From owner-cougaar-developers@cougaar.org Wed Jan 31 01:43:00 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id BAA08453
for cougaar-developers-outgoing; Wed, 31 Jan 2001 01:43:00 -0500
Received: from x_serv_2.dsta.gov.sg ([202.42.198.37])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id BAA08450
for ; Wed, 31 Jan 2001 01:42:57 -0500
Received: from xpress.dsta.gov.sg (XPRESS [202.42.198.52]) by x_serv_2.dsta.gov.sg with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2650.21)
id DDZR1FYA; Wed, 31 Jan 2001 14:42:17 +0800
Received: by XPRESS with Internet Mail Service (5.5.2650.21)
id ; Wed, 31 Jan 2001 14:29:19 +0800
Message-ID: <9F70B25D4804D411ACF300508B08F126142D85@XPRESS>
From: Lee Chew Hung
To: "'cougaar-developers@cougaar.org'"
Subject: Standards
Date: Wed, 31 Jan 2001 14:29:18 +0800
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
boundary="----_=_NextPart_001_01C08B4F.23DCB81E"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_001_01C08B4F.23DCB81E
Content-Type: text/plain
Hello,
Does anyone know if cougaar follows any standards like the FIPA
specifications? Are there plans for cougaar to conform to any standards?
Thanks and regards.
------_=_NextPart_001_01C08B4F.23DCB81E
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable
Standards
Hello,
Does anyone know if cougaar follows any standards =
like the FIPA specifications? Are there plans for cougaar to =
conform to any standards?
Thanks and regards.
------_=_NextPart_001_01C08B4F.23DCB81E--
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Wed Jan 31 18:47:59 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id SAA09060
for cougaar-developers-outgoing; Wed, 31 Jan 2001 18:47:59 -0500
Received: from mailgate1.darpa.mil (mailgate1.darpa.mil [192.5.18.61])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id SAA09057
for ; Wed, 31 Jan 2001 18:47:57 -0500
Received: from mailhub1.darpa.mil (mailhub1.darpa.mil [158.63.2.41])
by mailgate1.darpa.mil (8.10.1/8.10.1) with ESMTP id f0VNhjY24464;
Wed, 31 Jan 2001 18:43:45 -0500 (EST)
Received: from MSX-SMTP2.darpa.mil (msx-smtp2.darpa.mil [158.63.1.61])
by mailhub1.darpa.mil (8.9.1/8.9.1) with SMTP id SAA22356;
Wed, 31 Jan 2001 18:43:53 -0500 (EST)
Received: from 158.63.1.61 by MSX-SMTP2.darpa.mil (InterScan E-Mail VirusWall NT); Wed, 31 Jan 2001 18:47:25 -0500 (Eastern Standard Time)
Received: by MSX-SMTP2.darpa.mil with Internet Mail Service (5.5.2653.19)
id ; Wed, 31 Jan 2001 18:47:25 -0500
Message-ID:
From: tcarrico
To: Lee Chew Hung
Cc: "'cougaar-developers@cougaar.org'"
Subject: RE: Standards
Date: Wed, 31 Jan 2001 18:47:18 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: multipart/alternative;
boundary="----_=_NextPart_001_01C08BE0.25C27510"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_001_01C08BE0.25C27510
Content-Type: text/plain;
charset="iso-8859-1"
Lee:
Cougaar is not currently FIPA compliant. We have been watching the
evolution of the FIPA standards
and know what it will take to make it compliant, but due to other schedule
constraints and the continued
evolution of the FIPA definition, we have elected to hold off on any such
move for the time being. We
have not set a date, but I would look at Summer 2002 at the earliest. It
would be about a months worth of
work to achieve compliancy according to our estimates.
In addition to tracking FIPA compliancy, we are also monitoring the progress
of the DAML activities
and expect we will incorporate DAML compatibility at that time as well.
Todd
-----Original Message-----
From: Lee Chew Hung [mailto:LEE_Chew_Hung@DSTA.GOV.SG]
Sent: Wednesday, January 31, 2001 1:29 AM
To: 'cougaar-developers@cougaar.org'
Subject: Standards
Hello,
Does anyone know if cougaar follows any standards like the FIPA
specifications? Are there plans for cougaar to conform to any standards?
Thanks and regards.
------_=_NextPart_001_01C08BE0.25C27510
Content-Type: text/html;
charset="iso-8859-1"
Standards
Lee:
Cougaar is
not currently FIPA compliant. We have been watching the evolution of the
FIPA standards
and know what
it will take to make it compliant, but due to other schedule constraints and the
continued
evolution of
the FIPA definition, we have elected to hold off on any such move for the time
being. We
have not set
a date, but I would look at Summer 2002 at the earliest. It would be about
a months worth of
work to
achieve compliancy according to our estimates.
In addition
to tracking FIPA compliancy, we are also monitoring the progress of the DAML
activities
and expect we
will incorporate DAML compatibility at that time as well.
Todd
Hello,
Does anyone know if cougaar follows any standards like the
FIPA specifications? Are there plans for cougaar to conform to any
standards?
Thanks and regards.
