Re: [office-metadata] [Fwd: Re: ODF and semantic web]

["Bruce D'Arcus" <bdarcus@gmail.com>]

Oops ...

On 10/22/07, robert_weir@us.ibm.com <robert_weir@us.ibm.com> wrote:
>
> We'll probably get asked this question over and over again.  So do we have a good list of reasons why we cannot use the generic RDFa serialization?  Is it because it is un-doable hard?  Or is it that the incremental complexity is much greater than the incremental benefit?  Is this something that could be solved once, say in the form of an open source library, and then reused everywhere, so the cost/complexity issue is not a problem?

I think in a nutshell there was a worry that using RDFa proper--the
whole enchilada--would indeed add too much complexity for too little
benefit. My sense was the problem was probably less the actually XML
parsing code than the UI. If you have complex content marked up with
RDFa, what happens when a user edits that content? It seems the
primary focus of RDFa is web markup, often hand-authored and
relatively static.

We really have two choices:

1) keep things as they are, with the expectation we will not augment it later
2) change our attribute names such that our model is a proper RDFa
subset and could more easily add more of it later

Bruce