feat(cam): headless CPU rendering-platform mockup (CAM/viz)

patman commented

2026-07-10 00:12:39 +00:00

Owner

New cam-viz crate — the CAM rendering platform shell as a dependency-free,
headless CPU renderer. Rasterizes a representative CAM viewer frame (isometric
machine bed, serpentine pocket toolpath with in-progress cut cursor, X/Y/Z axes,
parameters panel, status chrome) and encodes it with a pure-std PNG encoder. No
GPU, no window, no third-party deps. Mockup geometry only — NOT the CAM path
core. Toolpath positions stay in typed LengthVec3 until the projection seam.

Adversarially reviewed (REWORK -> addressed): usize size math in the PNG encoder,
NaN-safe polygon sort, dead API removed, pixels() accessor over a pub field,
backend-seam doc claim toned down.

Note: ~840 lines (font.rs is bitmap data); larger than the <200 stacked-diff
target — cohesive mockup, splittable into png / canvas+font / scene on request.

Stacks on #43.

Co-authored-by: patman-assist patrick-ai@kgroo.co

New cam-viz crate — the CAM rendering platform shell as a dependency-free, headless CPU renderer. Rasterizes a representative CAM viewer frame (isometric machine bed, serpentine pocket toolpath with in-progress cut cursor, X/Y/Z axes, parameters panel, status chrome) and encodes it with a pure-std PNG encoder. No GPU, no window, no third-party deps. Mockup geometry only — NOT the CAM path core. Toolpath positions stay in typed LengthVec3 until the projection seam. Adversarially reviewed (REWORK -> addressed): usize size math in the PNG encoder, NaN-safe polygon sort, dead API removed, pixels() accessor over a pub field, backend-seam doc claim toned down. Note: ~840 lines (font.rs is bitmap data); larger than the <200 stacked-diff target — cohesive mockup, splittable into png / canvas+font / scene on request. Stacks on #43. Co-authored-by: patman-assist <patrick-ai@kgroo.co>

patman added 1 commit

2026-07-10 00:12:39 +00:00

feat(cam): headless CPU rendering-platform mockup (CAM/viz)

CI / build-test (pull_request) Has been cancelled

Details

fbc49a5f11

New cam-viz crate — the CAM rendering platform shell as a dependency-free,
headless CPU renderer. Rasterizes a representative CAM viewer frame (isometric
machine bed, serpentine pocket toolpath with in-progress cut cursor, X/Y/Z axes,
parameters panel, status chrome) and encodes it with a pure-std PNG encoder. No
GPU, no window, no third-party deps. Mockup geometry only — NOT the CAM path
core. Toolpath positions stay in typed LengthVec3 until the projection seam.

Adversarially reviewed (REWORK -> addressed): usize size math in the PNG encoder,
NaN-safe polygon sort, dead API removed, pixels() accessor over a pub field,
backend-seam doc claim toned down.

Note: ~840 lines (font.rs is bitmap data); larger than the <200 stacked-diff
target — cohesive mockup, splittable into png / canvas+font / scene on request.

Stacks on #43.

Co-authored-by: patman-assist <patrick-ai@kgroo.co>

patman-assist added the

AI Assisted

CAM

Feature

labels

2026-07-10 00:12:58 +00:00

patman merged commit cd92a55051 into main

2026-07-10 16:37:27 +00:00

patman referenced this pull request from a commit

2026-07-10 16:37:28 +00:00

Merge pull request 'feat(cam): headless CPU rendering-platform mockup (CAM/viz)' (#67) from patman/cam-viz into main

Rows
Columns

feat(cam): headless CPU rendering-platform mockup (CAM/viz) #67