history of CP3.2

I was supposed to check into the history of CP 3.2 (AI-20030107-4).  We 
were trying to figure out about the SHOULD clause.

I can't understand quite how it came about.  It appeared in in the editing 
cycle before the 20021108 publication, when I put a "To fulfill" clause 
into every checkpoint.  I think it is attempting to capture some subtlety 
in earlier text.  I'm not sure what that subtlety is, but it seems to be 
something like:  spec versioning/errata MUST be supported in test materials 
-- the most important end result of all of the QA deliverables -- and other 
QA deliverables such as planning documents ought to factor them in all 
along the way (SHOULD).

What I have done for the interim is reword the fulfillment criteria 
accordingly:

>Conformance requirements: the Working Group MUST ensure that the final 
>published test materials support specification versioning/errata, and 
>SHOULD address the topic of specification versioning/errata support in any 
>other QA deliverables such as intermediate planning documents.

That's the best I can do for now.  I think we will be taking another look 
at it in the Last Call review cycle.

Any objections to leaving it like that for now?

-Lofton.

Received on Monday, 20 January 2003 15:32:17 UTC