From xemacs-m  Tue May 20 20:34:24 1997
Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1])
	by xemacs.org (8.8.5/8.8.5) with SMTP id UAA17295
	for <xemacs-beta@xemacs.org>; Tue, 20 May 1997 20:34:24 -0500 (CDT)
Received: from Eng.Sun.COM ([129.146.1.25]) by mercury.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id SAA17162; Tue, 20 May 1997 18:49:00 -0700
Received: from kindra.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3)
	id SAA06910; Tue, 20 May 1997 18:33:46 -0700
Received: from xemacs.eng.sun.com by kindra.eng.sun.com (SMI-8.6/SMI-SVR4)
	id SAA28765; Tue, 20 May 1997 18:33:45 -0700
Received: by xemacs.eng.sun.com (SMI-8.6/SMI-SVR4)
	id SAA14245; Tue, 20 May 1997 18:33:43 -0700
Date: Tue, 20 May 1997 18:33:43 -0700
Message-Id: <199705210133.SAA14245@xemacs.eng.sun.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Martin Buchholz <mrb@Eng.Sun.COM>
To: carney@gvc.dec.com
Cc: xemacs-beta@xemacs.org
Subject: RE: XEmacs 20.3-b1 build mostly successful on Digital UNIX V4.0B
In-Reply-To: <9705191433.AA13769@fornet.gvc.dec.com>
References: <9705191113.AA18060@fornet.gvc.dec.com>
	<199705191144.EAA12098@xemacs.eng.sun.com>
	<9705191433.AA13769@fornet.gvc.dec.com>
X-Mailer: VM 6.31 under 20.2 XEmacs Lucid
Reply-To: Martin Buchholz <mrb@Eng.Sun.COM>

>>>>> "Steve" == Steve Carney <carney@gvc.dec.com> writes:

Steve> mrb@Eng.Sun.COM (Martin Buchholz) writes,
Steve> in <199705191144.EAA12098@xemacs.eng.sun.com>:

>> LIBS_TERMCAP does not seem to be defined by your s&m files.  

Steve> The problem turns out to be that configure sets libs_termcap to " " (via
Steve> sed at line 2018) if LIBS_TERMCAP is not defined.  Consequently, on
Steve> configure line 8366 the "test -n" always succeeds, which prevents adding
Steve> -lcurses or -ltermcap to LIBS.  Either, sed should set libs_termcap to
Steve> "", or the "test -n" should become whitespace aware.

Thank you for the very helpful investigation.  The next version of
configure.in will (try to) strip out any trailing blanks found in libs_termcap.

Steve> In configure lines 8361-8363, -lcurses is checked before -ltermcap.  In
Steve> lines 8367-8368, however, -ltermcap is checked before -lcurses.  Is
Steve> there a reason for switching the library search order?  If -lcurses is
Steve> checked first in the else clause, then I won't have to define
Steve> LIBS_TERMCAP in a system configuration file.

I think you are correct.  Is there some standard that defines the
names of the standard libraries and what functions they should define?
XPG4 only seems to talk about header files.  On Solaris, libcurses,
libtermlib, and libtermcap all exist and are in fact identical.  They
were hard to find, carefully hidden in /usr/ccs/lib.  I added the test
for -lcurses and -ltermcap precisely so that most systems would not
have to define LIBS_TERMCAP.  The next release will have the libraries 
in your suggested order.

Martin

