Notes for draft review meeting at Hewlett Packard in Palo Alto
	on Thursday 9/15/05 in the Sigma Room in Building 3.  Peter
	Markstein hosted us.  Jeff Kidder, Dick Delp, Jon Okada, Jim
	Thomas, Peter Tang, John Harrison, Debjit Das Sarma, Alex
	Fit-Florea, Mike Cowlishaw, David Hough, Prof Kahan, John
	Crawford, Ivan, Godard, & Dan Zuras attended.  Mark Erle, Eric
	Schwartz, Alex Liu, Roger Golliver, & Steven Carlough were on
	the phone.


	We began with the outstanding issues led by Dave.  The
	discussion ranged through non-canonical declets, the taxonomy
	of varying precision implementation styles, pause versus abort,
	the utility of alternate exceptions that are standard across
	platforms within a given language versus across languages
	within a given platform.

	After the 3:00 break we heard: a presentation by Jeff on
	converting binary to BCD (which, among other things, argued
	that conversion between DPD & BID is easier than had been
	argued in the past); a presentation by Peter Tang on further
	results for BID (involving fast divide-by-powers-of-ten for
	the purpose of correctly rounding the result of an arithmetic
	operation in BID using existing an binary multiplier of 64x64);
	a rebuttal by Mike (involving the problems of the various
	'software biased' format proposals versus DPD); and an
	announcement by me.

	My announcement was:


	I have come to the conclusion that this committee is not going
	to finish its work the way we are going.  To his credit, Dave
	has been telling me this for a long time now but I held out
	some hope that we could find a way to make it work.  The meeting
	last month convinced me that he was right & I was wrong.

	Last week I discussed the matter with the chairman of the MSC
	committee & the chairman of the IEEE Standards Board.  As a
	result of that discussion & provisions within the SA that allow
	a chairman to declare that no progress is being made, I am
	suspending these meetings until further notice on those grounds.

	The chairman of the standards board said that, in this case,
	there is provision for a committee chair to submit an incomplete
	draft standard where a committee can't reach agreement.  I said
	no.  The current draft is both incomplete & inconsistent & not
	something that can be usefully submitted.  He urged me to get it
	into a form that could be submitted & best represented that
	portion of the standard for which there was wide agreement.  I
	said I would but that I still wanted to give my committee a
	chance to form a consensus on the result.

	Therefore, I urge those of you who are involved in disputes over
	this standard to spend this time trying once again to come to
	some mutually beneficial agreement.

	I have asked David & Peter Markstein to aid me in preparing the
	skeletal framework of a draft.  For the rest of you, please don't
	be offended if I didn't include you in this group.  I did want
	to ask others but there were practical considerations such as
	the time you have available & communications difficulties to
	consider.

	If you want to make your opinions known, though, I have created
	a mail alias for you.  It is frame@nonabelian.com.  Mail sent
	there will be read & acknowledged but there will be no further
	reply or comment.

	I would like to ask those of you who have offered to host
	upcoming meetings to keep those dates & rooms available as long
	as possible.  I will let you know at least a week in advance if
	it will be used.  If we spill into 2006, I may ask you to host
	again in later months.  It will take as long as it takes.  The
	SA will give us whatever time we need.

	If any of you reach a useful compromise on your issues, please
	let me know.  I will let the mail reflector know when we begin
	again.  And when we do, the rules of behavior on this comittee
	& in this room will be VERY different from today.  But more on
	that when the time comes.

	One final thing.  Paul Zimmermann is flying out here in December
	& he wishes to speak to this committee.  Paul is one of those
	people who advised us on arbitrary precision arithmetic some
	months ago.  The only part of this that remains in the draft is
	what is now called varying precision.  I would like to have this
	committee get a chance to meet him but I don't know if we will
	be meeting at the time.  David is scheduled to host the December
	meeting at Sun in Menlo Park on the 14th & 15th.  In that event,
	perhaps we can get together on Thursday the 15th for an informal
	meeting anyway just to hear what he has to say.

	That is all.