A proposed methodology for seeking consensus

Hi all,

the discussion from yesterday's meeting got me thinking quite a lot. In 
particular:
* I share Andy's concern about convergence
* I like what Gregg said about trying to build a common proposal rather 
than compare different proposal
* I agree with Thomas that the less controversial point seems to be the 
concrete Turtle syntax
* as many have repeated before, we should ground our design choices in 
use-cases

So may be we should start from a proposal for the concrete syntax,
and try to express our different use cases with this syntax.
Then, for each use-case, we identify some SPARQL queries that could be 
exectued agains the UC data,
and what would be the desirble outcome for that use case.
Note that I we should come up with SELECT queries, but also ASK queries 
(the latter being a proxy for talking about simple entailment).

 From the desired outcome, hopefully, we should be able to drive an 
abstract syntax that satisfies them, and a semantics.

Any thoughts?

     pa