Re: Drastic revisions in draft 1.6.0

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

> What we should do at this point is try very hard to fix mistakes in
> the current draft, perhaps to add a clarification sentence here or
> there, and possibly to delete troubled subsections -- but not to
> rewrite entire clauses which would then have to be read and commented
> on by yet another round of ballotting.

Here's where opinions differ.    Although everybody who has touched the 754R
effort has probably had moments of fatigue and regretting getting involved,
many of us have had 25 years as well to regret the things that we didn't get
right enough or clear enough in 754.    I think from the point of view of
application programmers, the paramount issues are expression evaluation and
exception handling (and indeed when you get down to it, 
expression evaluation is exception handling too, for inexact
and exponent spill)... and being able to specify what you
want in a portable way, or even better a portable efficient way.     I see
no value in closing the process while there are clear improvements available
in these areas.

Other people approach 754R from other points of view, and come to different
conclusions.

We have spent a lot of the
last 7 years arguing about many matters which don't seem to concern application
programmers, and so the working group ran out of time without resolving the
points above or even addressing some of the issues in the PAR.    
The wording of the PAR is a whole other discusssion... 
but what we should have done and when is not as important as
what we should be doing now with what we have.