Re: [PATCH] Fix FMT= with CHARACTER arrays (PR fortran/39865, take 3)

[Tim Prince <TimothyPrince at sbcglobal dot net>]

Walter Spector wrote:
> Gents,
> 
> As an old F66 hacker, I'd like to offer the following:
> 
> 1.) Hollerith constants could only be placed into integer,
> real, and logical variables - i.e., those occupying a single
> numeric storage unit.  So supporting Holleriths in complex and/or
> double precision was an extension.
> 
> 2.) A Hollerith format spec *must* be an array name - not a
> scalar or an array element.  Supporting a scalar or array element
> would be an extension.
> 
> Since F66 had no obvious mechanism for keeping track of how
> long arrays were (when passed in from other procedures), the
> I/O library would basically have to count parenthesis to figure
> out where the end of the format spec was.
> 
> The Hollerith-related stuff is scattered out a bit in the
> F66 Standard.  Easiest is to look in Appendix C of the F77 Standard,
> which grouped everything together.
> 
> Walter
I checked the manual for the f66 compiler I learned on, and nowhere is the
fmt= specifier mentioned.  Nor does that F77 appendix C mention the
possibility of using fmt= with Hollerith.  Thus my belief that fmt= was
not in the standard until f77, when Hollerith had already been eliminated.