Fortran Versions (was JB Passes)

Chuck Guzis cclist at sydex.com
Tue Mar 27 01:51:14 CDT 2007


On 26 Mar 2007 at 22:05, Jerome H. Fine wrote:

> I don't have the time to compare FORTRAN IV and FORTRAN 77,
> but I seem to have the feeling that I can use code in FORTRAN 77
> that is completely acceptable in FORTRAN IV.  Maybe not every
> program can be compatible (especially DOUBLE PRECISION and COMPLEX),
> but a simple demo program should probably be identical.

Sans vendor-specific extensions, you're probably right.  At least 
that was the goal of F77.  I did a few X3J3 sessions with what was 
then known informally as the FORTRAN 8x ANSI group (vector extensions 
on behalf of ETA) and at least then (1985), the goal was to maintain 
backward compatibility, albeit with quite a number of "deprecated" 
features.  It was a nasty experience with threats by the IBM 
contingent to walk out of the proceedings. 

A big argument was "Should X3J3 simply select the best of the current 
extensions and ratify them or should we be about designing a new 
language?"  It was a serious issue; IBM wanted Vectran ratified; many 
wanted an alternative syntax.

Needless to say, with all of the contention, FORTRAN 8x wasn't about 
to meet the informal goal of becoming FORTRAN 88--it took until 1990--
and I'd been long gone.

Full-blown F90 looks about as much like F77 as F2003 looks like F90. 
It's kind of like going back to the old neighborhood after 30 or 40 
years--you recognize some things, but it's mostly foreign.

I rather liked Basic FORTRAN IV (that's what it was called).  A 
minimal functional subset; no logical IFs, ASSIGNed GOTOs, object-
time FORMAT or named COMMON.    Just a handful of statements that you 
could pretty much compile and run anywhere.

Cheers,
Chuck




More information about the cctech mailing list