Re: final install problems

[Phil Edwards <pedwards at disaster dot jaj dot com>]

On Thu, Nov 01, 2001 at 08:31:40AM -0800, Joe Buck wrote:
> > >>>>> Joe Buck writes:
> > 
> > Joe> Then why does configure decide to use it?  I recall many pieces of GNU
> > Joe> software had a configure that found an "install" program on my Solaris
> > Joe> system, tested it, didn't like it, and used a shell script included
> > Joe> with the distribution instead.
> > 
> > 	If you look at configure, it specifically avoids "install" on
> > particular systems even when it finds it.  It does not test if "install"
> > actually works, it simply is hardwired to avoid some systems.
> 
> I may have been hallucinating, but I could have sworn that I recalled
> messages implying that a test was being performed.

Perhaps I had the same hallucination; I could swear the same.

> Perhaps the sense of the test should be reversed: a known-good list
> instead of a known-bad list.  But even on a known-good system, there
> might be a different "install" on the user's path hiding the real
> "install".

For known-good systems, we could just use a hardcoded path.  Then the
test would merely be "does this file exist?"  and if it doesn't, just use
the script.

This might allow some of the known-bad systems to move onto the known-good
list.  Solaris, for example, has some bizarre funky crack-monkey install(1)
in /usr/bin, but the "correct" one in /usr/ucb.


Phil

-- 
If ye love wealth greater than liberty, the tranquility of servitude greater
than the animating contest for freedom, go home and leave us in peace.  We seek
not your counsel, nor your arms.  Crouch down and lick the hand that feeds you;
and may posterity forget that ye were our countrymen.            - Samuel Adams