------_=_NextPart_001_01C08BE0.25C27510--
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu Feb 8 10:18:36 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id KAA15634
for cougaar-developers-outgoing; Thu, 8 Feb 2001 10:18:36 -0500
Received: from portal.east.saic.com (portal.east.saic.com [198.151.13.15])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with SMTP id KAA15631
for ; Thu, 8 Feb 2001 10:18:35 -0500
Received: from mclmx.saic.com by portal.east.saic.com
via smtpd (for alp-61.alp.isotic.org [4.22.165.61]) with SMTP; 8 Feb 2001 15:18:35 UT
Received: from mcl-its-ieg01.mail.saic.com by mclmx.mail.saic.com; Thu, 8 Feb 2001 10:18:10 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
by mcl-its-ieg01.mail.saic.com (NAVIEG 2.1 bld 63) with SMTP id M2001020810135203005
for ; Thu, 08 Feb 2001 10:13:52 -0500
Received: by mcl-its-exbh01.saic.com with Internet Mail Service (5.5.2650.21)
id <1GC47GDG>; Thu, 8 Feb 2001 10:18:18 -0500
Message-Id:
From: "Coleman, Paul"
To: "'cougaar-developers@cougaar.org'"
Subject: cougaar and Visual J++
Date: Thu, 8 Feb 2001 10:09:39 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
My name is Paul Coleman and I am possibly interested in using COUGAAR with
Microsoft Visual J++. Do you know of any incompatibilities between Visual
J++ / Microsoft JVM and COUGAAR?
Thanks in advance.
- Paul Coleman
colemanp@saic.com
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu Feb 8 10:30:08 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id KAA15649
for cougaar-developers-outgoing; Thu, 8 Feb 2001 10:30:08 -0500
Received: from zima.bbn.com (ZIMA.BBN.COM [128.89.8.110])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id KAA15646
for ; Thu, 8 Feb 2001 10:30:08 -0500
Received: from ncombs (NCOMBS.BBN.COM [128.89.27.170])
by zima.bbn.com (8.9.3/8.9.3) with SMTP id KAA22873;
Thu, 8 Feb 2001 10:30:03 -0500 (EST)
Message-Id: <4.1.20010208102457.00d29c70@zima.bbn.com>
X-Sender: ncombs@zima.bbn.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1
Date: Thu, 08 Feb 2001 10:30:47 -0500
To: "Coleman, Paul" ,
"'cougaar-developers@cougaar.org'"
From: Nathan Combs
Subject: Re: cougaar and Visual J++
In-Reply-To:
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
At 10:09 AM 2/8/01 -0500, Coleman, Paul wrote:
>
>My name is Paul Coleman and I am possibly interested in using COUGAAR with
>Microsoft Visual J++. Do you know of any incompatibilities between Visual
>J++ / Microsoft JVM and COUGAAR?
>
i don't believe microsoft JVM supports more recent JDK versions (e.g. > 1.2)
which would be required.
http://support.microsoft.com/support/kb/articles/Q243/0/22.ASP
> NOTE: The Microsoft virtual machine (Microsoft VM), does not support JDK
versions later than JDK 1.1.4. Packages must be found within or be > compatible
with the JDK version 1.1.4.
Nathan Combs
ncombs@bbn.com BBN Technologies
(617) 873-4495 10 Moulton Street
(617) 873-2794 (fax) Cambridge, MA 02138
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Feb 9 10:49:08 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id KAA16641
for cougaar-developers-outgoing; Fri, 9 Feb 2001 10:49:08 -0500
Received: from david.siemens.de (david.siemens.de [192.35.17.14])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id KAA16638
for ; Fri, 9 Feb 2001 10:48:37 -0500
X-Envelope-Sender-Is: Martin.Schneider@mchp.siemens.de (at relayer david.siemens.de)
Received: from mail3.siemens.de (mail3.siemens.de [139.25.208.14])
by david.siemens.de (8.11.0/8.11.0) with ESMTP id f19FcMu09397
for ; Fri, 9 Feb 2001 16:38:22 +0100 (MET)
Received: from mchp9daa.mch.sbs.de (mchp9daa.mch.sbs.de [139.25.137.99])
by mail3.siemens.de (8.11.1/8.11.1) with ESMTP id f19FcM724962462
for ; Fri, 9 Feb 2001 16:38:22 +0100 (MET)
Received: by mchp9daa.mch.sbs.de with Internet Mail Service (5.5.2448.0)
id <1K5R49AJ>; Fri, 9 Feb 2001 16:39:04 +0100
Message-ID: <12D31B803B18D4119958009027FD42B8148365@mchp952a.mch.sbs.de>
From: Schneider Martin
To: "'cougaar-developers@cougaar.org'"
Subject: Question about Message Transport
Date: Fri, 9 Feb 2001 16:38:57 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
Content-Type: text/plain;
charset="iso-8859-1"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hello there!
I'm a newbie to Cougaar and in the moment I'm reading a lot
and trying to understand it.
Since there are some pages under construction (for instance the search page)
I don't find everything I need to know.
And so I have a question about message transport in CougaarSE:
default is RMIMessageTransport. OK.
What other MessageTransports are already realized and which
will be supported in the future? (e.g. http)
thanks!
Martin Schneider
=================================================
Siemens AG
CT IC 6 Intelligent Autonomous Systems
tel.: +49 89 636-44257 fax.: +49 89 636-41423
e-mail: Martin.Schneider@mchp.siemens.de
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Feb 9 14:33:09 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id OAA16859
for cougaar-developers-outgoing; Fri, 9 Feb 2001 14:33:09 -0500
Received: from mail.graphtronics.net (mail.graphtronics.net [216.28.32.93])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with SMTP id OAA16856
for ; Fri, 9 Feb 2001 14:33:08 -0500
Received: (qmail 5049 invoked from network); 9 Feb 2001 19:09:17 -0000
Received: from 209-164-211-83.telares.com (HELO regulusnt) (@209.164.211.83)
by mail.graphtronics.net with SMTP; 9 Feb 2001 19:09:17 -0000
Message-ID: <005401c092cf$1bb92d20$3700000a@clarksweng.com>
From: "Pat Clark"
To:
References: <12D31B803B18D4119958009027FD42B8148365@mchp952a.mch.sbs.de>
Subject: Re: Question about Message Transport
Date: Fri, 9 Feb 2001 14:32:57 -0500
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4522.1200
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Martin,
My name is Pat Clark, I led the team developing the BooksOnLine advanced
tutorial and some military applications of Cougaar.
