(j3.2006) Question about co-arrays

[Andy Vaught]

On Fri, 9 Feb 2007, Aleksandar Donev wrote:

> The "feature" is trivial. It is more work for implementors to prohibit C_PTR 
> in I/O than to allow it....after all, "implementation dependent" means they 
> can simply print a blank character if they are really that lazy.

  Despite the silly suggestion, I can assure you that you're incorrect on
this point.

> Telling people to use horrid things like TRANSFER to do a simple task is, 
> well, horrid...IMO, of course.

  But what is wrong with TRANSFER()?  You're asking for a construct that
does the exact same thing.  In the case of pointers, TRANSFER() is even
portable across architectures.

        Andy