Chess

[Mr SQL]

Very interesting! The display looks equally gorgeous on my CRT but I can't see the thicker lines, just a shimmer from the lower Hz.

 

 

[+Andrew Davie]

4 hours ago, Nathan Strum said:

Well, it would make for an interesting mental challenge having to keep track of your pieces like that.  

I suppose I could make a PAL-only version where I deliberately build malformed TV frames to disable the PAL colour circuitry/decoding.

 

[+Andrew Davie]

I have managed to make a PAL version that does interlacing and displays correctly on my TV - no colour artifacts/errors.

I'm curious to see if this works on other PAL TVs.  Use the right difficulty switch to change back and forth between interlaced/non-interlaced.

If you have a PAL setup, could you please test and indicate results - does the interlaced display look OK for you?
 

 

chess3E+20200709PALi.bin

[+Al_Nafuur]

31 minutes ago, Andrew Davie said:

I have managed to make a PAL version that does interlacing and displays correctly on my TV - no colour artifacts/errors.

I'm curious to see if this works on other PAL TVs.  Use the right difficulty switch to change back and forth between interlaced/non-interlaced.

If you have a PAL setup, could you please test and indicate results - does the interlaced display look OK for you?

Tested on PAL 2600jr with PlusCart:

 

[+Al_Nafuur]

The interlaced version works on my PAL TV. It is flickering not much but noticeable. Compared to the non-interlaced version the lines seem to be (slightly) thicker.

[+Andrew Davie]

2 hours ago, Al_Nafuur said:

The interlaced version works on my PAL TV. It is flickering not much but noticeable. Compared to the non-interlaced version the lines seem to be (slightly) thicker.

Many thanks!

I guess 2/2 isn't bad. Will wait to see what other results come in before calling it a success.

 

[+Andrew Davie]

The attached binary is now switchable between NTSC and PAL TV formats, and also switchable between interlaced/non-interlaced display.

Use the P1 difficulty (LEFT) switch to change TV format between NTSC and PAL. If you see squares any colour than blue, you have the wrong position. The switch is only detected on initial power-up - you cannot change mid-game. I'm reserving that for an easter-egg   The P2 difficulty (RIGHT) switch changes between interlaced and non-interlaced display. This appears to be working for both NTSC and PAL - at least on my TV. I guess this makes Chess3E+ the very first fully functional dual TV system dual display format cartridge, so it's already a collectable

After every move you will see "CHECK" briefly displayed all over the screen.  That's just part of a new system I'm too lazy to work on right now.

Anyway, have at it. If you have a TV this doesn't work on, please let me know. I think it's highly likely.
 

 

chess3E+20200711NTSCPALi.bin

[+Andrew Davie]

I don't really know if I'm generating an interlaced signal. All I know is that stella's scanline count is constant, and the TV alternates between the slight shimmer + scanline displacement when I toggle it on/off.  It looks like what I think interlaced should look like. If it looks like a duck, swims like a duck, and quacks like a duck, it's probably a duck.

 

[+Pat Brady]

7 minutes ago, Andrew Davie said:

If it looks like a duck, swims like a duck, and quacks like a duck, it's probably a duck.

 

But it might be a dragon.

[+fdr4prez]

42 minutes ago, Andrew Davie said:

If it looks like a duck, swims like a duck, and quacks like a duck, it's probably a duck.

[+Andrew Davie]

Just playing.... bit of fun... looks OK actually!
Let's call it "pride chess".
 

 

Edited July 11, 2020 by Andrew Davie

[+Andrew Davie]

Fiddling while Rome burns.

Bit of an experiment with non-black background ($F0)

Placing an image here for posterity.

... and $A0

 

Edited July 12, 2020 by Andrew Davie

[Keatah]

The Rainboard scheme looks nice. Would be even better if the colors could be stopped on demand..

 

F0 - blech looks like an aging LCD with some sort of copper dissolution or something.

A0 - blech too not enough contrast..

 

[+Andrew Davie]

3 hours ago, Keatah said:

The Rainboard scheme looks nice. Would be even better if the colors could be stopped on demand..

That's exactly what I called it in my code! Have you been peeking?
 

    IFCONST RAINBOW
    DEF RainBoard

                    clc
                    lda base
                    adc #4
                    sta base

                    and #$F0
                    beq RainBoard
                    
                    ora #8

                
.ROWBANK SET [SLOT2]
    REPEAT 8
                    ldy #.ROWBANK
                    sty SET_BANK_RAM

                    sta@RAM SMCOL1+1

                    clc
                    adc #16

.ROWBANK SET .ROWBANK + 1
    REPEND

                    rts
    ENDIF

 

[Keatah]

No. But I did watch the Rainboard.mp4 vid ?

It only makes sense that that's the name.

 

Edited July 12, 2020 by Keatah

[+Andrew Davie]

1 hour ago, Keatah said:

No. But I did watch the Rainboard.mp4 vid ?

It only makes sense that that's the name.

 

Ah, I didn't know you could see the name of the video. Doesn't show for me.

 

[Keatah]

6 hours ago, Andrew Davie said:

Ah, I didn't know you could see the name of the video. Doesn't show for me.

Have my browser settings show the file instead of a preview/play.

[+Karl G]

So - how soon do we get to see a version with swearing? 

