Skip to content

cache: tessellation segment count varies with --size, so changing output size is a full cold re-eval #828

Description

@ecto

Rendering rose-pro-skinned at --size 800 after a warm run at 700 was a cold cache (~10 min): segment count is derived from output size and is part of the root-mesh cache key. Also: STEP export is always cold because only meshes are cached, never the BRep.

Ask: quantize segment count to 2-3 tiers (or cache the finest tessellation and decimate down), and consider serializing Solids so vcad export STEP goes warm too. Numbers: skinned body 253 roots, ~10 min cold, 0.8 s warm at the same size.

🤖 Generated with Claude Code

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions