Mailing-List: contact dev-help@apr.apache.org; run by ezmlm
Precedence: bulk
Message-ID: <01fa01c32c4f$53fa0a80$78d96f83@pomello>
From: "Max Bowsher" <maxb@ukf.net>
To: <dev@apr.apache.org>
References: <019401c32c4c$86d04a80$78d96f83@pomello>
 <0.1054919470@scotch.ics.uci.edu>
Subject: Re: Retrying: [Patch] Add fallback method to libtool.m4 location
 algorithm
Date: Fri, 6 Jun 2003 18:16:11 +0100
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

Justin Erenkrantz wrote:
> --On Friday, June 06, 2003 17:56:08 +0100 Max Bowsher <maxb@ukf.net>
wrote:
>
>> Reposting - What do people think about this?
>
> I'm leery of using automake in any fashion, so I'm not overly enthusiastic
> about using aclocal even in such limited fashion.

Why? This change does not introduce a dependency - it merely takes advantage
of it if it needs to and it is available.

> Yet, I should note that you should probably test for aclocal first before
> you use it.  See how we search for libtoolize with PrintPath.  That'd be a
> better solution than just assuming aclocal is present before executing it.
> Otherwise, people will just see 'aclocal is not found' and then 'aclocal
is
> not available' - that's not very clean behavior.

OK, I will clean this up.

> BTW, what lame-brain OSes require this?  -- justin

Ones which use clever scripts to permit transparent usage of multiple
autotool versions.

In my case, Cygwin.


Max.