Thanks, Robert ... good observations and comments.
I experienced whiplash when I learned that the review target was DITA 2.0, not the DITA 1.3 multimedia domain that we had long planned. But if 2.0 is the target, I don't believe we would have designed the multimedia support in the way that we did. Using child elements to specify properties makes the vocabulary much more verbose than otherwise (9 element types instead of 2), and is especially awkward for Lightweight DITA.
I like the idea of adding audio and video to the DITA 2.0 base. I would favor defining attributes to specify properties, as does HTML5.
We can still release a DITA 1.3 multimedia domain as currently
designed, if it's the will of the TC to do so.
I'll note that this approach would have ramifications for Lightweight DITA, which I have barely begun to think through.
-Alan
Keith had a fascinating comment in the multi-media review that got me thinking - does the presence of audio and video supersede the need for the object element?
My gut reaction was - maybe so, but only if we make audio/video part of the base vocabulary (they can't be based on object and still mean object is unnecessary). But digging further, it's clear other uses are possible, so I don't think we can say the new elements supersede it. I've only used objects for audio/video, but here's a good HTML5 example of using the element to embed a PDF: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/object
At a minimum, Keith's comment points out that we need to clean up our reference topic for <object> so that it no longer talks about audio/video.
Beyond that, this comment -- and other chatter on the list during the review -- has me wondering about how much simpler things would be if audio/video were just base elements, rather than specializations of object. I know why we didn't consider that initially, and I probably risk the wrath of Kris or Chris in asking, but I wonder if at this point it's worth reconsidering? It would give us more flexibility in the definition to address some of the review comments that have come in. The down side is that it would rule out a backwards-compatible domain that works with DITA 1.3 and DITA 2.0. That said, the currently-defined domain markup would have a simple migration path into a DITA 2.0 model that uses base elements.
I don't want to go too far down that path without more discussion though...
Robert D. Anderson
DITA-OT lead and Co-editor DITA 1.3 specification
Marketing Services Center
E-mail: robander@us.ibm.com
11501 BURNET RD,, TX, 78758-3400, AUSTIN, USA
-- Alan Houser Group Wellesley, Inc. Consultant and Trainer, Technical Publishing arh on Twitter 412-450-0532