Re: Powerloom, a first stab... [somewhat OT]

Gerard Toonstra <[email protected]> Tue, 30 Mar 2010 09:06:14 +0000
Newsgroups gmane.comp.ai.powerloom
Message-ID <[email protected]>
--===============1099455658==
Content-Type: multipart/alternative;
	boundary="_91fe2293-fb5f-42d8-9edc-23df372bb21b_"

--_91fe2293-fb5f-42d8-9edc-23df372bb21b_
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable



Thanks very much for the replies and insight.

I think that at least for the moment=2C using SHOP2 together with powerloom=
 would be
a more complicated route and possibly overkill. However=2C a generic proble=
m solver=20
would be an ideal candidate for the combination. In my case=2C the problem =
is well defined=2C=20
needs data from the environment and I feel that I have more control over th=
e algorithm=20
when done in native code.

Best rgds=2C

Gerard

> Date: Fri=2C 26 Mar 2010 11:22:12 -0500
> From: [email protected]
> To: [email protected]
> CC: [email protected]=3B [email protected]
> Subject: Re: [PowerLoom Forum] Powerloom=2C a first stab... [somewhat OT]
>=20
> A quick follow-up=2C based on Thomas Russ's posting:
>=20
> The Common Lisp version of SHOP2 has a very open API for integration of
> outside solvers such as Common Lisp.  It's relatively easy to write Horn
> clauses that will execute (pretty) arbitrary code (although there  are
> some pitfalls).
>=20
> So it might be possible to use Powerloom as Thomas suggests.  A caution:
>  SHOP2=2C like most planners=2C will backtrack during its search=3B a
> discipline would have to be developed for having backtracking in the
> planner have corresponding undo effects in Powerloom.
>=20
> best=2C
> r
 		 	   		 =20
_________________________________________________________________
Hotmail: Trusted email with powerful SPAM protection.
http://clk.atdmt.com/GBL/go/210850553/direct/01/=

--_91fe2293-fb5f-42d8-9edc-23df372bb21b_
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<html>
<head>
<style><!--
.hmmessage P
{
margin:0px=3B
padding:0px
}
body.hmmessage
{
font-size: 10pt=3B
font-family:Verdana
}
--></style>
</head>
<body class=3D'hmmessage'>
<br>Thanks very much for the replies and insight.<br><br>I think that at le=
ast for the moment=2C using SHOP2 together with powerloom would be<br>a mor=
e complicated route and possibly overkill. However=2C a generic problem sol=
ver <br>would be an ideal candidate for the combination. In my case=2C the =
problem is well defined=2C <br>needs data from the environment and I feel t=
hat I have more control over the algorithm <br>when done in native code.<br=
><br>Best rgds=2C<br><br>Gerard<br><br>&gt=3B Date: Fri=2C 26 Mar 2010 11:2=
2:12 -0500<br>&gt=3B From: [email protected]<br>&gt=3B To: rpgoldman@sift=
.info<br>&gt=3B CC: [email protected]=3B [email protected]<br>&gt=
=3B Subject: Re: [PowerLoom Forum] Powerloom=2C a first stab... [somewhat O=
T]<br>&gt=3B <br>&gt=3B A quick follow-up=2C based on Thomas Russ's posting=
:<br>&gt=3B <br>&gt=3B The Common Lisp version of SHOP2 has a very open API=
 for integration of<br>&gt=3B outside solvers such as Common Lisp.  It's re=
latively easy to write Horn<br>&gt=3B clauses that will execute (pretty) ar=
bitrary code (although there  are<br>&gt=3B some pitfalls).<br>&gt=3B <br>&=
gt=3B So it might be possible to use Powerloom as Thomas suggests.  A cauti=
on:<br>&gt=3B  SHOP2=2C like most planners=2C will backtrack during its sea=
rch=3B a<br>&gt=3B discipline would have to be developed for having backtra=
cking in the<br>&gt=3B planner have corresponding undo effects in Powerloom=
.<br>&gt=3B <br>&gt=3B best=2C<br>&gt=3B r<br> 		 	   		  <br /><hr />Hotma=
il: Trusted email with powerful SPAM protection. <a href=3D'http://clk.atdm=
t.com/GBL/go/210850553/direct/01/' target=3D'_new'>Sign up now.</a></body>
</html>=

--_91fe2293-fb5f-42d8-9edc-23df372bb21b_--

--===============1099455658==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
powerloom-forum mailing list
[email protected]
http://mailman.isi.edu/mailman/listinfo/powerloom-forum

--===============1099455658==--