[+Nathan Strum]

16 hours ago, Karl G said:

So - how soon do we get to see a version with swearing? 

How do you know Andrew isn't swearing in every version?  

[+Andrew Davie]

On 7/15/2020 at 11:43 AM, Karl G said:

So - how soon do we get to see a version with swearing? 

I do my own swearing at the moment.
I just spent the last 3 years of my life finding the "castle bug". That's where the computer steadfastly refused to castle, despite me offering huge bonuses and incentives to do so. 49 virgins, and all that. No dice. It took me, as I said, three years without sleep to find it, but find it I did - and this version is now quite happy to castle when the situation suits. The bug was extremely obscure - but basically castle moves are really two moves in a row by the same side - first the king moves, then the rook moves. Well, what was happening was that the king moved, and then when the rook moved a temporary variable holding the piece type was not set to "rook" but instead remained as "king". So the move for the rook came along, it looked at the piece type and decided "hey I'm not a king anymore" - so it removed the value of a king from the evaluation, and added in the value of a rook. Now this is usually active for pawns promoting - take away the old piece, put in the new piece. But in this one situation that code was kicking in because the "old type" was incorrectly set. Took forever to find. Now I can get some sleep.

The video shows a nice game including castling where, at the end I just couldn't be bothered finding a good move and effectively conceded. Many of the moves by the computer seem reasonable/plausible, which is great considering this is only 3-ply (Q4).  

Binary attached - use left difficulty to switch NTSC/PAL (so if the colours look screwy, flip this and power OFF/ON).  Right difficulty switches interlaced mode - if the picture jiggles up and down ever so slightly and you don't like that, then flip right difficulty switch (effective immediately). Also don't forget you can hit SELECT (at any time it is your move) to swap sides. And finally, joystick button when the computer is thinking will either give you an epileptic fit, or show a pretty pattern, depending on the user.
 

This binary will still have that "testing to see if you're paying attention" en-passant "feature".  But that's not going to bother most people (i.e., 3 of the 4 of you who are actually following this). For the odd one out, you'll just have to stop moving pawns.

 

 

chess3E+20200716b.bin

Edited July 16, 2020 by Andrew Davie
Goddammit, Q is on a black square again. I swear to the 49 virgins, this keeps flipping randomly! Binary updated and now correct, even if the video is wonky.

[+Al_Nafuur]

1 hour ago, Andrew Davie said:

Binary attached - use left difficulty to switch NTSC/PAL (so if the colours look screwy, flip this and power OFF/ON).  Right difficulty switches interlaced mode - if the picture jiggles up and down ever so slightly and you don't like that, then flip right difficulty switch (effective immediately). Also don't forget you can hit SELECT (at any time it is your move) to swap sides. And finally, joystick button when the computer is thinking will either give you an epileptic fit, or show a pretty pattern, depending on the user.

This binary will still have that "testing to see if you're paying attention" en-passant "feature".  But that's not going to bother most people (i.e., 3 of the 4 of you who are actually following this). For the odd one out, you'll just have to stop moving pawns.

 

I gave that version a chance to show it's "features", but this time no creative chess rule interpretations showed up.

 

I prefer the non-interlaced version it is clear and calm. The lines of the interlaced version are slightly thicker, but the noticeable flickering irritates me a bit.

 

 

Edited July 16, 2020 by Al_Nafuur

[+Andrew Davie]

5 hours ago, Al_Nafuur said:

I prefer the non-interlaced version it is clear and calm. The lines of the interlaced version are slightly thicker, but the noticeable flickering irritates me a bit.

 

Thanks for the feedback. I tend to agree - but I'm going to leave the interlacing in as an option basically because I think it's a bit of a first and... why not.

[+Al_Nafuur]

33 minutes ago, Andrew Davie said:

 

Thanks for the feedback. I tend to agree - but I'm going to leave the interlacing in as an option basically because I think it's a bit of a first and... why not.

I think playfield-interlacing might be much better for games with more "action" like Sokoboo or BoulderDash..

[+Andrew Davie]

Here's an experimental version with a taller board (more square squares).

Pics for comparison...  new board on the left.  It's quite subtle - the board pieces are slightly better spaced, I think.

 

 

This increases the scanlines to 282 for NTSC, which is probably pushing things.

Let's see how it goes.

chess3E+20200716wide.bin

[+Andrew Davie]

I found and fixed the en passant bug, that allowed en-passant long after the victim pawn moved off the home rank. It was a flag clear that some optimisation code from long ago bypassed.

 

Quote

The real problem is that programmers have spent far too much time worrying about efficiency in the wrong places and at the wrong times; premature optimisation is the root of all evil (or at least most of it) in programming.

-- Donald Knuth

On a separate issue, I've been struggling to get the AtariVox to do any meaningful talking.

I have plain vanilla non-bankswitching code  - just a few hundred bytes - that talks as expected. But when I transplant that code directly into a Chess3E+ bank, the AtariVox just doesn't work at all. I step through it with the debugger, it all looks OK. But the AtriVox doesn't talk. I don't understand it. It's almost as like there's a hardware bug in the 3E+... I'm that paranoid.
 

chess3E+20200717ep.bin

Edited July 17, 2020 by Andrew Davie