(j3.2006) Interesting F2003-ism

[Aleksandar Donev]

On Monday 26 February 2007 17:12, Bill Long wrote:

> Well, it is not currently a Cray extension.  It is a directive, and one
> supported by more vendors than just Cray.
None of the compilers I use supports it, or if they do, it is not advertised. 
All have compilation switches that say "do away with interface checking", 
which is at best a hack.

> As said earlier, the idea of forcing the Fortran programmer to write 
Did I propose to force the programmer??? At present, we are *preventing* the 
C/Fortran/whatever programmer from writing more than 1 interface with the 
same binding label. Allowing such a thing is *permitting* the programmer to 
get things done, not forcing. It hardly prevents us from fixing TKR 
genericity in the future (but I have yet to see a promissing 
proposal...obviously directives are not an option).

>  This is the WRONG solution to this problem.
Do we at least agree a problem exists? What is the right solution?

> Feedback from the TR vote suggested that the information visible to the
> C programmer be in terms of C concepts.  So the TK information should be
> in terms of names line int, long, float. double. ...
Yes, I was thinking of predefined constants like TYPE_C_INT etc., based on the 
names already used in the interop module.

> That's assuming such C/C++ libraries would be written.  I'm convinced
> that the initial use will be the other way
But who is arguing that? What difference does it make what the initial, 
primary, or any subset of the use will be. The question is whether we should 
do a minimal effort to try to do better. We are writing a standard for 
everyone and for use for the next 10+ years.

Are you even seriously interested in what C/C++ libraries use or would like to 
use from Fortran interop? If you are serious (that is, you want to try to 
*collaborate* with people writing such libraries, rather than tell them what 
you think is right or wrong), I can ask the people at LLNL from these 
projects:

http://www.llnl.gov/CASC/components/
http://www.tstt-scidac.org/intro/index.html

both of which are concerned with ebing available in multiple languages, 
including Fortran. And they both use Craig's descriptor library 
(http://chasm-interop.sourceforge.net/), which does what the TR is supposed 
to standardize...

If you are not really serious and just want to prove yourself right, don't 
bother, I have other stuff to do...

Aleks