754R ballot comments

[David Hough 754R work <754r@xxxxxxxxxxx>]

I have revised 

http://754r.ucbtest.org/msc-ballots/ballot5.html

and added comments in .xml form

http://754r.ucbtest.org/msc-ballots/ballot5a.xml
http://754r.ucbtest.org/msc-ballots/ballot5b.xml
http://754r.ucbtest.org/msc-ballots/ballot5c.xml
http://754r.ucbtest.org/msc-ballots/ballot5d.xml

The most significant change was induced by probing comments from 
Michel Hack.     Clause 10 on Expression Evaluation was the most important
repair of 754 omissions, yet being optional
it has perhaps been under-reviewed and had
deteriorated from excessive patching when a rewrite was in order.
The most important changes are that some material that was more generally
applicable was moved from the widento section to the general section,
and "widenTo" was replaced uniformly by "minWidth" because it's really
about the narrowing destinations rather than widening operands;
754R arithmetic operation unrounded results depend on the values of the 
operands, not on the widths of the operands.

I have attempted such a rewrite

http://754r.ucbtest.org/msc-ballots/clause10.pdf

and will continue to work on it as I and other reviewers notice things.

It is supposed to be a clearer, more logical, less redundant exposition
of the same requirements, but if the text it is supposed to replace is
indeed less logical and redundant then there might be several incompatible
interpretations of those same requirements in reader's minds.   A clearer
exposition may bring those incompatible interpretations to light sooner
rather than later.

I hope that some application programmers and language designers and implementers
will at least take a glance and respond, by suggesting improvements 
or if necessary by redrafting clause 10 according to what they see as important.