https://docs.google.com/document/d/16l2YoU4Portp5E89YEYMs_t0w48aBQut9Y0KjU-TnPs/edit#heading=h.50h2zzws4jyy

• Which of these can be done using the specifications, as is?
• Which can be done using client and server tools that already exist?
• Which should the IIIF community document as common use cases or patterns, perhaps as part of a Cookbook?

User Experience

• How do I know if I can comment on it?
• If I can, how do I locate and specify an Annotation server that I have an account on?
• Once I do that, how can I create a working draft?
• How can I publish it?
• How can I edit/delete it?
• Can I use annotations to deskew images? For example, by creating and rotating bounding boxes as annotations on that content?

Authentication and authorisation

• How can I set read/discover permissions on my annotation so that access is limited to a class working on a transcription project?
• What is the absolute [meta]data that every annotation must contain about itself to be trusted?

Discovery and Notification

• How do others discover my annotation?
• How can I indicate to someone (an editor or teacher) that my annotation is ready for their review?
• How do I search across annotations and across annotation servers that may host annotations about my resource?
• How can I search or display only translations, or only transcriptions, or only commentary?
• Is it useful to search annotations made as part of an incomplete or unpublished study/transcription/commentary?
• How do I incorporate annotations into an authorative data store?

Preservation / Versioning

• Should a “published” annotation/collection be made as durable as an edition of a book? What happens when annotations are responded to with other annotations?
• Where does versioning intersect with collaboration?
• How should machines resolve sets of content following different conformance levels, versions, and standards?
• Will is be helpful to know in 200 years that sarah3123402@mail.facebook.com was absolutely authenticated as the commenting agent on this annotation?
• Is there is "an easy way" to extract all the annotations for later text analysis.

Semantics

• How do we distinguish between annotations about an object and annotations about an image? * How does that relate to Canvases, and also the presentation api as being about Presentation, not semantics or identity?

General

• I’m also interested in machine-generated and machine-consumed annotations—OCR, but also object recognition and other annotations not intended for direct display to the user.