Re: Draft telcon minutes

Thanks, Sandra.  Two small comments...

At 09:50 AM 10/22/02 -0400, Sandra Martinez wrote:

>[...]
>Ckpt.  10.3: Follow Web Accessibility Initiative and Internationalization 
>Guidelines." Do we really want to get on the business to make other GL 
>apply?  Is this within the scope of the SpecGL?
>
>         This particular checkpoint is mentioned in the introduction 
> section of the SpecGL.

The point was more like:  "Pubrules and Style Manual are mentioned in the 
Introduction, as other documents needing attention in addition to SpecGL, 
and WAI and Internationalization could be handled similarly."  [My draft 
Intro doesn't mention WAI and IG currently, nor this checkpoint.]

>Lynne recommended the deletion of this checkpoint and the modification of 
>the introduction section to include more information in this particular 
>area.  Editors will draft appropriate text for the introduction.
>
>         Deviations from the agenda; Dom, asked for comments in the 
> introduction section of the SpecGL. Lynne, requested the WG to send 
> comments on the introduction section. Lofton added that the group should 
> also look at the @@ in the SpecGL by midweek.  He commented that in the 
> process of rewriting the SpecGL, guidelines 14 is collapsing with GL-15, 
> therefore removing the granular grammar related checkpoints. The 
> relationship between automatic generation and granular grammar is lost. 
> Loftton move the information related to granular grammar to an appendix 
> so it would not be lost. Dimitris volunteered to generate appropriate 
> writing on the subject.
>
>Ckpt. 12.1: "If an ICS is included as part of the specification, indicate 
>whether it is a normative or informative part of the
>specification." How an ICS can be normative? To what class of product 
>would it apply?
>
>Lofton agreed with Dom on the comment that ICS can't be normative. Lynne 
>explained that the inclusion of an ICS is beneficial in claiming 
>conformance especially for implementations that implement discretionary 
>behaviors and values, for example  XSLT specification where many options 
>and choices are allowed. DOM still not sure if the SpecGL should required 
>ICS.  Lofton added that it is not within our scope to address conformance 
>claim but that under some conditions it might be useful to have an ICS.

Hmm, perhaps I mis-spoke.  I think it *is* within our scope to define how 
to make conformance claims (ie., I agree with revised CK11.2 that imposes 
minimal content requirements on conformance claims), and I think that 
SpecGL should require that specs not only have an ICS (if only a default 
sample), but also in principle could require that specs require those as 
part of the minimal content of a conformance claim.  [Note.  I said "in 
principle could", because I haven't made up my mind yet, whether or not it 
*should* -- however, it is a possibility.]

>Andrew stated that he sees it at the role of the certification 
>organization to have the  ICS, so it would not belong in the SpecGL.  It 
>was suggested the checkpoint to be moved to the TestGL. Kirill 
>agreed.  Mark did not agree and added that requirements belong to the 
>specification not to a test suite, and stated that the ICS does belong in 
>the SpecGL. Lofton felt that this discussion was leading to reopening 
>issue 96 and decided to take the discussion offline.

All for now,
-Lofton.

Received on Wednesday, 23 October 2002 15:36:18 UTC