Re: Extensibility and Uniformity; defining "error" behavior

On Dec 30, 2008, at 7:53 AM, Shane McCarron wrote:

> Ray Denenberg, Library of Congress wrote:
>> In any case "interoperable" is not the proper term to use. I argued  
>> this for several years and eventually gave up.  But I have to back  
>> Larry up on this. His definition is the traditional one in standards.
> +1 - I have been doing this nonsense since 1985 and that's always  
> the definition we use.

+2  I've always understood the word in more or less Larry's sense.  
However, it don't see (contra other comments) that the criteria in 
   are necessarily at odds with Larry's definition. This refers to  
test suites, which are collections of examples provided by a specs  
authors which are used to illustrate required or expected  
interoperation behaviors, such as not crashing or delivering a certain  
type of response under certain conditions. The RDF and OWL test  
suites, for example, consisted largely of example parsings,  
entailments and non-entailments appropriate for fragments of RDF and  
OWL syntax. None of this requires uniformity of the user agent's  
implementation, only in certain aspects of its behavior (those  
required for interoperability, in fact.)

IHMC                                     (850)434 8903 or (650)494 3973
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile

Received on Tuesday, 30 December 2008 17:58:30 UTC