Cougaar started out using Voyager, a proprietary product, and moved to
RMI in the last two years. There has been some talk of supporting CORBA, and
the architecture leaves a nice slot to slip that in (and pulling RMI out), but I
don't believe anyone's actually implemented it as a replacement transport. RMI
will work transparently for you, as long as you're not required to communicate
bidirectionally with masqueraded IP addresses. As long as all your agent
machines have ip addresses that are accessible "outside" any firewalls, you
don't even see RMI. All the plugin interfaces handle the inter-cluster
communication transparently, you don't have to create remote objects, bind them
to a registry name, or anything like that.
The exception is if you have agents that are not visible to the net on which
other agents communicate with them (i.e. they have private IP addresses, like
10.0.0.1, etc.). If Cluster A has a visible IP and Cluster B is masqueraded,
normal communication (meaning task allocation, result propagation, etc.) can't
occur, because A cannot initiate contact with B when he has something to
communicate. In this case, B must poll A occasionally to see if there is
anything available. He does this by tapping into A's http server, which Cougaar
calls a PlanServerPlugin. Cluster A must have PlanServerProviders (PSPs)
written which B can connect to via http protocol - PSPs are essentially Cougaar
equivalent of servlets which the PlanServerPlugin (aka web server) administers.
PSPs use the port 5555 by default, so B might connect to A.com:5555/Status.PSP,
for example.
If you download the BooksOnLine tutorial, you'll see examples of PSPs that both
do read only actions on another cluster, and ones that send control input in to
another cluster. If you download lesson 4, it'll give examples of all kinds,
including end user PSPs and cluster-to-cluster PSPs (see ui.PSP_BolUI.java for
the former and warehouse.PSP_CatalogSearch.java for the latter).
Pat Clark
Clark Software Engineering, Ltd.
5100 Springfield St. Suite 308
Dayton, OH 45431
937.256.7794
fax: 937.256.7842
patkclark@earthlink.net
pkclark@clarksweng.com
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu Mar 8 08:27:24 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id IAA11079
for cougaar-developers-outgoing; Thu, 8 Mar 2001 08:27:24 -0500
Received: from ss3000e.cselt.it (ss3000e.cselt.it [163.162.41.5])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id IAA11076
for ; Thu, 8 Mar 2001 08:27:13 -0500
Received: from rabadan.cselt.it (rabadan.cselt.it [163.162.4.12])
by ss3000e.cselt.it (PMDF V5.2-31 #43137)
with ESMTP id <0G9V004E1FPRL7@ss3000e.cselt.it> for
cougaar-developers@cougaar.org; Thu, 8 Mar 2001 10:01:03 +0100 (MET)
Received: by rabadan.cselt.it with Internet Mail Service (5.5.2650.21)
id ; Thu, 08 Mar 2001 10:01:26 +0100
Received: from ibm9151.cs.columbia.edu (ibm9151.cselt.it [163.162.96.29])
by EXC2K01.cselt.it with SMTP
(Microsoft Exchange Internet Mail Service Version 5.5.2650.21)
id F3BC0BS3; Thu, 08 Mar 2001 10:01:21 +0100
Content-return: allowed
Date: Thu, 08 Mar 2001 10:01:20 +0100
From: Peppo Valetto
Subject: constraining tasks wrt parent
X-Sender: valetto@canal.psl.cs.columbia.edu (Unverified)
To: cougaar-developers@cougaar.org
Cc: valetto@cs.columbia.edu, ncombs@bbn.com
Message-id: <5.0.2.1.0.20010308092649.02589148@EXC2K01.cselt.it>
MIME-version: 1.0
X-Mailer: QUALCOMM Windows Eudora Version 5.0.2
Content-type: text/plain; charset="us-ascii"; format=flowed
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi all,
as a newbie to COUGAAR programming, I've gone through the basic tutorial
exercises
and I have now a couple of issues for which I need some "higher wisdom"
First of all a preliminary question:
in class SettableConstraintEvent
the method setValue() takes 3 parameters
from javadocs:
setValue(double value, int constraintOrder, double slope)
I do not understand how the "slope" parameter works and javadocs do not
help me much (see below)
slope - specifies the rate at which the score function degrades on the
allowed side of the constraint. The disallowed
side always has a failing score.
Can you shed some light? How does the slope work here?
Second:
I have a trouble in setting correctly a constraint in the followign scenario
Task T has preferences (start=0, end=12)
and expands into T1, T2, T3,
Constraint c1: T1 must end before T2 begins
Constraint c2: T2 must end before T3 begins
Constraint c3: T3 must end before T (its parent) ends (i.e. in my intention
before 12)
The purpose of Constraint c3 is as follows: say, the allocation of T2, T3,
T4 and their duration cause T4 to
end after 12. I would like the expander to reschedule the expansion of T1
among available assets to try to
finish with T1 (and its expansion) before 12. Of course, in case instead
T4 ends before 12, nothing should happen
and everybody should be happy.
