ClausB

Thanks for testing! Yes, well, it could have been red originally. In 1983 I made it from memory (I didn't have YouTube then, not even VHS) so the shape isn't quite right either. I was lucky it was even orange, since I did it on a B&W TV!

Had some trouble after sync'ing to VCOUNT to wait for the display list to finish before clearing the bitmap for the next frame. Discovered that some frames take a bit too long for 15 fps. Was getting 14.5 on average. Had to resort to a look-up table for the *40 routine (instead of the optimizations credited above). Also had to disable the deferred VBIR with a SEI. Now it averages 15 fps, though a few frames are long and a few are short. Tested in emulation only. Please test on real hardware and report back. Runs in 40K RAM: TRON2.XEX Edit: Don't forget to press START after it loads and the screen goes black.

I meant the ST. 😛 Very hard to see.

Ah, yes, the famous wood varnish model.

Hmmm, I struggle to see the usefulness of 1 millibit. Did you mean 1 MB?

Working on my TRON Recognizer animation from 1983. Originally (below) it rendered 72 frames of GR.22 bitmaps, 2K each, and played them back at 15 fps on an 800 equipped with the Axlon 128K RAMdisk. Using my ported SubLogic FS1 line plot routines, now it can draw and erase each GR.23 frame in less than 1/15 second, so no need for so much RAM. The raw speed is measured at 1152 lines per second! Of course, most are short lines, and a few of the 59 lines per frame are clipped away, and the CPU runs faster in GR.23 than in GR.24. The original BASIC program rendered the animation over an hour or so and saved 144K of bitmaps to 2 SD floppies. An assembly program played it back on the Axlon at 15 fps with sound effects. I modified the BASIC program to save and clip line endpoint coordinates instead, only 17K. An assembly program feeds those coordinates to the FS1 plotter. A side-effect of using GR.23 is more colors, so I took advantage (above). Some work remains, such as double-buffering and sound. Thanks to @tebe and @Rybags for inspiring some optimizations.

Great idea! I think the Atari 8-bits' potential was stunted by the cheap c64's success, so it's good to see a mass market homage to the 400. My first Atari was an 8K 400 in 1980. Loved it!

112 means 11th week of 1982. In another thread you said you got your first 400 in 1979. Can you post that label?

I'll take a crack at it: I'll call the ring buttons N, S, E, W. The rear buttons L, R, B. Fire button F. L decreases throttle. R increases throttle. N toggles front and aft views. E cycles computer, tracking, none. S shows galactic chart. If stick has moved on chart, S again jumps to front view and hyperwarp. W toggles shields. B and N together shows long range scan. B and E together selects manual target. B and F together pauses.

OOOF! Don't do that! If you plug a 16K RAM board in there you will short S5 and RD5. At best your carts won't work. At worst you'll damage a motherboard chip!

OOOF! Don't do that! If you plug a 16K RAM board in there you will short S5 and RD5. At best your carts won't work. At worst you'll damage a motherboard chip!

Congrats! Did you post the label numbers in the "Calling all 800s" thread?

Star Raiders or go home!

Please post a photo of the underside label in this thread:

Bonus! Pre-yellowed, for those who don't want to wait 40 years!

My kid's old Android 2.2 tablet has Colleen and Droid800. Also the Zed Ex ZX81 emulator and a Sudoku game. Oh, and X-Plane 9. Still has decent battery life, so those apps keep it going.

Wait, some story missing? Did your family buy an A8?

I see your improvement to SCRNP. That was my code, not original FS1. Thanks for that. Looks like your test case is one diagonal line from 0,191 to 159,0 and you wait for VCOUNT=0, draw the line, and read VCOUNT again to measure speed, right? So, you get $49 which is 73 but VCOUNT increments every 2 scan lines so that's 146 scan lines, right? At 15,700 scan lines per second, we have 15,700/146= 108 lines per second, right? But that does not account for VBI service cycles lost, right? Or am I missing something?

I don't understand. The number of scan lines for how many lines? How long are the lines? What slope?

What are these numbers? Some kind of speed measurement? The added code in front of PLOTL appears unnecessary, since the original code checks direction and swaps with BCC L1C3E.

I first ran LINETEST drawing within a 140x96 area, which is what A2-FS1 uses, and got around 400 lines per second.

BTW those two examples of self-modifying code were present in FS1, but in different forms. 1. The addition or subtraction to the screen pointer was more complex for the A2, given its non-contiguous memory map. 2. The screen clear loop had hard-coded addresses but the loop branch condition was modified. Avoiding zero page indirect addressing saves about 7000 cycles.

This was 1980. Was reading BYTE about all the offerings. Wanted an Apple but oh, the price! Considered COSMAC ELF or VIP but not sold locally. Thought about TI-99. I had dismissed the Atari 800/400 as games machines but after seeing Star Raiders and learning programming details from the salesman, I was hooked. I later built a modem for it and upgraded it to 48K RAM and a used 810. Typing was OK since I wasn't a touch typist and I learned all the abbreviations. Two years later the 800 price came down and I upgraded. Eventually added 128K RAM and an ATR-8000. Two years later got an 800XL and upgraded it to 256K. Came full circle and published an article in BYTE!