autoconf
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: LIBOBJS and ANSI2KNR fix?


From: Harlan Stenn
Subject: Re: LIBOBJS and ANSI2KNR fix?
Date: Mon, 25 Mar 2002 01:31:43 -0500
User-agent: EMH/1.10.0 SEMI/1.13.7 (Awazu) FLIM/1.13.2 (Kasanui) XEmacs/21.1 (patch 14) (Cuyahoga Valley) (i386--freebsd)

> > Date: Mon, 25 Mar 2002 00:17:45 -0500
> > From: Harlan Stenn <address@hidden>
> > 
> > The NTP software currently requires 2.49,
> 
> That's too bad.  2.49 was never an official release.

(I should have said 2.49 or later.)

Yes, but it worked, and NTP used to require 2.13 with patches to both
automake and autoconf (included with the NTP release) until we started
getting bitten by bugs that were only fixed in the CVS releases.

It would be no trouble for me to bump the requirement to 2.50; I would just
prefer to know that 2.50 works and not really require 2.52 or later.

> I suppose you can use AC_LIBOBJ if it's defined, and if not, you can
> assign to LIBOBJ indirectly, in a way that Autoconf can't tell about.
> E.g. "eval 'LIB''OBJ'='value'".  It'd be ugly, but it'd work, I think.
> But you're on your own if you play tricks like that....
> 
> Personally, I would just bump the version of Autoconf to 2.53.  If you
> want to develop software, you gotta keep your tools sharp.  It's
> unwise for NTP to require developers to use an unreleased, unsupported
> version of Autoconf.

I don't (think I) require unreleased/unsupported versions of autoconf - the
last time I checked we worked OK with 2.49 and most folks are now using
2.52.  I'm trying to be liberal in what I accept.

Requiring 2.53 is fine with me, but the Linux camps screamed the loudest
about 2.13 not working even after 2.50 had been out for months.

H



reply via email to

[Prev in Thread] Current Thread [Next in Thread]