From: gsw AT aloft DOT att DOT com
Date: Wed, 30 Jun 93 14:20:12 EDT
Original-From: aloft!gsw (gerald.s.williams)
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Subject: Re: The continuing saga of getcwd()

Jeremy Mathers <pynq AT midway DOT uchicago DOT edu> writes:

> [ ... ] I think the default names ought to get functions with Unix-like
> behavior.  Versions that give behavior unique to DOS ought to have the
> _dos_ prefix. [ ... ]
> 
> Thus, if you want a version of getcwd() that returns a drive letter, you
> are just going to have to call _dos_getcwd().

One could argue that if you want getcwd() to be more UNIX-like, it SHOULD
include the drive letter.  Under UNIX, getcwd() always returns the FULL
path.  Under DOS, the FULL path must include a drive letter.

My point is this: there are some functions that SHOULD work differently
under DOS.  These should be changed in libc.a.  The reason for libunix.a
is to allow people to get the UNIX implementation in those cases where
either might be needed (and only those cases).

Making getcwd() return a path without a drive letter by default does not
necessarily facilitate ports from UNIX.  Drive letters still must be
dealt with.  It does, however, impede ports from other DOS compilers.

Face it, if you are compiling for DOS, you have to deal with DOS, whether
you like it or not.  And like it or not, DJGPP is a DOS compiler.

It was me who suggested libunix.a, by the way.

,-----------------.
|Gerald S Williams|
|gsw AT aloft DOT att DOT com|
|  (215)439-7237  |
`-----------------'

      _  |     ____/    _  |
     /   /    /        /   /      /                         /
    /   /  ____ |     ____/   _  _/     __ |     __ |      /       ____/
   /   /        /    /          /      /   /    /   /     /     ____
______/  ______/  __/         ___/   _____/   _____/    ___/   ______/