Re: A few comments on SpecGL

At 11:18 AM 8/2/02 -0400, David Marston/Cambridge/IBM wrote:
>[...]
> >I've been out of things for awhile - so, I don't understand what the
> >problem is. Conformance clause has been an accepted and widely used
> >term. What is the objection to using the 'clause'.   Is there really a
> >confusion over the use of the term or are we anticipating that people
> >aren't able to figure out what is meant.
>
>It's just about what document structure you have if you satisfy the
>checkpoint to have the necessary conformance verbiage present, but you
>don't satisfy the checkpoint to have a separate conformance section.
>You don't want that verbiage spread all over, a sentence here and a
>sentence there, due to the findability checkpoint. Maybe if QAWG were
>to do priority resetting on those 10.x checkpoints right away,

We could put this on next (6-aug) telecon.  The general plan is to work 
substantive SpecGL on 6-aug, with an extra telecon on 13-aug for all 
priorities, and any leftover substantive.

>it
>would clear up what possibilities are allowed beneath the ideal of "an
>easily-findable conformance section that points to all pertinent
>parts of the spec and provides a packaged description for use by a
>test lab."

It is an idea to consider, unifying the checkpoints.  The single reason 
that I can see for the P1 "conformance clause" (possibly scattered) and the 
P2 "single section" is:  allow level-A compliance for legacy 
specifications, that might have gotten all of the right stuff, but in 2-3 
different subsections.

This could be handled with a single checkpoint as well.  The can be 
multiple, bulleted individual requirements under a single checkpoint, so 
the "single section" bullet could have a legacy-specification exemption 
(i.e., "not applicable for those specifications which pre-date this document").

-Lofton.

Received on Friday, 2 August 2002 16:56:32 UTC