Termination stuff

Fergus Henderson fjh at cs.mu.oz.au
Sat Oct 11 22:36:55 AEST 1997


Christopher Rodd SPEIRS, you wrote:
> 	Back to the topic at hand.  Firstly, in the NEWS file, there is
> an XXX which will need to be replaced with the final resting place of
> lp_solve.  Could you please do this when you decide where lp_solve will
> be.

OK, I'll do that.

I've put lp_solve in the top-level directory for the moment;
we should probably move it to the `extras' directory, before
the next major release, but hopefully we can get rid of the
dependency on it altogether before then.

> Secondly, I have not documented the MERCURY_LPSOLVE environment
> variable anywhere.  I was not sure where the best place to document it
> was.

In the "Environment variables" section of the Mercury User's Guide.

But in fact it is not really necessary to document this.
It is an implementation detail that users don't really need to know about.
I've made the installation install `lp_solve' in the Mercury bin
directory, and the installation instructions say that users should
put this in their path.  That should be sufficient.

So, please document it, but I'm not going to hold up the release for that.

> The only other thing, was that I noticed that there is now an
> lp_solve 2.2, and maybe we should say somewhere that we have only tested
> our system with version 2.0.

The fact that we use lp_solve at all is an implementation detail.

The only potential problem is if the users have another executable
named `lp_solve' in their path.  The right fix to that is for
the `mmc' script to set the MERCURY_LPSOLVE environment variable,
if not already set, to @BINDIR@/lp_solve.


scripts/mmc.in:
	Set the MERCURY_LPSOLVE environment variable to the full path
	"@BINDIR@/lp_solve", if not already set, so that it won't matter
	if users have another command called `lp_solve' in their path.

Index: mmc.in
===================================================================
RCS file: /home/staff/zs/imp/mercury/scripts/mmc.in,v
retrieving revision 1.2
diff -u -u -r1.2 mmc.in
--- mmc.in	1997/07/27 15:09:37	1.2
+++ mmc.in	1997/10/11 12:32:21
@@ -29,6 +29,8 @@
 HAVE_DELAY_SLOT=@HAVE_DELAY_SLOT@
 HAVE_BOXED_FLOATS=@HAVE_BOXED_FLOATS@
 DEFAULT_OPT_LEVEL=${MERCURY_DEFAULT_OPT_LEVEL="-O2"}
+MERCURY_LPSOLVE=${MERCURY_LPSOLVE=@BINDIR@/lp_solve}
+export MERCURY_LPSOLVE
 
 # The default optimization level should be after
 # all the options that describe the machine configuration.

-- 
Fergus Henderson <fjh at cs.mu.oz.au>   |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>   |  of excellence is a lethal habit"
PGP: finger fjh at 128.250.37.3         |     -- the last words of T. S. Garp.



More information about the developers mailing list