Re: NSPathUtilities Patch - 5 - NSPathUtilities.m

[Peter Cooper]

On Thu, Apr 22, 2004 at 12:04:15PM -0700, Alex Perez wrote:
> On Thu, 22 Apr 2004, Sheldon Gill wrote:
> 
> > On Wed, 21 Apr 2004 17:56, Kazunobu Kuriyama wrote:
> > > Sheldon Gill wrote:
> > > IIRC, some people were not in favor of setting
> > > GNUSTEP_CONFIGURATION_FILE to
> > > the fixed path /etc/GNUstep/GNUstep.conf.  Did you reached some
> > > agreement with
> > > them?
> >
> > The definition of the fixed path is to be a compile time configuration 
> > option.
> 
> This is wrong, wrong, wrong, wrong. On BSD (ie Non-Linux) systems, this
> is absolutely, categorically, without a doubt the incorrect place to put this
> file. BSD's have /usr/local/etc and /usr/etc for a reason. Just because
> 99% of Linux distros don't use this models does not mean it shouldn't be
> used when the mechanism exists. GNUstep is not Linux specific and we
> should not be so presumptuous as to try and make it so. This patch should
> not be hard coded, and should be something that is determined by the
> configure script based on the OS.

And of course, Alex, in terms of *BSD, not all the world's FreeBSD ;-)

On NetBSD the magical path string is /usr/pkg/etc.

I believe that the hardcoded path should be defined at compile-time
for each platform.

This is even fairly easy to do with some nested #if's although I am a
bit concerned because I personally don't necessarily install things in
standard locations when doing test installs of new versions. Therefore
a "one-size-fits-all" approach even on one platform isn't necessarily
enough. So a well-defined and documented environment variable (and
equivalent for platforms that don't use them :-) override mechanism
is probably unavoidable.

I'd like to commend Sheldon on the hard work he's been doing. I see
definite merit in this approach, and I'd suggest that people with the
most vehement concerns consider providing him with constructive ways
to address those issues.

Regards

Peter