Re: User Facing Documents

There are at least two issues:

	1) Whether to deliver outreach material sooner rather than later. I  
actually prefer delivering our outreach *deliverables* later, after  
the design of the language has been settled. I am, as evidenced by my  
constant evagelization, more than happy to see outreach to users. OWL  
1.1, itself, is the result of outreaching to users.
	2) What exactly the outreach deliverables will looklike.

On 2 Nov 2007, at 11:59, Jim Hendler wrote:

> actually, what I had in mind was to actually write our charter  
> deliverables including the "descriptive specification"

It's still open whether this will be a distinct document. I've  
expressed before that I think the structural specification could be  
made to handle both descriptive material and formal specification.  
(Though that must be done carefully.)

One model for this is the RDF/XML Syntax Spec:
	http://www.w3.org/TR/rdf-syntax-grammar/

Section 2 is the single best non-tutorial intro to the language. I've  
used it in classes; I've used it myself.

I know you don't like the structural specification, but I don't like  
either splitting it out or making the descriptive spec too elaborate.  
As we disagree, I hope reasonably, there is some work to be done to  
get consensus.

> and the "user guide" -- I don't see where we have any choice on that.

Similarly, we've not done *any* work to determine what the WG as a  
whole would think is an acceptable and useful user guide. Some people  
have expressed a desire for diff or near diff documents. I've  
expressed some qualms at that.

Both you and Jeremy (he in telecon) have tried to make the my  
disagreement to be a matter of my ignoring the charter. Please stop  
that. My current questions are about timing (I *don't* think it's  
necessary or helpful to produce a *comprehensive* descriptive spec  
before the design is done) and the form.

In fact, I've not heard a coherent description of the particular form  
of the documents, or the users they are targeting. For example,  
descriptive material targeting web developers, vs. targeting  
triplestore authors, vs. targeting HCLS modelers can look very  
different. Documents targeting one might be rather less successful or  
even fail with others.

Furthermore, Jeremy apparently is proposing producing Working Drafts  
that aren't rec track, or submission track, but /dev/null/ track  
(i.e., deliberately designed to be dropped after one or two  
versions). I'm not sure how I feel about that on several levels.

Cheers,
Bijan.

Received on Friday, 2 November 2007 12:27:41 UTC