Jean Vezina wrote:
>> It should be noted also that an explicit interface is not required
>> for character dummy arguments having a length of (*). Consequently,
>> the test program is standard conforming.
--On Saturday, April 3, 2004 2:13 AM +1000 Robin <[log in to unmask]>
wrote:
[in response to someone else
> While one is not required (as are most of the old style Fortran
> procedure calls) it is recommended that one be included,
Sure, but that has nothing to do with the subject. I also recommend
looking both ways before crossing a street.
The recommendation to use explicit interfaces is completely separate
from any question about character*(*) dummies. There was a prior
statement in this thread (I forget who from, so I won't say, lest
I misattribute it) that an explicit interface was required in this
case. That statement was incorrect. jean was pointing that out and
also answering the original question - about whether the code was
standard conforming.
I haven't seen anyone even suggest that explicit interfaces
aren't a good idea. However, I think it important to avoid confusing people
about the distinction between good ideas and requirements of the standard.
That now appears to be clear here (I hope).
--
Richard Maine | Good judgment comes from experience;
[log in to unmask] | experience comes from bad judgment.
| -- Mark Twain
|