Chess Game Graphics

[+MrFish]

Getting this out from the other topic and continuing on...

 

A few more changes here -- mainly the blue lettering and numerals.

 

 

[+MrFish]

Here's my first draft of a hi-res version of the same design.

 

I'm tempted to say I prefer the hi-res version for the definition, but it's nice having 6 colors (and more available for board decoration, which I've already tested) for the lower-res version compared to 1.5 + 1.5 colors the hi-res. Maybe, all things considered, it's pretty equal. What do you think?

 

 

Edited August 1, 2017 by MrFish

[Justin Payne]

I like em both.

Edited August 1, 2017 by Justin Payne

[+MrFish]

This one uses a dither for the black squares. The main benefit here is that there is more contrast between the black squares and the black pieces themselves compared to using a solid color for the black squares. The tradeoff is you lose some definition in the area between each piece's border and the dither.

 

The other benefit of using a dither is to give the appearance of an additional color not available when using black and white alone (as with the GUI, for instance).

 

 

 

 

Edited August 2, 2017 by MrFish

[Stormtrooper of Death]

MrFish, do you also have programmed the Chess logic already ? I am very interested in a new chess program for the Atari 8bit computers.

[+MrFish]

MrFish, do you also have programmed the Chess logic already ? I am very interested in a new chess program for the Atari 8bit computers.

 

No, not yet; I'm mainly interested in seeing what is possible graphically, right now.

 

However, I have looked into some of the existing algorhythmic strategies (of which there are many) and also what features have been used for chess games on the Atari up until now.

Edited August 2, 2017 by MrFish

[+MrFish]

Refined the notation graphics and added a colored border:

 

 

 

Edited August 3, 2017 by MrFish

[snicklin]

Don't forget this....

 

Writing Atari Strategy Games from John White

http://www.atarimania.com/documents/Writing-Strategy-Games-On-Your-Atari-Computer.pdf

[+MrFish]

Thanks for the heads up. I hadn't seen that book before (although it was already in my collection).

 

I see already the author concedes that a good chess program can only be written by an experienced machine coder -- which he isn't. I have the feeling a well optimized version could be written in Action, DMSC's Fast BASIC, or Mad Pascal to play a good game, though.

Edited August 3, 2017 by MrFish

[ricortes]

You've probably seen it so I am just making sure. There is Microchess for the KIM, early 6502 SBC, that has been documented by the author. http://www.benlo.com/microchess/

 

It has been hacked to work on an Arduino running a 6502 emulator which while irrelevant, I found interesting.

Edited August 3, 2017 by ricortes

[+MrFish]

You've probably seen it so I am just making sure. There is Microchess for the KIM, early 6502 SBC, that has been documented by the author. http://www.benlo.com/microchess/

 

It has been hacked to work on an Arduino running a 6502 emulator which while irrelevant, I found interesting.

 

Thanks, I had seen it already. I haven't played it but I expect it's quite primitive at < 1K.

 

There's a nice Chess Programming Wiki; they have a page on MicroChess there: Chess Programming - MicroChess

Edited August 3, 2017 by MrFish

[+MrFish]

Larger board squares, so less crowding of the pieces, smaller coordinate markers to help, and section dividers on the info panels.

 

 

 

Edited August 4, 2017 by MrFish

[+S.D.W.]

This one looks great!

[+mytek]

Yep I like that last one the best.

 

- Michael

[snicklin]

They all look really, really good but I think that I will go with the last one, though only by a small amount.

 

I just hope though that all the CPU time isn't taken up by DLI's. That CPU time is critical for a chess game.

[Rybags]

I don't think any DLIs would be needed - that could probably all be done with PM overlays.

Of course the problem in all cases is the DMA load would be fairly heavy. Actually, using stuffed GRAFn registers instead of PM DMA and having DLIs do that work would probably save cycles.

 

For such a game, maybe some sort of secondary display methodology could be used during "thinking" to give the CPU that extra 20-25% worth of cycles. Maybe a narrow mode window of reduced size that scrolls around the playfield. Or if bitmap is in use, maybe an alternate DList that only shows every second scanline.

[JamesD]

I know the last version probably looks the neatest, but if I were staring at a chess board for hours, I'm not sure I'd want the board to be blue.
Maybe swap colors between the stats and the board?

[+MrFish]

They all look really, really good but I think that I will go with the last one, though only by a small amount.

 

I just hope though that all the CPU time isn't taken up by DLI's. That CPU time is critical for a chess game.

 

Thanks.

