Talk:Core Platform Team/Initiatives/IIIF API

FWIW, after all this page is obsolete, I recently released an extension that may cover (in part) some of the aims outlined on this page:

Extension:IIIF

I thought I'd suggest some first "user scenarios". These two linked scenarios focusing on how Wikimedia Commons could function as a "IIIF Server Service" to institutions that have hi-res and openly licensed media, but don't have the resources to set up a IIIF Server of their own.

Wikimedia Commons as a "free IIIF-service"

The majority of GLAMs don’t have the financial resources or technical capability to set up their own IIIF-server. While this may change as vendors of specialised content management and digital asset management systems gradually add support for IIIF this means that IIIF is currently out of reach for most small and medium GLAMs.

Given the above, it could be very attractive a to GLAMs and other institutional users of Wikimedia Commons if all images and videos uploaded to Commons would be “IIIF-ified” and made available via a Commons IIIF Image API and, ideally, IIIF Presentation API.

A service like this would ideally be Wikimedia Commons IIIF-viewer that can be embedded off-Commons (see below).

Possible follow-up features:

- The Category representing the media upload - "Media from GLAM X" - could also be made available as a IIIF Collection.

- While it would be natural to start with images IIIF also supports video and audio. Support for "IIIFifying" video and audio could thus be a feature to add after releases focusing on images. Note that support for video and audio is recent to IIIF and there are not as many (or any?) open source IIIF Viewers that support them.

- When uploading PDFs, or other files with OCRed texts, publishing them also via a Commons IIIF Search APIwould lay the foundations for offering a better online reading experience, many open source IIIF-viewers have display modes optimised for this, including letting users search within the text.

- Allowing the user to annotate media with annotations made available externally in a standards compatible manner, like W3C Web Annotations and that many IIIF Viewers can consume and display. For an image annotation focused proof of concept see Wikidata Image Positions. User annotations could of course also be made possible to apply on texts (highlighting, commenting on or even translating words and paragraphs), video and audio (commenting at certain time-codes and with video also regions). User annotations is a big and complex scenario of its own really!

Support for this case

Based on personal communications with colleagues in GLAMs I think a “Wikimedia Commons as a free IIIF-server” would be a substantial draw. One important reason for this is that a service like this can save money. This is always a powerful argument in cash-strapped organisations and even more so when the COVID-pandemic has left most GLAMs strapped for cash while at the same time having the need to improve their digital services.

Wikimedia Commons as a provider of an embeddable media viewer

For the same reasons that less resourced GLAMs can’t buy or develop IIIF Server support themselves they are also unable to develop/adapt/adopt a IIIF media viewer. For those GLAMs the possibility to use, to embed, a viewer/player/reader from Wikimedia Commons on their own sites is an added benefit.

Another added benefit to the Wikimedia movement is that an embeddable IIIF-viewer will likely also be used in some cases where a user currently would download a media file and then upload it to their own website. This provides the opportunity for more representative view statistics which in turn makes the business case for institutional partnerships stronger. The drawback could be an increased load on the Wikimedia Commons servers.

+1 on how welcome an IIIF image-hosting service would be. (Though many GLAMs in the UK would have to think whether they wanted to try to continue to restrict their content to be CC-BY-NC, because that would not ever be an acceptable licence for Commons).

And it's not just small GLAMs that would be interesting. Even the biggest GLAMs can have departments with very cash-restricted budgets, and/or no money for experiment with new ways of making their content available. As reported a couple of years ago in , the maps library of one very prominent UK GLAM would love Commons IIIF service, for a large number of maps that the public have georeferenced. The (3rd party) georeferencing application needs the old maps to be available on an IIIF service, so that the app can pull in just the bit of the map it wants, at just the resolution it wants. (Image files for maps can be very big, to have enough resolution to allow words & detail to be legible). At the moment the GLAM's old map unit is having to pay a certain amount a year for 3rd-party IIIF hosting for those maps, money the unit doesn't have. It would *very* much like to be able to use the copies of the maps on Commons instead.

I must have read Tom Crane's blog post back in 2018 https://gist.github.com/tomcrane/70e879884a744ce69d329cc6a33a34ac 'cause really what I wrote above is pretty much what he already suggested there!

Seconding this: let's please start this :) What is involved, what's the up-front work needed and what would the marginal cost be?

Also, since you mentioned maps @Jheald, how does this compare to the overhead for maintaining a tileserver?

If we're going to use this page as a serious forum for discussing issues around IIIF development, it would be useful to have it as a conventional talk page, rather than a page of Flow threads. Flow is hopeless when it comes to following a discussion in sequence from the beginning, because everything is presented in most-recent-first order. It also doesn't give us the flexibility to organise and group sub-threads -- eg to group all threads discussing manifests together, or threads raising particular end-use cases. For this kind of thing, a conventional wiki talk page is more useful.

If this page can't be turned into a conventional talk page, then the talk page at eg either m:Talk:Wikimedia_Foundation_GLAM_team or c:Commons_talk:International_Image_Interoperability_Framework might be a better venue.

I like the Commons page, but since there may be broader proposals for integrating into Mediawiki installs, or connecting directly with other WM Projects, I've started a page on Meta: https://meta.wikimedia.org/wiki/IIIF