Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!killer!ames!pasteur!agate!helios.ee.lbl.gov!nosc!ucsd!ucsdhub!hp-sdd!hplabs!hpl-opus!hpccc!hp-sde!hpda!hpcuhb!hpcllla!hpclisp!hpclcdb!cdb
From: cdb@hpclcdb.HP.COM (Carl Burch)
Newsgroups: comp.lang.fortran
Subject: Re: Double Precision COMPLEX
Message-ID: <6690018@hpclcdb.HP.COM>
Date: 22 Jun 88 05:37:45 GMT
References: <10423@ames.arc.nasa.gov>
Organization: HP ITG/ISO Computer Language Lab
Lines: 22


> Is the data type Double Precision COMPLEX part of the
> FORTRAN 77 standard?  
No.

> How about arbitrary precision COMPLEX in the new 8x standard?
No, not "arbitrary" precision - the committee calls it "generalized
precision", but (as a compiler-writer) I prefer to call it "compiler-
selectable" precision.  The user specifies the precision needed for
the problem, and the compiler selects the machine data type giving
that precision (and exponent range) or greater.  COMPLEX is also
generalized in this way.  One of the problems with the current draft
is that the results are "processor-dependent" when the request exceeds 
the data types on the machine.

The political prognosis of all this in the F8x standard is kind of
mixed - there seems to be a good bit of sympathy in the committee for
simplfying the generalized precision feature, but there seems to be
a consensus that some form of DOUBLE COMPLEX is a high-priority
requirement.
					- Carl Burch
					  HP Fortran