This is my code for defining those constraints:
// End(t1) must be before Start(t2)
NewConstraint c1 = theLDMF.newConstraint();
c1.setConstrainingTask(t1);
c1.setConstrainingAspect(AspectType.END_TIME);
c1.setConstrainedTask(t2);
c1.setConstrainedAspect(AspectType.START_TIME);
c1.setConstraintOrder(Constraint.BEFORE);
constraints.addElement(c1);
// End(t2) must be before Start(t3)
NewConstraint c2 = theLDMF.newConstraint();
c2.setConstrainingTask(t2);
c2.setConstrainingAspect(AspectType.END_TIME);
c2.setConstrainedTask(t3);
c2.setConstrainedAspect(AspectType.START_TIME);
c2.setConstraintOrder(Constraint.BEFORE);
constraints.addElement(c2);
// End(t3) must be before End(parent task)
NewConstraint c3 = theLDMF.newConstraint();
c3.setConstrainingTask(new_wf.getParentTask());
c3.setConstrainingAspect(AspectType.END_TIME);
c3.setConstrainedTask(t3);
c3.setConstrainedAspect(AspectType.END_TIME);
c3.setConstraintOrder(Constraint.AFTER);
constraints.addElement(c3);
I had a case running my code that T3 ends at 8 (so should not violate c3)
but it seemingly does
according to COUGAAR since T (its parent) takes the value 8 of T3 as its
own end somehow
(I guess because it is after all the end time of the whole of its expansion)
and therefore I end up with a situation like 8 AFTER 8 which is false.
Can you help?
Peppo
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu Mar 8 14:41:40 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id OAA11402
for cougaar-developers-outgoing; Thu, 8 Mar 2001 14:41:40 -0500
Received: from zima.bbn.com (ZIMA.BBN.COM [128.89.8.110])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id OAA11399
for ; Thu, 8 Mar 2001 14:41:35 -0500
Received: from ncombs (NCOMBS.BBN.COM [128.89.27.170])
by zima.bbn.com (8.9.3/8.9.3) with SMTP id OAA09774;
Thu, 8 Mar 2001 14:41:31 -0500 (EST)
Message-Id: <4.1.20010308141124.00cdbd90@zima.bbn.com>
X-Sender: ncombs@zima.bbn.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1
Date: Thu, 08 Mar 2001 14:42:36 -0500
To: Peppo Valetto
From: Nathan Combs
Subject: Re: threading in plugins
Cc: cougaar-developers@cougaar.org
In-Reply-To: <5.0.2.1.0.20010308180238.02589148@EXC2K01.cselt.it>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
hi peppo --
At 06:24 PM 3/8/01 +0100, Peppo Valetto wrote:
>Hi Nathan,
>I know that, normally, plugins are not allocated a Java thread each, for
>economy reasons I guess,
>but they are scheduled by the PlugInScheduler of the cluster.
>I am starting to write generic plugins that can be worklet-ized on the fly:
>I've figured out a number of things about how to inject worklet code in a
>plugin already.
>However, the main worklet transport and mgmt class which I need to instantiate
>to allow Cougaar plugins to send/receive/execute worklets (named WVM)
>extends Thread.
>In my early experiments I chose to instantiate a WorkletPlugIn as an
>extension of SimplePlugIn and
>the first thing I do in setupSubscriptions() is instantiating and starting
>a WVM
>but when I invoke start() on the WVM I get the following
>
>java.lang.IllegalThreadStateException
> at java.lang.Thread.start(Native Method)
> at coolets.WorkletPlugIn.setupSubscriptions(WorkletPlugIn.java:38)
>
>I strongly suspect this has to do with the threading model of SimplePlugIn
>If so, am I forced to use the more heavyweight SINGLE_THREAD model?
if i understand you -- i think u're right. i don't think you should be
"starting" your WVM Plugin Thread. Load the plugin via Clsuter and let
Clsuter manage start.
>With what drawbacks?
>What is your advice on these threading matters?
>
i believe all simpleplugins are executed round-robin (ala plugin scheduler)
re-using single thread, this is adequate for most cases. its an
efficiency to minimize cluster cpu footprint for most use cases. plugins
which can't live with this threading model are free to run in their own
private thread (as u point out, i think its SINGLE_THREAD). turns out must
plugins can live in shared world well enough can't really think of many
cases where this would be a liability 'cept where perhaps if you had a lot
of plugins and had a steady stream of subsription items which you wanted to
be processed to as close to the rate of publish as possible (vs. having
them bunched up too much). my advice is always to start simple unless u
know its not good enough ;-)
thread model applies to the "execute()" execution of plugin (stimulated by
incoming subscription) and not "background threads" which a plugin may
launch and be running in background.
>Another question is about when I can set or change subscriptions in a plugin.
>setupSubscriptions() is executed once and for all at init time, right?
>What if I later on I use subscribe() and/or unsubscribe()? I mean what is
>the mechanism with which
>a later subscription is accepted and processed in the cluster? When exactly
>does the change
>take effect?
setupsubscription called once. you can dynamically set up subscriptions
subscribe()/ unsubscribe() later, however, as the subscription setup is
fairly expensive -- try to do it in setup once. turns out most plugins
are comfortable with this. ie. they know a priori what they are interested
in subscribing and are interested in it for the duration. an example
exception might be some ui plugins which may need to handle ad hoc queries.
-n.
