# pouët.net

## Using Blitter drawing polygons, to perform binary algebra on pixels in bitplanes (another nail to my coffin???).

category: code [glöplog]

Yo! Hatefull peeps!

Since we figured out how to set minterms, octants and how PROPERLY configure BLTAPT for SING lines. We drew a polygon.

Polygon was replacing existing data in bitplane.

If we, however define, A as a source with polygon (1bitplane), B as a overflow and C as a background. We can, by using minterms \$96, sum AB and C to get binary result in D. By uising minterm \$e8, we can get overflow of adding ABC, that should be carried on to the next bitplane. \$e8 blit has to be performed BEFORE \$96 blit. Overflow from the last bitplane returns back to first bitplane.

This way we realise binary algorythm of summing on pixels where the polygon is drawn. It means that all pixels are turned into counters and pixels where the polygon is drawn increase counter value by one.

Regards!!
added on the 2023-04-11 11:54:25 by Laffik
How long before he discovers glenz and goes on a rampage?
added on the 2023-04-11 11:59:40 by Preacher
sorry, I used to think that demoscene is to show skills of a coder in certain set of disciplines. it's not a point to download source from your pal. It is about your own approach to the subject of particular, classical effect. Is it not like this? I thin it's always been! So, my way of solving octants is fastest of published. If I copied it, like you persuade, it would take three times longer to find it.

How many more of you there with similar farts?

Daisy chain of sh&t-shovers, frankly!
added on the 2023-04-11 12:14:28 by Laffik
category: rant [glöplog]

Seriously, dude.. WTF?

Regards! :]
added on the 2023-04-11 12:56:29 by ham
Here's an example of a novel, outside-the-box approach to a classic problem - solving a maze by simulating gas particles.

Would you improve that by finding the fastest gas flow model possible, or would you improve that by simply using A*?
added on the 2023-04-11 12:58:47 by jobe
Recommended book: The Fart of Computer Programming. :D
added on the 2023-04-11 13:05:42 by ham
Quote:
to show skills of a coder in certain set of discipline

added on the 2023-04-11 13:20:18 by d0DgE
"bare metal platform"? wtf....
added on the 2023-04-11 14:40:36 by Laffik
well the question is: does it have unlimited detail?
added on the 2023-04-11 14:43:33 by prost
"unlimited detail"??

there's no such a thing as unlimited detail. it's reality.

who are you talking to???
added on the 2023-04-11 15:05:32 by Laffik
Quote:
there's no such a thing as unlimited detail.

Euclideon begs to differ.
added on the 2023-04-11 16:15:35 by Scali
reality! reality man! computers are real! limited pixels, limited tick per second, limited resorces. wakey-wakey!!! there's no unlimited detail.
added on the 2023-04-11 16:20:09 by Laffik
Are you calling Bruce Dell of Euclideon a liar?
added on the 2023-04-11 16:28:40 by Scali
you should know there are hypotheses, not proven to exist and abstracts - existing only in virtually as well as theories copying or simplifying the reality. there's plenty of fish. Point is that all computers have LIMITED resources, and will not present infinity of detail. as well as infinity of quantity. utopia. does it say anything (and it's not about mythical realm)? there were plenty of liars in publications adopted or adapted to science, so I wouldn't wonder if it was +1. same as fractals - you can draw it and you will never reach the detail in infinity of it.
added on the 2023-04-11 16:43:06 by Laffik
This remembers me of a Game of Life algorithm done via 64b (int) or 128b (SSE2) boolean ops to calculate the cell values doing add etc. on a bit level like the blitting above. On a Pentium IV I got a several billion cells calculated per second.

I think, a fast 68k working in fast ram could be more efficient here. It's a simple calculation of mem and ALU bandwidth. Who likes to might also do a roofline analysis.
@pro Holy crap, that brings back memories XD
added on the 2023-04-11 20:18:35 by d0DgE
This has got to be Amiga Adok.
added on the 2023-04-11 23:21:58 by thec
So... after the rain I am sticking one more word about subtracting bobs:

Binary function of subtracting gives blit D - \$96, like before and blit O (overflow, which is now underflow, loan or borrow) - \$8e. A - background, B - object, C - carried overflow or borrow from previous bitplane, D - background. Blit of overflow has to be done before blit of D, as D overwrites A.

For usual shade bobs it is casual that overflow or loan from the last position is carried on to the first bitplane, resetting the counters. To stop it and get saturation effect, function for the last bitplane has to be changed from adding to binary OR - \$fe, for blit D and unconditional 0 for blit O.

For subtraction minterms \$90 for blit D and \$80 for blit O, should do the saturation right.
added on the 2023-04-25 13:50:57 by Laffik
lol, the Blitter saga continues... :)
added on the 2023-04-25 15:26:37 by Defiance
To blit, or not to blit? that is the question.
Whether ’tis nobler in the mind than copper.
added on the 2023-04-25 22:44:57 by ok3anos