Date: Mon, 28 Aug 1995 09:31:48 +0100
From: Thomas Eifert <EIFERT AT reze-1 DOT rz DOT rwth-aachen DOT de>
Subject: Re: problems with F2C libraries
To: lrobins AT enh DOT nist DOT gov (Lawrence H. Robins), djgpp AT sun DOT soe DOT clarkson DOT edu
Reply-To: eifert AT rz DOT rwth-aachen DOT de
Organization: Rechenzentrum RWTH Aachen

> Hi, I have some questions about the F2C program for converting
> Fortran-77 source to C, specifically about the libraries called
> LibF77 and LibI77.  Apologies for posting to this group, because
> I'm not using DJGPP (at least not yet), rather another C compiler

at least, f2c works very good together with djgpp, so You should
consider a change ... ;-)

> for MS-DOS, but I thought someone here might be able to help.
> 
> Anyway, I obtained the following files
> from the NETLIB math software repository:
> (1) MS-DOS executable version of F2C itself
> (2) Second executable version of F2C, uses extended memory
> (3) Header file f2c.h
> (4) Two libraries named LibF77 and LibI77
> My questions are about the libraries, which according to the
> documentation contain support routines for Fortran
> intrinsic functions and I/O:
> 
> (1) The LibF77 and LibI77 files from NETLIB aren't plain
>     source code.  They begin
> 
> >   # to unbundle, sh this file (in an empty directory) 
> 
> [What's "sh"?]
> and continue (for example)
> 

These files are "shell archives" : bourne shell scripts which "unbundle"
when interpreted by "sh" - I have some doubts whether You'll get happy 
with them. 
Perhaps You should get "ordinary" zip-files ?

> 
> How can I "unbundle" or otherwise use these files?

s.a. 

> 
> (2) Realize you might not know the answer to this, but if the
> library files were "unbundled", would I be able to compile
> them, and then to compile and run F2C'ed Fortran programs,
> using some random C compiler unknown to this group
> (e.g. my copy of Borland Turbo C++ 3.0)?

Yup

> 
> Thanks for any help with my confusion!
> 
> 
Thomas Eifert