Nathan Combs
ncombs@bbn.com BBN Technologies
(617) 873-4495 10 Moulton Street
(617) 873-2794 (fax) Cambridge, MA 02138
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Sun Mar 18 22:46:24 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id WAA09417
for cougaar-developers-outgoing; Sun, 18 Mar 2001 22:46:24 -0500
Received: from x_serv_2.dsta.gov.sg ([202.42.198.37])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id WAA09414
for ; Sun, 18 Mar 2001 22:46:14 -0500
Received: from xpress.dsta.gov.sg (mail.dsta.gov.sg [202.42.198.52]) by x_serv_2.dsta.gov.sg with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2650.21)
id FZ9NKPG0; Mon, 19 Mar 2001 11:45:32 +0800
Received: by XPRESS with Internet Mail Service (5.5.2650.21)
id ; Mon, 19 Mar 2001 11:31:15 +0800
Message-ID: <9F70B25D4804D411ACF300508B08F126142D87@XPRESS>
From: Lee Chew Hung
To: "'cougaar-developers@cougaar.org'"
Subject: Sending objects between clusters.
Date: Mon, 19 Mar 2001 11:31:11 +0800
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: multipart/alternative;
boundary="----_=_NextPart_001_01C0B025.0D572A8E"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_001_01C0B025.0D572A8E
Content-Type: text/plain;
charset="iso-8859-1"
Hello,
I'm trying to write a simple cougaar society consisting of two clusters.
How do I send my own object (eg. class Foo) between these two clusters
without making Foo a task or an asset?
Thanks in advance
------_=_NextPart_001_01C0B025.0D572A8E
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sending objects between clusters.
Hello,
I'm trying to write a simple cougaar society =
consisting of two clusters. How do I send my own object (eg. =
class Foo) between these two clusters without making Foo a task or an =
asset?
Thanks in advance
------_=_NextPart_001_01C0B025.0D572A8E--
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Mar 19 10:45:09 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id KAA10045
for cougaar-developers-outgoing; Mon, 19 Mar 2001 10:45:09 -0500
Received: from zima.bbn.com (ZIMA.BBN.COM [128.89.8.110])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id KAA10042
for ; Mon, 19 Mar 2001 10:45:08 -0500
Received: from ncombs (NCOMBS.BBN.COM [128.89.27.170])
by zima.bbn.com (8.9.3/8.9.3) with SMTP id KAA03980;
Mon, 19 Mar 2001 10:44:54 -0500 (EST)
Message-Id: <4.1.20010319102720.00cee380@zima.bbn.com>
X-Sender: ncombs@zima.bbn.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1
Date: Mon, 19 Mar 2001 10:46:12 -0500
To: Lee Chew Hung ,
"'cougaar-developers@cougaar.org'"
From: Nathan Combs
Subject: Re: Sending objects between clusters.
In-Reply-To: <9F70B25D4804D411ACF300508B08F126142D87@XPRESS>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
At 11:31 AM 3/19/01 +0800, Lee Chew Hung wrote:
>
> Hello,
>
> I'm trying to write a simple cougaar society consisting of two clusters. How
> do I send my own object (eg. class Foo) between these two clusters without
> making Foo a task or an asset?
>
> Thanks in advance
you can write a "logic provider" plugin which will route objects published on
Blackboard to remote clusters (ref. PDG for details on logic provider
plugins). If you write/load your own logic providers, may want to provide
your own implementation of:
org.cougaar.domain.planning.ldm.Domain
(again, consult PDG for how to do this).
In terms of a minimalistic LP example, something like below can route a Foo
instance to Cluster named "cid_MyFavoriteCluster"
-n.
public class SendRemoteRequestEnvelopeLogicProvider implements
EnvelopeLogicProvider {
private ALPPlanServesLogicProvider myPlan;
private ClusterServesLogicProvider myCluster;
public SendRemoteRequestEnvelopeLogicProvider(
ALPPlanServesLogicProvider alpplan,
ClusterServesLogicProvider cluster )
{
myPlan = alpplan;
myCluster = cluster;
}
public void execute(EnvelopeTuple m, Collection changeReports)
{
ClusterIdentifier source =
myCluster.getUIDServer().getClusterIdentifier();
ClusterIdentifier dest =
ClusterIdentifier.getClusterIdentifier("cid_MyFavoriteCluster");
// AAIDirective is an implementation of
org.cougaar.domain.planning.ldm.plan.Directive
// straightforward wrapper.
//
AAIDirective aaid = new AAIDirective(
new Foo(), // Send Foo
source,
dest );
myPlan.sendDirective(aaid);
}
}
}
}
}
Nathan Combs
ncombs@bbn.com BBN Technologies
(617) 873-4495 10 Moulton Street
(617) 873-2794 (fax) Cambridge, MA 02138
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu Mar 22 14:45:23 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id OAA02764
for cougaar-developers-outgoing; Thu, 22 Mar 2001 14:45:23 -0500
Received: from zima.bbn.com (ZIMA.BBN.COM [128.89.8.110])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id OAA02761
for ; Thu, 22 Mar 2001 14:45:23 -0500
Received: from ncombs (NCOMBS.BBN.COM [128.89.27.170])
by zima.bbn.com (8.9.3/8.9.3) with SMTP id PAA08262;
Thu, 22 Mar 2001 15:17:38 -0500 (EST)
Message-Id: <4.1.20010322150540.00ce8170@zima.bbn.com>
X-Sender: ncombs@zima.bbn.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1
Date: Thu, 22 Mar 2001 15:17:59 -0500
To: cougaar-developers@cougaar.org
From: Nathan Combs
Subject: Re: Sending objects between clusters.
In-Reply-To: <4.1.20010319102720.00cee380@zima.bbn.com>
References: <9F70B25D4804D411ACF300508B08F126142D87@XPRESS>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
At 10:46 AM 3/19/01 -0500, Nathan Combs wrote:
>
>
>you can write a "logic provider" plugin which will route objects published on
>Blackboard to remote clusters (ref. PDG for details on logic provider
>plugins).