 

There is only a single DLI needed for the ones with colored borders (the last one included). No DLI's necessary for any of the other hi-res layouts, just PM's, as Rybags states.

 

The low-res version would need DLI's for all the character sets (Antic 4). So the hi-res version is much better for that reason, but I think the hi-res looks better overall anyway.

 

 

I don't think any DLIs would be needed - that could probably all be done with PM overlays.

Of course the problem in all cases is the DMA load would be fairly heavy. Actually, using stuffed GRAFn registers instead of PM DMA and having DLIs do that work would probably save cycles.

 

I'm actually just seeing what PM's look like. I think all black and white (or whatever single color) is fine.

 

 

For such a game, maybe some sort of secondary display methodology could be used during "thinking" to give the CPU that extra 20-25% worth of cycles. Maybe a narrow mode window of reduced size that scrolls around the playfield. Or if bitmap is in use, maybe an alternate DList that only shows every second scanline.

 

Maybe not a bad idea...

 

 

I know the last version probably looks the neatest, but if I were staring at a chess board for hours, I'm not sure I'd want the board to be blue.

Maybe swap colors between the stats and the board?

 

It's all just experimentation. I'd make it configurable no matter what anyway. I can't say I'm in love with any color for the chessboard either (and I've tried them all).

[devwebcl]

I like more the first version

[invisible kid]

As you already stated, in the dithered black square ones it's easier to quickly tell the difference between the black and white pieces, where on the others it's harder to differentiate due to the pure inverse thing, but they do look sharper. Either way, looking real good!

Edited August 5, 2017 by invisible kid

[Rybags]

I think with hires the overlay makes it look better, agreed probably B&W for the board though, or just have it in user preferences.

 

Isn't Atari's chess cart only 4K? Possibly it could be reverse-engineerd, then redo the graphics. Or maybe find the "best" engine for 6502 based machines and use that.

[+MrFish]

As you already stated, in the dithered black square ones it's easier to quickly tell the difference between the black and white pieces, where on the others it's harder to differentiate due to the pure inverse thing, but they do look sharper. Either way, looking real good!

 

Thanks. You're touching on what is probably the most important part of making a hi-res display for a chess set and board on the Atari look the most clear. The subject is actually a little more wordy than I want to get right now, but suffice it to say that there are more ways than one to skin a cat. If you look at the difference between how I designed the first set of hi-res miniatures for the notation and how I designed the later miniatures, then you'll get a clue as to what I'm talking about.

Edited August 5, 2017 by MrFish

[+MrFish]

I like more the first version

 

I still like it too. I plan to see if I can expand the board squares without having to remove the coordinate markers. I may have to redesign the pieces themselves in order to accomodate.

 

I did do a version without coordinates, which was posted in the DOS Applications thread where I started this topic. I haven't updated the panel graphics for that one yet though.

 

Meanwhile, here's an updated version for the smaller board that uses coordinates.

 

Edited August 5, 2017 by MrFish

[+MrFish]

I think with hires the overlay makes it look better, agreed probably B&W for the board though, or just have it in user preferences.

 

As I mentioned in my post to the invisible kid, there are other factors at play as well. So experimenting with redesigning some of the inversions will likely present some improvements.

 

 

Isn't Atari's chess cart only 4K? Possibly it could be reverse-engineerd, then redo the graphics. Or maybe find the "best" engine for 6502 based machines and use that.

 

The Computer Chess cart is 8K.

 

I think the best ones on the Atari are Sargon, Colossus, and ChessMaster.

 

I like the idea of at least using a decent opening book, for the higher level settings.

Edited August 5, 2017 by MrFish

[devwebcl]

I think with hires the overlay makes it look better, agreed probably B&W for the board though, or just have it in user preferences.

 

Isn't Atari's chess cart only 4K? Possibly it could be reverse-engineerd, then redo the graphics. Or maybe find the "best" engine for 6502 based machines and use that.

 

Maybe you can start with the simple First Chess

https://manillismo.blogspot.cl/2016/01/first-chess-cc65.html

 

 

 

As I mentioned in my post to the invisible kid, there are other factors at play in as well. So experimenting with redesigning some of the inversions will likely present some improvements.

 

 

 

The Computer Chess cart is 8K.

 

I think the best ones on the Atari are Sargon, Colossus, and ChessMaster.

 

I like the idea of at least using a decent opening book, for the higher level settings.

 

Colossus Chess is far the best player (not the gfx)

https://manillismo.blogspot.cl/search/label/ajedrez