Re: bugs with squash ?

From: Warwick Harvey <wh_at_icparc.ic.ac.uk>
Date: Tue 22 Jan 2002 04:03:52 PM GMT
Message-ID: <20020122160351.A29007@tempest.icparc.ic.ac.uk> 
Hi Eric,

On Tue, Jan 15, 2002 at 03:44:42PM +0100, Eric.Bensana wrote:
> But i got a strange behavior with squash, which were not present in the
> 5.2 version

[...]

> But the same with squash with log does not answer
> ria:(A::0.0..3.0), ria:(C::0.5..0.65), ria:(C =:= cos(A)), squash([A], 0.01, 
> log).

Thanks for spotting this.  In prior versions of ECLiPSe, there were problems
with locate (and squash?) when using log splitting with an interval spanning
0, so we rewrote the log splitting code.  Apparently we didn't get it quite
right.  :(

I'm currently trying to decide the right way to fix the problem; in the mean
time, can you tell me which architecture(s) you're using, so I can send you
fixed versions of the libraries when it's done?  (It may be a while before a
new build including the fix appears on the FTP site.)

Cheers,
Warwick
Received on Tue Jan 22 16:03:53 2002

This archive was generated by hypermail 2.1.8 : Wed 16 Nov 2005 06:08:12 PM GMT GMT