Hi - received some mail on my previous post, some public clarification.
Previous msg presented a basic pattern for augmenting "Cluster messaging
infrastructure" by adding logic providers to send arbitrary blackboard
objects between clusters.
An alternative pattern/possibilty would be to use the alternative "UI
infrastructure" to communicate serialized objects between clusters.
Consult PDG about the HTTP "LogPlanServers" and PSPs (servlets) and how
they work, but the gist is that if you wanted to communicate objects
between clusters WITHOUT modifying the cluster messaging mechanism, you
could write a PlugIn which could open a URL conneciton to a receiving PSP
at a remote cluster and either PUSHED or PULLED serialized objects over
that connection.
In the PULL case you'd need to have the receiving cluster open a connection
back to the origin Cluster (to PSP) and either POLL, or WAIT on a
Keep-Alive connection (see PDG for description of how keep-alives work).
In the PUSH case, the origin Clsuter would open a connection to the target
cluster and POST serialized object into PSP which can then add it to remote
blackboard.
Crudely, the trade-off is likely that the UI infrastructure may be easier
to implmeent (esp if already familiar with writing plugins and psps) but
will entail more connection overhead, and is likely a tad less efficient.
The one trick is that the plugin will need to access the URL of the remote
cluster (given its String name). This can be "looked-up" per snip below.
import org.cougaar.lib.planserver.server.FDSProxy;
import org.cougaar.lib.planserver.server.NameService;
import org.cougaar.lib.planserver.server.ProxyMapAdapter;
FDSProxy fds = new FDSProxy();
NameService myNameServiceProvider = new ProxyMapAdapter( fds );
//
// base_url is URL to cluster LogPlanServer... need to append
your target PSP and path
//
URL base_url = base_url = myNameServiceProvider.lookupURL(name);
-n.
Nathan Combs
ncombs@bbn.com BBN Technologies
(617) 873-4495 10 Moulton Street
(617) 873-2794 (fax) Cambridge, MA 02138
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Mar 26 09:45:01 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id JAA00603
for cougaar-developers-outgoing; Mon, 26 Mar 2001 09:45:01 -0500
Received: from mailgate1.darpa.mil (mailgate1.darpa.mil [192.5.18.61])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id JAA00594;
Mon, 26 Mar 2001 09:44:55 -0500
Received: from mailhub1.darpa.mil (mailhub1.darpa.mil [158.63.2.41])
by mailgate1.darpa.mil (8.10.1/8.10.1) with ESMTP id f2QFFAo21597;
Mon, 26 Mar 2001 10:15:10 -0500 (EST)
Received: from MSX-SMTP2.darpa.mil (msx-smtp2.darpa.mil [158.63.1.61])
by mailhub1.darpa.mil (8.9.1/8.9.1) with SMTP id KAA25771;
Mon, 26 Mar 2001 10:15:37 -0500 (EST)
Received: from 158.63.1.61 by MSX-SMTP2.darpa.mil (InterScan E-Mail VirusWall NT); Mon, 26 Mar 2001 10:19:11 -0500 (Eastern Standard Time)
Received: by MSX-SMTP2.darpa.mil with Internet Mail Service (5.5.2653.19)
id ; Mon, 26 Mar 2001 10:19:11 -0500
Message-ID:
From: tcarrico
To: "'cougaar-information@cougaar.org'" ,
"'cougaar-developers@cougaar.org'"
Subject: FW: Next Cougaar Plug-In Developer training April 10-12
Date: Mon, 26 Mar 2001 10:19:03 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
charset="iso-8859-1"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Reminder:
Cougaar Training News
The next Cougaar Plug-In Developer training course will be held April 10-12,
2001 at the DARPA TIC (Technology Integration Center) in Arlington, VA. Each
session will run from 9:00AM to 5:00PM.
The course is intended to train developers for writing Cougaar PlugIns and
developing Cougaar clusters and societies. The course takes a experiential
approach to training, mixing some presentation material with 'hands-on'
programming work in the lab. We expect that each participant will, on
completion of the training, have written and run a moderately complex
Cougaar society of clusters and PlugIns that represent many of the key
Cougaar architecture concepts and programming constructs.
The following is the URL with directions and more information about the TIC
:
If you have a clearance, please forward your clearance to TIC security so
that you can be unescorted during the course of the training.
The DARPA Director requires all Foreign Nationals visiting DARPA facilities,
which includes the TIC, to complete and submit a DARPA Foreign National
Visit Request Information Form at least seven days before the visit date.
For more information or to receive a copy of this form, please contact
Charlotte Young at either (703) 526-6620 or cyoung@snap.org.
In terms of prerequisites for the course, we expect that participants will
be comfortable with programming and debugging in the Java programming
language, and be familiar with standard Java packages (java.io, java.util,
etc.). Further, we expect that participants will be comfortable with the
Windows NT environment, including executing of batch files and editing text
documents. Finally, and probably most significantly, we expect all
participants in the course to have downloaded and familiarized themselves
with the Cougaar PlugIn Developer's Guide (PDG), which is available at the
Cougaar Open Source WebSite : .
Breakfast and lunch will be provided. A small registration fee will be
collected at the first session. No credit cards will be accepted. The fee
will be determined and everyone who has signed up for the course will be
notified of the amount via email several days in advance of the course.
Please contact Steve Summers (ssummers@bbn.com) or Todd Carrico
(tcarrico@darpa.mil) to register for the Cougaar training session.
We look forward to seeing the participants at the Cougaar Training at the
TIC on April 10 at 9:00 AM.
----------------------
-Michael E. Dyson
-Schafer Corporation
-4310 N. Fairfax Dr.
