Annotating images with Annotea.

Hi,

One thing I want to do with Annotea, is annotate resources other than *ML
pages, in particular images.

SVG paths and drawings seem to be the obvious way of doing having the
actual annotation information, but what's a good way of storing it in
Annotea?

I see two options for that, either simply providing a a:body of
content-type image/svg+xml which then any Annotea can just query - the
advantage of this is the simplicity of doing it and extending existing
clients (they just have to display the SVG document, no processing
needed.) In fact, I've already done this, if you look at the annotations
on http://jibbering.com/imgs/shepherds.jpg you'll see they include
http://annotest.w3.org/annotations/body/1016716304.319541 .  Amaya
doesn't try to get annotations on image content-type urls, Snufkin
displays a link to the SVG and reports the content-type, and Annozilla
won't install on my 0.9.8 [*]

The other alternative is for the Annotations to be in RDF, and the SVG
then generated from this RDF, this will give better flexibility in the
display, and allow for annotations to be combined, however the parsing
will be more complicated - http://jibbering.com/svg/AnnotateImage.html
has an image annotation tool in SVG which generates RDF and SVG type
annotations. My thoughts on what the RDF might look like is in
http://jibbering.com/2002/3/samplesvg.rdf
(http://jibbering.com/rdf/foafwho.html already parses something similar.)
This would also mean you could have the body containing non-svg'd
annotation for those clients without SVG support.

My questions are.
Is this a reasonable use of Annotea?
What do you think is the better idea, SVG content-Type, or RDF ones?

Cheers,

Jim.

[*] I'm prompted to download chrome://annozilla/content/addpanel.xul,
rather than Mozilla executing it, so it never appears in the sidepanel
which I assume is essential.

Received on Thursday, 21 March 2002 14:02:29 UTC