Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: mid-rad, inf-sup, a caution...

Dan,

I'd like first to hear any possible other opinions
concerning the importance of mid-rad vis a vis inf-sup.
What I was actually thinking however, in view of
Arnold's comment (but without additional input
from others) was to make inf-sup mandatory, but
specify mid-rad as optional and specify conversions,
provided mid-rad is provided.  However, I may be
overlooking other arguments that have been given
in favor of mid-rad.

Your suggestion (analogously to binary / decimal in
754) that a conforming implementation is allowed to
support either or both is intriguing.  In particular,
that opens the idea of other representations (other than
mid-rad or inf-sup);  I'm not sure how we would define
a standard in those terms, other than based on
the very simple command "thou shalt contain the
exact result."

Baker

On 5/10/2010 11:27, Dan Zuras Intervals wrote:

Date: Mon, 10 May 2010 10:58:28 -0500
From: Ralph Baker Kearfott<rbk@xxxxxxxxxxxx>
To: Arnold Neumaier<Arnold.Neumaier@xxxxxxxxxxxx>
CC: rbk@xxxxxxxxxxxxx, Dan Zuras Intervals<intervals08@xxxxxxxxxxxxxx>,
  =?ISO-8859-1?Q?J=FCrgen_Wolff_von_Gudenberg?=
     <wolff@xxxxxxxxxxxxxxxxxxxxxxxxxxx>,
  stds-1788@xxxxxxxxxxxxxxxxx
Subject: Re: mid-rad, inf-sup, a caution...

Arnold,

Thank you for your input and clarification.  Are there
additional comments concerning this?  Should mid-rad be
standardized as an optional type?

Baker



	Hossam's analogy was to binary versus decimal in 754.

	And in 754, a conforming implementation is permitted
	to support either binary or decimal or both.

	The analogy for us would be that a conforming 1788
	implementation would be permitted to support either
	mid-rad or inf-sup or both.

	That is, mid-rad is as optional as inf-sup but one or
	the other is required.

	Is that what you mean?

	Or do you mean to suggest we should have an inf-sup
	standard that specifies conversions to&  from mid-rad?

	Just so I understand it all...


				Dan




--

---------------------------------------------------------------
R. Baker Kearfott,    rbk@xxxxxxxxxxxxx   (337) 482-5346 (fax)
(337) 482-5270 (work)                     (337) 993-1827 (home)
URL: http://interval.louisiana.edu/kearfott.html
Department of Mathematics, University of Louisiana at Lafayette
(Room 217 Maxim D. Doucet Hall, 1403 Johnston Street)
Box 4-1010, Lafayette, LA 70504-1010, USA
---------------------------------------------------------------