Re: summarizing proposed changes to charter

On 8/13/2014 22:08, Eric Prud'hommeaux wrote:
> I think you want the charter to prescribe some characteristics of the 
> extension mechanism. I appreciate the desire for direction, but 
> caution that this would in effect exclude other possible designs like 
> grammatical extensions (as occurred in SPARQL 1.1) or the MUST/MAY 
> UNDERSTAND (e.g. SOAP) model. I believe the current charter permits 
> your design without prescribing it. What do others think? 

While this topic is really closed (see my parallel response), I cannot 
claim that I understand your line of reasoning here. Grammatical 
extensions such as from SPARQL 1.0 to 1.1 are "compile-time" changes. 
What the WG will need to address are "run-time" changes that allow any 
user to express constraints that are not covered by the collection of 
predefined shapes. And I don't know SOAP well enough to know what 
MUST/MAY UNDERSTAND means (and a quick Google search was not informative 
either).

I don't think my proposal was excluding any design, it was just stating 
a problem. In fact, many people will find the ability to define custom 
constraints far more important than the use of pre-defined shapes. A 
standard that does not deliver that extensibility would be a failure.

Thanks,
Holger

Received on Thursday, 14 August 2014 00:02:20 UTC