Model and Design for Annotations

[agildehaus6]

There aren't any specifications or guidelines on how annotations on images are going to be stored and presented.

Do any of the front end guys have designs or ideas of what they want annotations to contain or the functionality? I assume they are gonna need coordinates, an author, a title, an id, and the content itself. Possibly on top of that there could be r/w/delete privileges, comments on the annotation, links to other annotations ect.

[magnostherobot]

Image annotations are mentioned
in the HCI protocol document. This is enough information for a single HCI group to be able to create, store, retrieve and understand its own annotations.

If, however, HCI groups want to be able to interact with other groups' created annotations, then a protocol will have to be designed specifically for that.

[magnostherobot]

What specifically would HCI groups want to be able to support in annotations?
First of all, the shape:

• Single-point annotations?
• Quadrilaterals?
• Circles?
• Irregular polygons?
• Combinations of several (possibly disconnected) shapes?

I imagine text associated with annotations is a good idea, but do we want to specify a markup format within, to format the text neatly?
What about special visual features, like colour; outline thickness; filling opacity?
Is editing history important to annotations?
And how do HCI retrieve them from BE?

Personally, I don't want to make this myself, since I don't know what an HCI group would desire.