Re: Proposed addition to Charter

[Danny Ayers <danny666@xxxxxxxxxxx>]

Sam Ruby wrote:

> Danny Ayers wrote:
>
> > I can't see much of a downside - Tim suggests that it might slow down 
> > the whole effort, but I really don't think any significant delays 
> > would be associated with this, it can happen in a separate processing 
> > thread. There is plenty of expertise available here and nearby, and 
> > less potential for disagreement given that the schema would be 
> > derived directly from the spec prose.
>
>
> I for one don't have a problem with the work being done (presuming 
> that there are volunteers), or with the inclusion of transforms or 
> other mappings into the final document.
>
> The sticky point to me is putting it in the charter, making it a 
> gating deliverable.
>
> FWIW, I feel the same way about schemas (xml, RelaxNG, or others).

Ok, put together that seems reasonable. I think the parallel with other 
schemas is valid - yes, anyone can use the format based on prose + 
examples, but having schemas ensures greater alignment between systems.

> On the plus side, such work increases rigor and sometimes exposes 
> questions which were not obvious otherwise.  The potential down side 
> is when recommendations come in to make compromises in order to make 
> the mapping easier.

I agree. If Atom is specified well as an XML web format, and the mapping 
isn't relatively straightforward, then it's RDF that needs to be looking 
for solutions (I don't think it will be necessary, but it seems a fair 
caveat).

Cheers,
Danny.

--
----
Raw
http://dannyayers.com