[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
Re: motion for Aug: move mixed-radix comparison to "purple text"
This is item 16 on the August agenda. Items 11 through
15 are tabled motions 23 through 26 & 28 from July. - Dan
Date: Thu, 27 Jul 2006 10:21:03 -0700
From: Jim Thomas <jwthomas@xxxxxxxxxx>
Subject: motion for Aug: move mixed-radix comparison to "purple text"
Dan, I'd like to make the following motion at the August meeting.
-Jim
-----------------
Motion: move specification about mixed-radix comparison to "purple text"
Rationale:
The current draft in 7.12, paragraph 2 says:
"Comparison between decimal and binary floating-point numbers shall be
provided only if this operation reflects the true relationship between
the actual values .... Such comparison operations should not use the
same operation names or operator notation as same-radix comparison."
The first sentence is strange in that it is normative partial
specification of an operation that isn't in the draft standard (754R
doesn't define a mixed-radix comparison operation). We haven't
undertaken to enumerate and prohibit bad behavior of extensions to the
standard. The second sentence makes a recommendation about binding (for
an operation that isn't in 754R), though 754R generally does not address
binding. The gist of 7.12 paragraph 2 will fit as implementation
guidance, but not as part of the normative standard.
Text changes:
7.12, paragraph 2: Reword the first sentence to: "Comparison between
decimal and binary floating-point numbers, if provided as an extension
to this standard, should reflect the true relationship between the
actual values ...."
7.12, paragraph 2 and footnote 1: Move to "purple" text.