Search
13 results for “grz0zrg”
-
Made a #forth ARMv2 (macro) assembler (tested with #gforth) and took the opportunity to revamp my light #ARM Forth a bit by adding useful words with some light changes :
https://github.com/grz0zrg/ARM-ForthLite/blob/main/examples/ARMv2_assembler.fs
The goal was to have one mnemonic = one word so 90% of the source is a lot of repetition but it looks close to the conventional syntax. (In a postfix way)
Does not require much primitives.
Wrote a small write-up : https://www.onirom.fr/wiki/blog/20-12-2025_writing_an_armv2_assembler_in_forth/
-
Added short form syntax (shift is now optional) and ADR pseudo instruction (at last) to my #arm #forth assembler, mostly convenience stuff to avoid verbosity / manual compute of PC relative address, also changed labels word to -> and <- (--> for ADR labels), ADR implementation use a single pass back-patching algorithm that build a linked list until --> word is hit (will resolve each ADR), could be done more generically as well (for branch) but will skip it for now. Time to code some RO stuff !
-
Added short form syntax (shift is now optional) and ADR pseudo instruction (at last) to my #arm #forth assembler, mostly convenience stuff to avoid verbosity / manual compute of PC relative address, also changed labels word to -> and <- (--> for ADR labels), ADR implementation use a single pass back-patching algorithm that build a linked list until --> word is hit (will resolve each ADR), could be done more generically as well (for branch) but will skip it for now. Time to code some RO stuff !
-
Added short form syntax (shift is now optional) and ADR pseudo instruction (at last) to my #arm #forth assembler, mostly convenience stuff to avoid verbosity / manual compute of PC relative address, also changed labels word to -> and <- (--> for ADR labels), ADR implementation use a single pass back-patching algorithm that build a linked list until --> word is hit (will resolve each ADR), could be done more generically as well (for branch) but will skip it for now. Time to code some RO stuff !
-
Added short form syntax (shift is now optional) and ADR pseudo instruction (at last) to my #arm #forth assembler, mostly convenience stuff to avoid verbosity / manual compute of PC relative address, also changed labels word to -> and <- (--> for ADR labels), ADR implementation use a single pass back-patching algorithm that build a linked list until --> word is hit (will resolve each ADR), could be done more generically as well (for branch) but will skip it for now. Time to code some RO stuff !
-
Some HAKMEM 149 based randomised #circlepacking algorithm i am tinkering on, the algorithm is used to draw the circles (and fill them) but also to select a random point inside the root circle by reusing the HAKMEM circle fill by messing with the orbit a bit (it spiral inward), this makes the code rather tiny !
Inspiration came from https://rantonse.org/art/randomized-circle-packing
-
Wrote some more about extending the "HAKMEM 149 line algorithm" to a quad rasterizer to be able to draw billboards (sprites) up to 3D sprites, the HAKMEM approach (and forward mapping) lose a bit of interest when degrees of freedom (and projection) are added up though but it was fun to retrace the sprites history a bit. :blobcatcomfysmirk:
https://www.onirom.fr/wiki/codegolf/main/#HAKMEM_149_quad_rasterizer_/_sprite_scaler_+_rotation
-
Experiment with #Cistercian numerals extended to base 16, rearranged so it map to binary, a #glyph depict 16-bit numbers with quadrants depicting nibbles (4-bit), quadrant order is untouched.
4 original Cistercian numbers are mapped on power of two :
0001 -> 1 (Cistercian representation)
0010 -> 2
0100 -> 4
1000 -> 6All nibble numbers are then defined by mixing those.
Packed all my tiny intros with these glyphs for fun, replacing stave by a dot makes cool glyphs. (and is less ambiguous)
-
A very simple way to draw HAKMEM 149 based structures by using whole 32 bits range with upper bits as coordinates using unsigned shifts.
This also works with a lower range but it loses precision.
The oblique version is even simpler by removing the last two operations.
The structure looks like Apollonian net or Hopalong attractors.
This drawing method also works with other attractors or Gingerbread man map etc.
#sizecoding #codegolfing #algorithm #attractor #fractal #computergraphics
-
A tiny way to compute a square / triangle wave at the same time without conditionals, only basic arithmetic and shifts.
This can be used for #oscillators or animations etc. #algorithm is a variant of HAKMEM 149.
Can also be used to draw a square. (And other shapes, see previous toot)
Code also shows how to draw a diamond shape by rotating the square by 45°.
There are two versions, optimized for size and optimized for speed.
#computergraphics #p5js #softwarerendering #codegolf #sizecoding
-
A try at realtime graphics programming / code golfing on the #AtariST #motorola68k, also wrote a guide (may works for other models such as Falcon) : https://www.onirom.fr/wiki/codegolf/atari_st/
My first :atari: ST program was made 20 years ago but i didn't go further than some GEM "hello world" at the time... a fun and easy platform overall although i dislike bitplane graphics and the exe header. ;) (no need for header / loader on Acorn Archimedes / RISC OS !)
-
A tiny way to draw an "isometric" (dimetric) cube in an iterative fashion with a single loop, all integers with an integer variant of HAKMEM 149 (so called Minsky circle) display hack. (Sin / Cos approximation)
Can also be used to draw a hexagon (it is a filled hexagon after all) and can easily render fractured cubes as well by varying "l" (second screenshot)
#cube #p5js #computergraphics #isometric #dimetric #hexagon #softwarerendering
-
Notes about this :
1. Most of the syntactic addition obviously serves as visual cues and the syntactic rules are NOT enforced, still okay for a small language with few users focus, there is great Forth-like high level language like #factor anyway for a more professional role
2. '}' serve as termination for 'if' 'for', may only work in some implementation where the scope is known, also didn't close the definition with ';' or '}' on two last ones, a mistake