Re: enable-foo vs. with-foo is illogical
On 9/28/05, Olivier Galibert <galibert(a)pobox.com> wrote:
On Wed, Sep 28, 2005 at 01:52:31AM -0500, Ben Wing wrote:
> I'm a little confused. What is the objection to just using `with-foo'
> always, and never using `enable-foo' at all, no aliasing provided?
1- I'd find annoying to have to change my configure-calling scripts
_again_ (but I'll survive)
2- I really, really don't want us to have to go through that crap
again in 6-12 months. Always with-foo was the situation before
whoever changed it. So it must have not been the right thing for
some (unfathomable to me) reason.
OG.
While I agree with Ben that the current situation is confusing, I
don't think changing all of the --enables to --withs buys us very
much. It simplifies the situation at the expense of backwards
compatibility.
I think the autoconf distinction is, at best, confusing and in the
case of XEmacs, the distinctions don't make much sense. We are
clearly beyond the pale as far as number and types of options with
respect to autoconf.
We also have some history to consider: I, too, like Olivier, would not
want to change my scripts again.
Therefore, if we're going to make any changes, I think the only change
that increases understanding while respecting history is to create an
alias between --with and --enable.
My $.02.
- Vin
--
Whoever you are, no matter how lonely,
the world offers itself to your imagination,
calls to you like the wild geese, harsh and exciting--
over and over announcing your place
in the family of things. Mary Oliver