I have not looked really hard for a workaround, but I have not found one.
What I have done in the past is to maintain parts of the code as f77, by
simply not adding an f90 suffix. Such inflexibility is certainly annoying
if one is used to being able to do casts in c, but as Dave says, we don't
really require it here...
Regards - Bob
----- Original Message -----
From: "David A. Case" <case.scripps.edu>
To: <amber-developers.scripps.edu>
Sent: Friday, December 02, 2005 6:03 PM
Subject: Re: amber-developers: fortran95 help requested
> On Fri, Dec 02, 2005, Michael Crowley wrote:
>>
>> fortcom: Error: _ew_recip.f, line 1716: The type of the actual argument
>> differs from the type of the dummy argument.
>
> This is reminiscent of a similar problem with the rstack()
implementation
> earlier. Generally, I think we should try to make sure that types
> explicitly
> agree. It's not clear to me that implementations of complex variables
are
> required to use two adjacent real variables -- so the compiler is
actually
> being correct in refusing this.
>
> Having strict type checking is sometimes a pain, but my guess is that
the
> advantages in catching errors outweigh the hassle in getting things to
> work
> the way the compiler wants....
>
> (But, maybe someone else knows a workaround.....)
>
> ...dac
>
>
http://www.rosswalker.co.uk/movie_sounds/
Received on Wed Apr 05 2006 - 23:49:46 PDT
