HP3000-L Archives

July 2001, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
John Clogg <[log in to unmask]>
Reply To:
John Clogg <[log in to unmask]>
Date:
Mon, 2 Jul 2001 12:13:21 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (72 lines)
Ken,
Your posting seems to use the terms HP 3000 and HP e3000 to denote different
machines.  Are you using those terms to distinguish the classic and PA-RISC
versions of the 3000?  Is this HP's convention, or one that is becoming a de
facto standard?  The distinction is one I haven't seen before...

-----Original Message-----
From: Ken Paul [mailto:[log in to unmask]]
Sent: Sunday, July 01, 2001 6:36 PM
To: [log in to unmask]
Subject: Re: Fortran66 & 77 HELP (long)


With 17 years of Fortran experience on the HP (e)3000, I have several
opinions on this subject that I would like to share.

Fortran 66

On the HP 3000, this was one of the best compilers and produced some of the
best run time code (along with SPL and BASIC).

The switch from Fortran 66 to Fortran 77 was not trivial and HP's first
attempt with the Fortran 77/V compiler was awful.  It was a slow compiler
which produced slow and large code (usually taking Fortran 66 code and
creating Stack Overflow problems on the Classic machines once/if the compile
ever finished.

The current version of Fortran 77/iX at least gives us Native Mode code and
I generally use the following compiler directives to go from Fortran 66 on
the HP 3000 to Fortran 77/iX on the HP e3000.

$STANDARD_LEVEL SYSTEM
$SHORT
$OPTIMIZE ON

The only bad things about the Fortran 77 compiler is that it uses relative
record numbers to report errors instead of actual line numbers so you are
constantly forced to renumber your source.  It also has a bug in the turning
off of Integer Overflow checking.  When I used the following compiler
directives:

$CHECK_OVERFLOW INTEGER_2 OFF
$CHECK_OVERFLOW INTEGER_4 OFF

the compiler was still generating Overflow checking code so I had to add:

$CHECK_OVERFLOW INTEGER OFF

which is redundant but had the desired affect.

<PLUG>

If anyone is looking for a Fortran 66 or Fortran 77 programmer on the HP
e3000, I am available!

<\PLUG>

Thanks for your time,

Ken

Ken Paul
Independent Consultant
[log in to unmask]
(303) 694-0920

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2