Going from "SVG/Canvas is wasteful" to a tile based approach is a giant leap. Wouldn't AI suggest delta encoding of polyline coordinates and discretization of deltas as small, maybe just 2-4 bit integers? Storing the X and Y delta vectors separately, rather than interleaved, could lead to a further increase to the compression ratio.<p>What was asked of Claude? The article is very sparse on this.
32x32 4bpp 16-color icon is 512 bytes. There definitely were programs that had a recognisable world map for an icon in those days, so the answer is yes.
Is there a good known algorithm which performs general purpose compression where the target is a given turing complete instruction set? Rather than relying on a fixed general purpose decoder and the associated compressed data.<p>I’m asking here instead of asking an LLM because that’s what humans used to do and it was pleasant.
A perfect implementation would be a Kolmogorov oracle. <a href="https://en.wikipedia.org/wiki/Kolmogorov_complexity#Halting_problem" rel="nofollow">https://en.wikipedia.org/wiki/Kolmogorov_complexity#Halting_...</a> suggests this is equivalent to a halting oracle. So, it depends what you mean by "good".
that sounds pretty related to Kolgomorov complexity, which is uncomputable in general. <a href="https://en.wikipedia.org/wiki/Kolmogorov_complexity" rel="nofollow">https://en.wikipedia.org/wiki/Kolmogorov_complexity</a><p>I too would be interested in approximations or heuristics if anyone has any
I didn't know you could do this with fetch() and a data: URI:<p><pre><code> fetch('data:;base64,1ZpLsgIxCEXnrM...==').then(
r => r.body.pipeThrough(new DecompressionStream('deflate-raw'))
).then(
s => new Response(s).text()
).then(
t => b.innerHTML = '<pre style=font-size:.65vw>' + t
)</code></pre>
Why is there a lake in Africa near Congo/Gabon istead of one near Uganda?<p>Also, can this be done with png? Most consecutive lines are very similar, so I'd expect the algebraic pass to be very useful.
Since the map just has two states - land or water - I wonder if there's a way to represent the same information more efficiently bits rather than bytes.
What's the point in doing things like this if you just get Claude to do it?
Experimenting with drawing a world map and micro-optimization.
Presumably you could precompute some parametric function (probably a Fourier sum) which draws a reasonably close map of the world, and get that into 500 bytes with a math-focused programming environment (R, Julia, etc.)? I might try throwing Fable at this and seeing what I can get.
Well, here's what Fable came up with in 499 bytes of R in about half an hour: <a href="https://pastebin.com/sBsiGD9t" rel="nofollow">https://pastebin.com/sBsiGD9t</a>, result: <a href="https://imgur.com/a/W3eDdIC" rel="nofollow">https://imgur.com/a/W3eDdIC</a>. Probably with sitting down and tweaking you could do even better, but I think this is a decent first start.
heck yea! 499 tera. no problem.