New table-based annex D revision

["David James" <dvj@xxxxxxxxxxxx>]

All,

My thanks to the many comments that have been received, both
in public and in private. Based on these comments, I have
updated the annex-D (table version), in preparation for
publication concurrently with the approval of the 754 standard.

The most recent changes have been in response to:
  >> Btw, Dave James:  you need a column for "n+0.5" in your table D.2 of
  >> your tabular version of Annex D.
And observations that columns are better oriented neg-to-pos,
for left-to-right scanning.

A copy of this revised text is attached.
As always, any/all comments are most appreciated.

HOWEVER, please be aware that the goal of the tables is to
clarify the 754 draft text, not to change its technical intent.
Please address complaints on the 754 draft content to the
others, who are empowered to change this text, not myself.

As with most of the 754 standard functions, the tables forced
significant editorial license, where the standard is unclear.

Yet another illustration of how reproducing the same results
is unlikely, given the multiple possible interpretations of
the document by distinct readers/users (sigh:<).

Cheers,
DVJ 

[2. application/pdf; dvj-Func754-2007-06-22.pdf]...