From windley@cheetah  Sat Jan 13 11:49:41 1990
Received: by iris.ucdavis.edu (5.57/UCD.EECS.2.0)
        id AA23971; Sat, 13 Jan 90 11:49:41 PST
Received: from cheetah.ucdavis.edu by clover.ucdavis.edu (5.59/UCD.EECS.1.11)
        id AA27202; Sat, 13 Jan 90 11:54:29 PST
Received: by cheetah.ucdavis.edu (AIX  2.1.2/3.14)
        id AA03356; Sat, 13 Jan 90 11:49:28 PST
Message-Id: <9001131949.AA03356@cheetah.ucdavis.edu>
To: info-hol@clover
Subject: Ultrix and make
Date: Sat, 13 Jan 90 11:49:27 -0800
From: Phil Windley <windley@cheetah>


This message concerns a bug in the Ultrix OS on VAXES that affects the
making of HOL.  If you aren't planning on making HOL on a VAX running
Ultrix (Ultrix on the Decstation is ok), you can safely ignore this
message.  If you are, this message will tell you how to work around the bug
until we can fix Ultrix ;-).

Typing "make hol" on a VAX running Ultrix results in an error when the
first lisp file is compiled.  The problem is not in the makefile
distributed with HOL, sh, or even make.  The problem is somewhere in the
bowels of Ultrix. As near as we can tell, the problem has to do with using
pipes inside an if_then_else statement inside a makefile (pretty obscure,
huh?).  If the problem is in make, it is found in at least two very
different versions of make.  I suspect it is in a system call somewhere.
All is not lost, however; there is an easy workaround.

Rather than typing "make hol", type "make -n hol >makehol.script".  This
will produce a script for sh that will make hol successfully.  To use it,
simply type "sh makehol.script".  If necessary, the script can be edited
(if you're brave), but I'd suggest editing the makefile if necessary and
then remaking the script from scratch.

Cheers,

--phil--


