Ray tracer in a boot sector | Hacker News Reader

Ray tracer in a boot sector

github.com

210 pointsblizdiddy21 comments

Open articleView on HN

If you like it, you definitely want to visit these two websites:

https://www.shadertoy.com/

https://www.pouet.net/

The last demo party was this one: https://www.pouet.net/party.php?which=1550&when=2024

And this is 64K: https://www.pouet.net/prod.php?which=96589

Something out of this world.

And this is 128 bytes: https://www.pouet.net/prod.php?which=63518

It appeared on HN too: https://news.ycombinator.com/item?id=7940212

Another masterpiece from Oscar Toledo. Context about him and the Toledo family: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Unbelievable!

I'd be very grateful for an explanation for how the first three lines in the Atari basic version that have only numbers and no alphabet work?

I just can't remember what Atari basic does, how that data is presented to be read into the array...

Thanks!

I think you're referring to just the column numbers (01, 02... 80) which aren't part of the program listing (?)

Thank you both!

Those first two lines are just column number indicators for easy reference... 01 to 80 - to show that it's 80 columns wide.

My first thought at seeing something pretty in so few bytes was: “perhaps you could just run through every value in each of those 484 bytes and find other gems in there too” then I did the math - or tried to… :)

Just a reminder that most of the modern cryptography uses keys that are at most 32 bytes.

Check out the BBC micro version in 432 chars (443 bytes). https://bbcmic.ro/?t=9ctpk ... That cuts your problem down by a factor of ~5.5e98. ;) Or even a lot more if you limit to readable ascii. I'd be curious if you did that how many of the tries would result in a valid runnable program.

You don't need to go anywhere near 400 bytes to find inscrutable programs, if your programming language is concise enough. Even 8 bytes, or 64 bits to be precise, offers plenty unchartered territory for finding new gems in lambda calculus [1], one such recent discovery being a 49 bit program whose output exceeds Graham's number.

[1] https://oeis.org/A333479

Honestly why bother with actual bytes? Why not just take Wolfram’s approach of looking through programs encoded as cellular automata?

From the oeis link: https://tromp.github.io/blog/2023/11/24/largest-number

Acutally, I only learned about the 49 bit program after writing that blog entry. It is described in https://github.com/tromp/AIT/blob/master/fast_growing_and_co...

> Thanks to Peter Ferrie for helping me to save 23 bytes

That's the spirit!

mov al,0x13 brings back memories :)

mode 13h in tasm while pretending to understand line algos takes me back

I love these little projects, and the efficiency of the resulting code. I wish I was skilled enough to turn it into a game or something, though.

The author has two books about writing small games that can run in the boot sector. First book also has a quick course in 8086 Assembly:

Programming Boot Sector Games

More Boot Sector Games

I have the authors Programming Atari 2600 Games and really enjoyed it. I started my professional programming career programming in Assembler on an OS/360 mainframe back in the mid 90s and it really helped me starting out at such a low level. It was enjoyable revisiting Assembler while going through his book.

Unreal codegolfing!

Reply on news.ycombinator.com