-Suite 700
-Arlington, VA 22203
-(703) 516-6009
----------------------
-Michael E. Dyson
-Schafer Corporation
-4310 N. Fairfax Dr.
-Suite 700
-Arlington, VA 22203
-(703) 516-6009
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to
majordomo@ultralog.net
with the line "unsubscribe ul-everyone" in the BODY of the message.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Sun Apr 22 20:44:34 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id UAA18583
for cougaar-developers-outgoing; Sun, 22 Apr 2001 20:44:34 -0400
Received: from pisces.starnet.gov.sg ([203.116.59.242])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with SMTP id UAA18580
for ; Sun, 22 Apr 2001 20:44:32 -0400
From: nkiabee@starnet.gov.sg
Received: (from nobody@localhost)
by pisces.starnet.gov.sg (8.10.1/8.10.1) id f3N0o5Q32697;
Mon, 23 Apr 2001 08:50:05 +0800
X-Authentication-Warning: pisces.starnet.gov.sg: nobody set sender to nkiabee@starnet.gov.sg using -f
To:
Subject: Re: How to send messages to other clusters
Message-ID: <987987005.3ae37c3d5692c@www.starnet.gov.sg>
Date: Mon, 23 Apr 2001 08:50:05 +0800
Cc:
References: <200104210519.BAA17232@alp-61.alp.isotic.org>
In-Reply-To: <200104210519.BAA17232@alp-61.alp.isotic.org>
MIME-Version: 1.0
Content-Type: text/plain
Content-Transfer-Encoding: 8bit
User-Agent: StarNet Mail System (SMS)
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi,
I'm new to Cougaar and I have a question regarding sending messages to
other clusters.
Besides sending Task to other clusters. Is there other ways to send
message to other clusters, something like a boardcast/multicast?
Thanks.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Apr 23 07:20:13 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id HAA18867
for cougaar-developers-outgoing; Mon, 23 Apr 2001 07:20:13 -0400
Received: from zima.bbn.com (ZIMA.BBN.COM [128.89.8.110])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id HAA18864
for ; Mon, 23 Apr 2001 07:20:12 -0400
Received: from dslloaner (TC016.BBN.COM [128.33.238.16])
by zima.bbn.com (8.9.3/8.9.3) with SMTP id HAA05437;
Mon, 23 Apr 2001 07:26:21 -0400 (EDT)
Message-Id: <4.1.20010423064307.00be33f0@zima.bbn.com>
X-Sender: ncombs@zima.bbn.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1
Date: Mon, 23 Apr 2001 07:06:28 -0400
To: nkiabee@starnet.gov.sg,
From: nathan combs
Subject: Re: How to send messages to other clusters
Cc:
In-Reply-To: <987987005.3ae37c3d5692c@www.starnet.gov.sg>
References: <200104210519.BAA17232@alp-61.alp.isotic.org>
<200104210519.BAA17232@alp-61.alp.isotic.org>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
At 08:50 AM 04/23/2001 +0800, nkiabee@starnet.gov.sg wrote:
>Hi,
>I'm new to Cougaar and I have a question regarding sending messages to
>other clusters.
>Besides sending Task to other clusters. Is there other ways to send
>message to other clusters, something like a boardcast/multicast?
>
You can send messages via HTTP (from PlugIns to PSPs or vice versa). if
you look at previous email in this archive (go to cougaar website / mail
archive), more detail.
should u need it, u can also use plugins to integrate other transport -
example: experimentation underway using multicast from PlugIn (Sender) to
Plugin (Receiver) for the fan-out (using cougaar normal message channel for
the pt-2-pt return reply). A placeholder shingle can be seen at
http://aai.bbn.com/cougaar/multicast.html
-n
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Apr 23 09:09:07 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id JAA18951
for cougaar-developers-outgoing; Mon, 23 Apr 2001 09:09:07 -0400
Received: from coopers.dsl.net (coopers.dsl.net [65.84.81.5])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id JAA18948
for ; Mon, 23 Apr 2001 09:09:07 -0400
Received: from regulus (65-84-104-66.client.dsl.net [65.84.104.66])
by coopers.dsl.net (Postfix) with SMTP id 6C63C1FD0A
for ; Mon, 23 Apr 2001 09:10:15 -0400 (EDT)
Message-ID: <008a01c0cbf7$9fa07e70$3700000a@regulus>
From: "Pat Clark"
To:
References: <200104210519.BAA17232@alp-61.alp.isotic.org> <987987005.3ae37c3d5692c@www.starnet.gov.sg>
Subject: Re: How to send messages to other clusters
Date: Mon, 23 Apr 2001 09:16:35 -0400
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2919.6700
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi!
Although the nomenclature is not real great (it's not really a TASK as
you'd think of one), you can use a task pretty easily to do what you want.
Make a task, maybe call it MessageTask, you can attach any kind of object
you want as the indirect object of a preposition. Just make sure everyone
you want to broadcast this to will have the class of this type in its class
path, or some clusters won't be able to deserialize it. If you want to have
a broadcast mechanism, you could make an expander that would generate a
bunch of child tasks and allocate them to each of the clusters he knows
about (from his assets list, where asset is instanceof organization and is
not self).
PSP method will work well, just might take you a little longer to do if
you're just starting out with cougaar.
Writing your own logic provider might be the most efficient, but is also
going to be the most painful - you will have to learn a lot about the
nitty-gritty internals of Cougaar, and could make your society not able to
play with others who aren't using those logic providers.
Pat Clark
Clark Software Engineering, Ltd.
