Skip to content

feat: host manifests and docs#2

Open
allison-truhlar wants to merge 12 commits intomainfrom
host-manifests-and-docs
Open

feat: host manifests and docs#2
allison-truhlar wants to merge 12 commits intomainfrom
host-manifests-and-docs

Conversation

@allison-truhlar
Copy link
Copy Markdown
Collaborator

This PR adds the viewer manifests and logos to the capability-manifest repo. Previously, they were hosted in fileglancer. The README is updated to reflect this change.

This is related to PR #307 in Fileglancer.

Note: before merging this PR, the URL for the logo location will need to be updated to point to this repo's main branch instead of the host-manifests-and-docs branch. This should also be accompanied by a new patch release and a corresponding version update for this package in the fileglancer package.json.

@krokicki

Add manifests/ directory with canonical capability manifests for
Neuroglancer, Avivator, OME-Zarr Validator, and Vol-E. These can be
loaded by URL so non-Fileglancer consumers can access them.

Also fix neuroglancer omero_metadata to true in public/viewers/.
Expand README with:
- Canonical manifests table linking to manifests/ directory
- Full viewer section schema (name, version, repo, template_url)
- Full capabilities section schema with all fields documented
- Table showing how validateViewer() maps each capability check
  to errors vs warnings
Replaces `compatible: boolean` on `ValidationResult` with two separate
fields: `dataCompatible` (hard gate — viewer is shown or hidden) and
`dataFeaturesSupported` (soft gate — warnings are shown). Reclassifies
`channels`, `timepoints`, and `hcs_plates` from hard errors to soft
warnings, since viewers that don't support these features can still open
the data. Fixes `getCompatibleViewersWithDetails()` to call
`validateViewer()` once per manifest instead of twice. Fixes version
comparison to use string equality via `.map(String)` instead of
`parseFloat`, avoiding future breakage with versions like "0.10".
Adds CoordinateTransformation and DatasetMetadata types modelling the
OME-NGFF coordinateTransformations spec (identity, scale, and translation
each supporting inline list or path reference). Adds a
hasTransformationType() helper and soft-requirement warning checks for
scale, translation, and bioformats2raw_layout. Adds a tracked TODO for
rfcs_supported explaining the spec-level gap that blocks implementation.
Documents the dataCompatible/dataFeaturesSupported distinction, updates
the checks table with a Level column showing which capabilities are hard
compatibility requirements vs soft feature support, adds rows for
scale/translation/bioformats2raw_layout, and notes the rfcs_supported
spec-level gap.
Zarr v2 stores compression info as `compressor: { id: string }` in array
metadata; Zarr v3 uses `codecs: [{ name: string }]`. The previous check
only handled v2 and had a broken fallback (`|| metadata.compressor`) for
the case where `.id` was absent. Replaces the single-codec check with a
collection step that reads from whichever field is present, then checks
each codec individually — emitting one error per unsupported codec and a
single warning when the viewer declares no codec support at all.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant