Re: "Variability in Specifications" WG draft

At 10:44 AM 8/25/2004 +0200, Dominique Hazaël-Massieux wrote:
>Le lun 23/08/2004 à 16:56, Lofton Henderson a écrit :
>[...]
> > ** I notice that the glossary of SpecLite does not contain definitions for
> > profiles, modules, etc.
>
>Karl, could you fix this?
>Or do you mean there should be another glossary in this document?

At least fix SpecGL Glossary.

I had not thought about the possibility of a ViS Glossary.  I guess I think 
of ViS sort of like an external appendix to SpecGL, so I would say it can 
point to SpecGL glossary.  (Yeah, that sweeps some other questions under 
the rug, like special words that appear in ViS but are not needed in SpecGL.)


> > ** "Specification Category" -- this is under-specified.  I.e., it is
> > difficult to understand both the explanations of the different categories
> > (they each need at least a sentence of definition/discussion), and how 
> this
> > is critical to variability analysis.  I myself would have difficulty
> > writing a Spec. Cat. analysis based on this.  We should develop the 
> section
> > better, or remove it.  For now, maybe "green flag" it to indicate that we
> > think it needs attention.
>
>I've added an introduction to the section that tries to relate the
>specification categories to classes of products and variability; I have
>flagged an issue at the start of the section for the other issues.

Another thought occurred to me after I sent my comments.  The Spec. Cat. 
stuff is not an end in itself, but (as self-described) is a methodical way 
to get at the Class of Product.  So move the SC section after the CoP 
section -- less distracting that way, and more logical since it is a 
technique of sorts.

-Lofton.

Received on Wednesday, 25 August 2004 13:05:24 UTC