Assessments

Measurements of what glTF assets contain, criterion by criterion.

Assets published by other projects are measured to learn what is in the field. They are not a basis for the profile’s rules: a field asset is as likely as not to be non-compliant, so what other projects ship does not decide what a delivery must be.

This section is raw, and should be read that way. It mixes measured results with ideas about what to do next, its corpus definition has moved twice, and some counts in the prose are stale against the current corpus.

• Community exemplars. Who outside this project ships glTF to Gazebo and RViz, and what their files contain.

Audits of our own deliveries against the profile are planned and will use the same tools.

The corpus assessment lives beside its inputs in assess/, so that the generated report and the hand-edited criteria it answers stay together:

• criteria.md: what is assessed and why, criterion by criterion

• examples.yaml: the list of assets

• findings.md: what the corpus says, written by hand

• ASSESSMENT.md: generated by gltf-assess, never hand-edited

• Community exemplars: who else ships glTF to Gazebo and RViz
  • Method
  • The cohort
  • The people, and the two organizations that turned out not to be doing this
  • What this changes for us
  • Where the argument is being had