)2) Breakthrough (1K minigame contest deadline, if FP is finished on time)An obscure but fun board game invented by Dan Troyka.3) Marble Jumper 2 (1K minigame contest deadline, if FP is finished on time)I won't spend too much time on this. Just some updates and new puzzles.4) Reversi Attack (TBA)A new twist on Reversi, conceived by AA member yuppicide.5
BreakthroughGame design by Dan Troyka (playable demo with dumb AI)Breakthrough is a fast paced board game with easy-to-learn rules. The winner is the first to get one of 14 pieces across to the last row. Pieces move forward one square vertically or diagonally, and they can capture diagonally. Because pieces always move forward, games end quickly. There are no draws.Your side is Silver. To move a piece, place the cursor there, hold the fire button, and push left or right to move diagonally or up
Here is the latest version of my four-in-a-row game. It is a challenge to win, but I'm able to do it from time to time. Usually the wait between moves is less than 20 seconds, and the longest time I measured was about 35.I am curious how people feel about the difficultly level now. I hope Thomas has some time to try it out, since he consistently won against an earlier release.Some other new features include a screen that starts out white and fades to black while the Atari is thinking. This gives
Just wanted to let you know that I'm still working on my homebrew. I've got a version that takes at most half a minute to play and is a challenge to beat, a vast improvement over the 3.5 minutes the last version took.I spent some time during the holidays working on alpha-beta pruning and got it working. This pruning allows the computer to recognize moves that will not be of any benefit and saves searching time. Also I've been able to put in a more intelligent heuristic. There is still work to d
In my last attempt at a chess kernel with playfield graphics, I came up with a technique that would only work with about 27K of code. Since then I've made some advances that make a 4K kernel possible. One trick I developed was a new 32 pixel asymmetrical playfield configuration that takes four loads and six stores. (I believe it's new at least.)
The playfield remains transposed rather than reflected, but unlike before I load the same byte for the right PF0 and PF2 data. The high nybble goes
The good news is I have made Four-Play smarter. If you can defeat it now, you know something about the strategy that I don't. The bad news is it takes about 3 1/2 minutes to plan a move. You can reduce the time by setting the frame rate to maximum in z26, for example "z26 -r1000 fourplay_dec11.bin". My pentium 4 unit can plan moves in about 30 sec.The next step will be to try alpha-beta pruning to see if I can cut the time to a reasonable amount. If not, I'll have to choose a different heuristic
Sometimes designing kernels is more fun than actual game programming. Anyway, the idea of doing a better chess kernel has been on my mind for a while. I've been skeptical of the notion that chess could only be done with Venetian blinds or flickering, and this weekend I came up with a design that needs neither. Like my other board game kernels it relies heavily on PF graphics and multiple color changes.Unfortunately I could not figure out a way to use self-modifying code. The PF must be updated e
About 2 1/2 years ago, AA member Atari Rescure Group asked in the forums whether Pente was possible for the 2600. At the same time I had been starting to think about how to port Go. The board I proposed back then was a monochrome 80 pixel wide sprite using flickering.I've learned a lot about VCS programming since then, but a board without flickering had still eluded me until this weekend. I knew I could set up 19 pieces with the PF, but there was no way to change the colors every 8 pixels (it ta
I've applied the classic minimax search to Four-Play, and I'm pleased with the results. The program will anticipate any traps within 7 moves and act accordingly. I was concerned that it might take a long time to search 7 moves ahead, but it turned out not to be too bad.The AI is not terribly difficult to beat at this point because the heuristics are simple. The search just checks whether there is a winning position or not. The next step will be to apply some more sophisticated heuristics, and t
I think I've found my next homebrew project after Four-Play. It's an original concept, and the attachment is a rough prototype that doesn't do much...In the remote mountains of Kashmir, maverick treasure hunter Jack Rascly has found the final resting place of the last survivors of Mu, the ancient empire which sank into the sea 12,000 years ago.The people of Mu clearly had an advanced civilization. The rooms were built from a titanium alloy and were well lit. The walls were strewn with bright cry
