Re: autoconf

[Ben Laurie <be...@gonzo.ben.algroup.co.uk>]

Dean Gaudet wrote:
> 
> On Fri, 27 Jun 1997, Roy T. Fielding wrote:
> > +1 on using autoconf for 2.0.
> 
> I'd really like to re-open the autoconf can of worms and find out why
> there is a dislike of it.  I've only had successes with it when I use it,
> my program worked on platforms I hadn't seen without any changes.  I just
> followed the autoconf docs and used every macro they suggested.  So you
> can count me as +1 for autoconf. 

I'll restate my objections to autoconf:

1. What we've got mostly works (this is a new objection ;-)

2. Changing what autoconf does requires an understanding of the gadget that
generates the scripts, whereas changing what we do only requires understanding
shell and .h - far more common skills.

3. Failing 2, it is next to impossible to fix autoconf scripts when broken.

4. On minority OSes, autoconf scripts are often severely broken.

Cheers,

Ben.

-- 
Ben Laurie                Phone: +44 (181) 994 6435  Email: ben@algroup.co.uk
Freelance Consultant and  Fax:   +44 (181) 994 6472
Technical Director        URL: http://www.algroup.co.uk/Apache-SSL
A.L. Digital Ltd,         Apache Group member (http://www.apache.org)
London, England.          Apache-SSL author

Re: autoconf

[Dean Gaudet <dg...@arctic.org>]

On Fri, 27 Jun 1997, Ben Laurie wrote:
> 2. Changing what autoconf does requires an understanding of the gadget that
> generates the scripts, whereas changing what we do only requires understanding
> shell and .h - far more common skills.
> 
> 3. Failing 2, it is next to impossible to fix autoconf scripts when broken.

They're really the same point.  But I can't see how it's much harder than
doing this:

cp config.h.in config.h
vi config.h
cp Makefile.in Makefile
vi Makefile

Which is about as hard as:

less PORTING
vi helpers/GuessOS
vi Configure
vi conf.h

Dean