Re: Issue 1144 regarding SpecGL 2.3

Le lundi 14 mars 2005 à 10:42 -0500, david_marston@us.ibm.com a écrit :
> I then proposed an alternative that focused on what could be measured
> in the document: 
> "The specification should provide sufficient detail in the normative
> references to prevent conflicting interpretations of the requirements
> imposed by reference." 
> 
> Dom then asks for it to be shortened, such as: 
> "Provide details about normative references" 

As Lofton mentioned, we can put more details as what we mean in our
"what does it mean" section.

> I don't think that is complete enough, because it doesn't include the
> goal. When we ask whether the Good Practice was followed, we mean
> something specific about the provided details, namely that they were
> used to qualify the other spec as needed for this purpose. The details
> are sufficient when there is no way for an implementer to twist the
> provisions of the other spec for selfish purposes. How about something
> like this? 
> "When imposing requirements by normative references, prevent conflicts
> and vagueness." 

It's going in the right direction, but I'm not fully satisfied yet; what
about: "Provide details on normative references to anticipate conflicts
and vagueness"?

Dom
-- 
Dominique Hazaël-Massieux - http://www.w3.org/People/Dom/
W3C/ERCIM
mailto:dom@w3.org