Results of voting on Discretionary Items report

[David_Marston@lotus.com]

Six committee members voted on the Discretionary Items report that I sent
on April 18th, and the results are mostly positive. I hope we can isolate
the content issues, clear them up as necessary, and improve the layout as
we move ahead.

The major question, in my view, is whether we have indeed found everything
in the specs and errata that is a deliberate grant of discretion. Based on
the history of this document and your votes, I would say we have.

You also seem to agree with the dividing line between testable (first two
groups) and non-testable (other two groups) in most cases, and I hope that
my subsequent explanations helped to assure you of the placement of the
items you questioned. Much of the uncertainty of your votes relates to the
divisons within testable (do now vs. postpone) and within non-testable
(core XSLT vs. out of scope). For the former, we can downplay "postpone"
to be just an "action flag" on each item. The non-testables could be
lumped together. That way, we'd have just two major sections. Kirill and I
will discuss possible refinements on Thursday, and propose whatever fixes
to the whole committee.

I think we can make wording changes like Ken's note about rounding in
xsl:number on an informal basis. It's also clear to me that we should
improve the wording about applying the discretionary flag on a test case.
Some of the feedback isn't so much about what the report says as it is
about whether a particular flag might apply to a particular case.

Once we have the set of items bundled and categorized acceptably, we
can XMLify the content to allow multiple renderings and better
indexing and linking. The citation pointers should be improved, too.
Both of these things could occur later, after the content meets the
needs. I think it's worth taking an extra week or so to settle the
remaining content questions.
.................David Marston