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

Motion 39.01 YES



I vote YES on the text of Clauses 1.1 through 1.6.

I do have minor issues.  It is possible to interpret the text in a
satisfactory manner (hence my YES vote), but the text could perhaps
be improved.

1.1:  I had already made a suggestion with regard to the mention
      of IEEE 754 types.  John replaced the really objectionable
      text with the already-agreed-to text of 2008, which is ok,
      as it can be interpreted in a way that does not require the
      IEEE types.  My suggestion was unambiguous:

         ...and at least one fully-specified numeric type
            such as an IEEE-754/2008 floating-point type.

1.4:  Exclusions  ...except for interval types derived from IEEE 754

      Ambiguous again.  My generous interpretation is that SOME aspects
      of (Level 3) bit patterns would be specified (but vacuously, as
      they would be the 754 specifications) -- but it could be interpreted
      as saying that the complete Level 2/3/4 representation of intervals
      based on 754-types would be specified, e.g. [lb, ub] as consecutive
      754-types, and not, say, [-lb, ub].  I would vote NO if the latter
      was the intention.  (Note that this would also have required a full
      specification of decoration encoding -- which we REALLY don't want.)

      My suggestion is simply to drop the 'except for...' part.

Michel.
---Sent: 2012-12-20 00:17:43 UTC