Joined February 2010
(Two days later) I understand now that this is a poorly written APL code.
1
47
That's old news now. I realized that I forgot to publish my reimplementation of MarkovJunior, which uses the Baker-Bird algorithm for 2-dimensional pattern matching. The classic Maze Backtracker implemented with two simple rules: "R.. => GGR" and "RGG => YYR".
1
1
1
68
For completeness: "R. => RR"
36
Unsure about what to call this. It is like tixy.land but follows a curve defined by L-system rules.
1
1
2
92
Some points may overlap, naturally.
1
30
Oof, this quality is terrible :facepalm: I should build my own GIF exporter.
28
A test interface for interacting with the model. I'm not sure why I keep going back to System 7.
2
63
A baby GPT model is learning to draw cats, cars, and parrots.
1
74
Hey @techthumb. This is Kirill. I know this is awkward, but I lost your contacts, and I'm really sorry about that. Hope you are doing great. Just wanted to drop a message (and it's hard to believe that I am doing this via Twitter) to say hi and see how you are doing.
1
93
I was reconstructing L-system rules for βΩ-Indexing, yet another obscure space-filling curve, and I made a mistake. This bug turned out to be visually pleasing but practically useless.
1
3
125
The reduction of the top-right quadrant of the 6th order gives this: B+F+C-F-E+F+Y+F+C-F-B+F+D-F-W -F- W-F-B+F+D-F-E+F+C-F-B+F+D-F-E There are extraneous +F+ -F-, which likely point to the bug, ie: C => BF-CFE-F-Y (the kernel) C' => B+F+C-F-E+F+Y (from the definition above)
1
48
Something is still off: -F- and +F+. They must alternate the direction and flip + -> - from one order to another: A => B+F+D-F-B+F+D B => C+F+B-F-D+F+W D => W+F+B-F-D+F+E C => B-F-C+F+E-F-Y E => Y-F-C+F+E-F-D Y => B-F-C+F+E-F-D W => C+F+B-F-D+F+E
33
Bootstrapping Forth for a project that has nothing to do with Forth, but to crunch numbers, and as one said, "when you have seen one Forth, you have seen one Forth."
1
56