Here is an unfinished project of mine to hack a classic title into a two-player cooperative game. I've had the idea for a couple of years now, and wanted to keep it under wraps until I had a finished game to show off. Unfortunately it was easier said than done.The idea was to bankswitch between two nearly identical copies of a game and to flicker two independent player sprites. The attached demo shows the basic idea. In this demo, player 1 can move freely, and player 2 is limited to horizontal m
I don't know if this has been done before, but this diagram is the secret to my time-sensitive kernels. It shows every clock and every cycle of the scanline. No more adding up cycles. If you know what your kernel needs to do at a particular point, the code almost writes itself!I've attached a blank diagram, and an example filled out for the main part of Marble Jumper. I'll take suggestions for improvement, and if it seems useful to enough people, it could go in the MiniDig.stak.zip stak.zip
EDIT 09/29/05: The ideas and mockup in this post have been made obsolete by recent developments. Supercat has produced a new demo, and I have designed an alternate board. See this thread for details.Many people have thought about how to port Archon to the VCS, but they worked with the classic design of three colors for squares and two colors for sprites. That is too many colors for the VCS unless you settle for flickering or Venetian blinds.With a little flexibility though, a good board is possi
As a treat for those of you who read my blog, here is a sneak preview of my 1K puzzle game. I've tripled the number of puzzles from the first version posted, and added a musical reward for winning. (thanks to Paul and Wolfgang ) It wasn't easy keeping the scanline count stable when resetting and selecting, while still keeping under 1K.A reminder of the rules: You play by jumping marbles over each other, like in checkers, except the jumps are only horizontal or vertical. You win when the last pie
If you've been browsing the homebrew forum, you have probably seen the Marble Solitaire game I've been putting together for the 2005 Minigame Competition. http://www.atariage.com/forums/index.php?showtopic=75152Since the original annoucement, I have been able to include some basic sound effects, and more than doubled the number of puzzles. Currently, I am looking for more puzzles to include in the final version. They are not hard to find, indeed the problem is actually narrowing them down. I've
Work on my board-game homebrew is going well. The current demo is a fully playable two player version where both players share one joystick. The controls are simple: push left and right to select a column, and push down to drop your piece. Use reset to clear the board. The rules are simple too. The winner is the first player to get four in a row horizontally, vertically, or diagonally.The software automatically ends the game if the four pieces in a row are vertical. Since this is a work in progr
I worked out the AI for Vertical Four-In-A-Row and put together a DOS program with a text interface that plays a pretty good game. The next step will be to implement the AI in 6502 assembly. Then it's on to finishing the kernel (see previous entry) and making an interface.You take turns with the computer dropping pieces into an upright board, and object of the game is to be the first to get four of your pieces in a row horizontally, vertically, or diagonally. As far as I know, this might become
.You guessed it! Congo is a board game. This chess-like game sets you up with a menagerie of African animals. Your pieces are a giraffe, a monkey, two elephants on either side of a lion, a crocodile, a zebra, and a row of 7 pawns. Each piece moves according to its own rules, and the object of the game is to capture the opponent's lion. For more information, check out chessvariants.org.The mockup above is definitely doable. An asymmetrical playfield with 9 color changes already exists in my Go k
.I have been working on a board game in which players drop pieces into an upright 7x6 board. The first to get four of a kind horizontally, vertically, or diagonally is the winner. Milton Bradley markets a similar game under the name Connect Four.The first picture is an actual screenshot of a single row, and the next part is a mockup of what the final kernel should look like. Making one row was the most difficult part, and repeating it will of course be simple. With any luck, I will be able to ma
Welcome to my Blog!What is it about the Atari 2600 that motivates so many to want to create their own games? Like a large number of AtariAge members, I have more projects in mind than actual time do to them. I have been programming for the Atari VCS since 2003, and am the creator of Combat Redux, the 2004 Stan Awards, and the chess hack Knightmate. Continuing projects include Tank AI, a 9x9 Go game, and Combat with an asymmetrical maze.It does not make sense to start a new thread every time I g
Like us on Facebook
Follow us on Twitter
AtariAge on Instagram
Watch us on YouTube
Chat on Discord