(j3.2006) Asymmetry of REAL, CMPLX intrinsics

[Robert Corbett]

Aleksandar Donev wrote:

> On Monday 07 January 2008 20:30, Robert Corbett wrote:
> 
>>You
>>might be surprised to learn how few people are pounding at the doors
>>for UDDTIO.
> 
> You might be suprised to learn that even I, having participated in the 
> creation of F2003, will not use features until they are implemented in *all* 
> the compilers of interest to me (presently only 2) cause otherwise I cannot 
> run my simulations, so what is the point?!? I may run some tests of course.

Some of Sun's customers have asked for some features of Fortran 2003,
even though they are not implemented by all existing compilers.

> And the customers that can pound on your door (and break it) have such large 
> inertia that they may never even realize Fortran has OOP...

Any customer can pound on the doors.  We has a mechanism called
"request for enhancement," which is available to all users.  Few
customers can break down the doors, but some can.

One large customer has asked for the object-oriented features of
Fortran 2003.  No one has asked for UDDTIO.

Bob Corbett