ArchiveLabs / iiif.archivelab.org

Internet Archive IIIF Image 2.0 Server
GNU General Public License v3.0
30 stars 13 forks source link

Most popular Grateful Dead recording manifest has no items #123

Closed hadro closed 1 year ago

hadro commented 1 year ago

Will dig further in a bit, but the manifest for this item is currently empty with no items:

hadro commented 1 year ago

Ah, the item page says "Stream Only", even though there seem to be derivs in the metadata. Maybe this is an item type that we need special handling for?

hadro commented 1 year ago

This item is not "Stream Only", and also has no items: Item page: https://archive.org/details/gd72-04-14.aud.vernon.23662.sbeok.shnf Metadata: https://archive.org/metadata/gd72-04-14.aud.vernon.23662.sbeok.shnf Manifest: https://archivelabs-iiif-archivelab-org-presentation-v3.ux-fnf-misc.archive.org/iiif/3/gd72-04-14.aud.vernon.23662.sbeok.shnf/manifest.json

hadro commented 1 year ago

Here's another that probably should work, but doesn't: Item page: https://archive.org/details/deadco2023-07-01.EJL95 Metadata: https://archive.org/metadata/deadco2023-07-01.EJL95 Manifest: https://archivelabs-iiif-archivelab-org-presentation-v3.ux-fnf-misc.archive.org/iiif/3/deadco2023-07-01.EJL95/manifest.json

hadro commented 1 year ago

As far as I can tell, everything listed here has in common a mediatype of "etree", which seems like it's connected to the fact that it's included in the "Live Music Archive" collection.
But while some of those are "Stream Only", some of them definitely seem like open items with audio derivatives that seem like fair game for IIIF manifests.

I wonder if we need additional logic to parse mediatype: "etree"

hadro commented 1 year ago

This page lists the following possible mediatypes (noting that there also may be others) -- this list might be the list worth creating logic for:

hadro commented 1 year ago

With the latest media parsing logic, all three of the items referenced above now have valid IIIF manifests that will play audio in IIIF players:

Before I close this up, @mekarpeles, can you help us confirm that this is expected and desired behavior especially for the first item?

hadro commented 1 year ago

This is fixed up on the new infrastructure at https://github.com/internetarchive/iiif 🎉