Who cares about Satuday programs? Corrupted Memory 0.4 and my XML parsing program are in the queue!

Because some people here actually made some good programs. And Monochromatic is just fantastic.

i know, I was kidding

But seriously, Mono* is awesome. I wish I had an 83+

...and Off 1.only for TI-89 is in the queue, too. These are three excellent programs, which is why they were featured.

Hint: Monochromatic belongs in THIS folder:

/pub/83/asm/graphics/ion/

I had just downloaded all these programs a few days ago...now I'll really have to try them!

Thank you Michael. I've never actually expected a green sticker, but perhaps it was the Ion style asm library I just added...WOOOO HOOOOO!!!! Party tonight and all the females get in free! Time to celebrate!!!

All the females get in free?!?! Can I come?????? (JK :-) ).

The Seklorean Warrior Shell looks really great, and I would definitely use it if it had support for standard YAS icons and YAS filetypes. I hate to be asking for more features on a shell that has so many already, but I really like editing TextView files and ZAC source code files on my calc with ATE. Also, I think some games (like blinded by the dark) use YAS filetypes to allow a saved game to be launched from the save file. Not essential, just a suggestion for if you're looking for more features to add...

Such ideas came to mind. As for the icons, it's based a bit off YAS as well. Not only does it support normal icons, but also 4 level greyscale ones too. ZAC and textview file I can easily add recognition for, but for editing them... the actual programs are best and i lack the skill to anyways (the whole checksum thing).

I didn't mean adding editing capabilities - just some code so that if ATE was on your calc, when you tried to run a TextView file, it would open ATE with the file loaded (like iShell...).

Every time I try to send it to my calc, it complains about a problem witha variable called 'MONOCHRround('.

Geez, just rename the variable.

I tried to test the 83+ version in VTI, but it kept on giving me a transmission error. Maybe ~22 KB is too big :-). I'll have to try it on my actual calc.

And this demo run perfectly on TI-82 with my "TI-83 emulator"! Good Job ;)!

Woo hoo! My multi-threading routines got featured! I never thought they would be.

For some reason some of the files still say version 0.9b, but it's really version 1.0. I thought I had fixed them. Just ignore them please. :)

Alright, so who has used or possibly would like to use my multi-threading in a program? Don't be shy! It could even be a C program; with an earlier version I wrote a working threaded Small C program with some C wrapper functions, so I know it can be done easily. I might include the wrapper functions if people want them.

I plan on revamping my threads sometime soon (changing the scheduler and thread management and adding timers etc.), but I'm more likely to do it sooner if people are using it.

Does anyone have any suggestions for making my threads better?

Questions? Comments? Money? :)

They sound useful for a multithreading shell like TSE...maybe I'll try something like that...

Oh yeah, BTW, congratulations to the other featured authors.

That Monochromatic 2 graphics demo looks pretty awesome! It's really good for a TI-83. I noticed the dithering patterns are different from a normal Bayer dither patterns. It's not bad or anything; it just gives the graphics a different "texture", which makes the demo more interesting. The patterns are even different from the first Monochromatic demo (which I think uses spiral dither patterns), and it has a different look because of that.

I'm just using pixel-doubling for the more "intensive" effects in mine - so the dither patterns are restricted to 100% (black), 50% (grey) and 0% (white). I can't think of a sensible way to "up" the palette, especially as the manner in which I pixel double vertically is a mucky hack anyhow; I'll have to see how much it slows if I add a 25% and a 75%.

In your what? Do you have a demo like Monochromatic? I looked around your site and your profile a bit, and I see you're into images and animations, but I don't think I saw anything which uses only 0%, 50%, and 100% dithering.

I wrote a Bayer dithering routine which returns whether a pixel is on or off depending on its intensity (0-255, 0 being black) and position on the screen. Would this work for you?

It works similarly to a FindPixel routine, being highly optimized like one, but like a FindPixel routine, it could be used in-line in a program so it's much faster. For one thing, you could maintain a pixel mask (which is used by the dither routine) and a row counter (just Y modulo 4) instead of computing them for each pixel. This can save a lot of time.

Alternatively, you could reduce the resolution of the screen by 1/4 in each direction (4 is the width and height of the dither pattern). I think the spirally/blobby graphics in the Monochromatic demo used 4x4 "pixels". 24*16 pixels can be computed much quicker than 96*64 pixels.

I'm the coder of Monochromatic 1 and 2.

You are correct that the dither-effects in the demos are done using 4x4 "tiles" of dither-patterns.

Nice that you spotted my new dither-algo! I'm not using any known algo, I just make it up as I go along. :)

My dither algo runs along the lines of 'if it's between x and y, use tile 1, if it's between y and z, use tile 2' and so on :(

I did a little experimenting with a grayscale image I made which is similar to one of the effects in your program. I made two versions: one is full resolution, and the other is lower resolution (with 4x4-pixel blocks). I converted both to black and white using Bayer dithering, and both images were almost identical! The low-res image looked slightly "fuzzier", but it wasn't very noticeable.

I guess fixed-pattern dithering obscures high-frequency (small) details anyway, so that's why a low-res image looks almost the same as the high-res image when it's dithered.

Not to steal your spotlight, I might try to write a graphics demo of either interfering waves which produces a Moire effect, or plasma. Each will use 4x4 blocks to reduce computation time by 1/16th.

I've done a monochrome (non dithered) moire which overlays (XORs) three concentric-circle patterns on top of eachother. OK, the pattern is a 192x192 bitmap that is shifted, (generated at runtime), but it runs very fast at full resolution...

Mine will calculate the brightness of points using the cosine of the distance from the center of a circle, so it might be a *little* slower than using XOR. :)

I'm currently writing a graphics demo with some old school effects in it... I have used 4x4 dithers before, if you notice in FT2 (flame effect).
Re: Your routine - I currently start at 0,0 then shift the bitmask right every pixel (for pixel doubling I start with a mask of %1100000, and shift twice every pixel) and wait for that to overflow, once it does I reset it and increment the pointer to PlotsScreen... it works nice and fast! If you could email me your routine, or point me to it, I'd be interested... *goes and looks at your profile*

Hey, if you use a circular rotate on your mask instead of shift then you don't have to reset it when it overflows. Rotates set the carry flag on overflow, so you can still test for it as with the shift.

My Bayer routine is in the file called bayer.zip (you never would've guessed :). You can browse my profile or click on my link above.

Oh, Bayer, is that what you used in your sunlight program? A patterened dither? I'll have a look into it! Thanks :)

Yes, Bayer is a patterned dither. It's quite easy to implement and also fast.

My sunlight program spends most of its time calculating the illumination, so I can't really tell how fast my Bayer routine is from that program, though. I estimate it could fill the screen (128x64) pixel-by-pixel in about 1/6 of a second.

I kinda like your fire effect more than Mono's.

Cool :)
I need to find a decent reference on Bresenham - all my circles are coming out wonkily :(

I found some circle-drawing routines here. Try these (all are under www.ticalc.org/pub/):

circle.zip (86/asm/source/routines)
circle.asm (85/asm/source/routines)
fastcirc_8x.zip (83plus/asm/source/routines)

I don't know how fast they all are. The first two are for the 85 or 86, but you could probably convert them for the TI-83(+) pretty easily. The first one draws filled circles.