On Tue, Mar 26, 2002 at 05:47:11PM -0800, Paul Eggert wrote:
> > From: Bruce Korb <[EMAIL PROTECTED]>
> > Date: Tue, 26 Mar 2002 14:26:42 -0800
> >
> > #ifndef FUNC_BROKEN
> > # define posix_func func
> > #else
> > # define posix_func autoconf_rpl_func
> > #endif
>
> [...]
>
> I think Autoconf would do well to follow in that tradition. In the
> end it is probably the only one that will actually work portably, for
> all POSIX functions.
>
> I hope I never have to use it myself, though, as it makes the code
> painful to maintain. For example, when importing code from other
> packages I would have to substitute "ac_open" for each instance of
> "open".
This would be really useful. But I think it would be better done
as a separate project, rather than within Autoconf. Partly for
the reason that Paul gives -- not to end up, even if
inadvertently, making it difficult for Autoconf users to avoid
it. But mostly for philosophical cleanliness. Autoconf should
continue to focus on what it does now, and does well -- detecting
and reporting system differences. This hypothetical new package
can then worry about the encapsulation layer that hides them from
the application programmer.
--
| | /\
|-_|/ > Eric Siegerman, Toronto, Ont. [EMAIL PROTECTED]
| | /
"Outlook not so good." That magic 8-ball knows everything!
I'll ask about Exchange Server next.
- Anonymous