5100 Springfield St. Suite 308
Dayton, OH 45431
patkclark@earthlink.net
937.256.7794
Fax: 937.256.7842
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Apr 23 15:45:01 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id PAA19126
for cougaar-developers-outgoing; Mon, 23 Apr 2001 15:45:01 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id PAA19123
for ; Mon, 23 Apr 2001 15:45:00 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA2DF8
for ;
Mon, 23 Apr 2001 12:50:42 -0700
Message-ID: <3AE48791.82E1D17B@trw.com>
Date: Mon, 23 Apr 2001 12:50:41 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: cougaar-developers@cougaar.org
Subject: Possible cluster initialization bug in Cougaar 7.2
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi All,
I seem to have found a scenario where a cluster doesn't get loaded.
On the day this occurred, I was able to repeat this several times.
However, the next day I was not able to replicate the problem, so I'm
not sure how real this problem is. I'd appreciate any comments, help,
confirmation, etc. Here's the scenario:
Consider a single node with 3 clusters, call them cluster A, B, and C. They
are listed in that order in my node initialization file (call it
TestNode.ini). At initialization time (in setupSubscriptions) in cluster B,
the cluster opens a JDBC connection to a remote Oracle database to fetch
some data. Suppose that for some reason, the machine with the Oracle db on
it is down. Eventually the connection attempt times out, and this
java.sql.SQLException is thrown:
"java.sql.SQLException: Io exception: The Network Adapter could not establish
the connection"
After this exception is printed to the console, Cougaar reports that it
has loaded cluster C, and finally reports "Plugins Loaded." Upon opening
TASKS.PSP, cluster C cannot be found in the list of loaded clusters - even
though Cougaar reports that it has been loaded. Cluster B is loaded.
If I rearrange the clusters so that cluster B is *last* in my TestNode.ini
file, then cluster C gets loaded, and cluster B throws the same exception
noted above. Has anyone else ever observed a similar error?
Robert
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Sr. P/A, Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Wed May 2 14:38:21 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id OAA24718
for cougaar-developers-outgoing; Wed, 2 May 2001 14:38:21 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id OAA24715
for ; Wed, 2 May 2001 14:38:20 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA25D
for ;
Wed, 2 May 2001 11:43:49 -0700
Message-ID: <3AF05563.D9165BD6@trw.com>
Date: Wed, 02 May 2001 11:43:47 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: cougaar-developers@cougaar.org
Subject: Predicate question
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi all,
What could make a unary predicate-type class stop matching correctly?
For example, consider this predicate from Cougaar exercise 4:
class myProgrammersPredicate implements UnaryPredicate{
public boolean execute(Object o) {
boolean ret = false;
if (o instanceof Organization) {
Organization org = (Organization)o;
OrganizationPG orgPG = org.getOrganizationPG();
ret = orgPG.inRoles(Role.getRole("SoftwareDevelopment"));
}
return ret;
}
}
Say you've put this class into it's own file and made it a public class
within the same package you're working on. Say also that you're no longer
looking for SoftwareDevelopment organizations that exist within your own
cluster - they're coming from another cluster within the same node. For
some reason, even though the subscription using this predicate is firing,
it is not correctly matching the role of SoftwareDevelopment to the
organization. Any ideas? Thanks :)
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Sr. P/A, Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Wed May 2 15:21:11 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id PAA24749
for cougaar-developers-outgoing; Wed, 2 May 2001 15:21:11 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id PAA24746
for ; Wed, 2 May 2001 15:21:10 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA40DE
for ;
Wed, 2 May 2001 12:26:55 -0700
Message-ID: <3AF05F7E.4D141AFF@trw.com>
Date: Wed, 02 May 2001 12:26:54 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Predicate question, clarified
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi,
Thought I'd make this a bit clearer... what is happening, is that
my subscription to this unary predicate is firing, the unary predicate
is returning true, but when my execute() method fires in the plugin
owning the subscription, the Organization that matched the predicate
isn't in the list of elements (e.g. the java.util.Enumeration class
contained in IncrementalSubscription). Either I'm doing something
terribly wrong, or something is broken. Any Ideas?
-------- Original Message --------
Subject: Predicate question
Date: Wed, 02 May 2001 11:43:47 -0700
From: "Robert Manning"
To: cougaar-developers@cougaar.org
Hi all,
What could make a unary predicate-type class stop matching correctly?
For example, consider this predicate from Cougaar exercise 4:
class myProgrammersPredicate implements UnaryPredicate{
public boolean execute(Object o) {
boolean ret = false;
if (o instanceof Organization) {
Organization org = (Organization)o;
OrganizationPG orgPG = org.getOrganizationPG();
ret = orgPG.inRoles(Role.getRole("SoftwareDevelopment"));
}
return ret;
}
}
Say you've put this class into it's own file and made it a public class
within the same package you're working on. Say also that you're no longer
looking for SoftwareDevelopment organizations that exist within your own
cluster - they're coming from another cluster within the same node. For
some reason, even though the subscription using this predicate is firing,
it is not correctly matching the role of SoftwareDevelopment to the
organization. Any ideas? Thanks :)
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Sr. P/A, Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Wed May 2 18:46:54 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id SAA24847
for cougaar-developers-outgoing; Wed, 2 May 2001 18:46:54 -0400
Received: from po1.bbn.com (PO1.BBN.COM [192.1.50.38])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id SAA24844
for ; Wed, 2 May 2001 18:46:50 -0400
Received: from bbn.com (ros-dun052-048.bbn.com [192.233.52.48])
by po1.bbn.com (8.9.1/8.9.1) with ESMTP id SAA01042;
Wed, 2 May 2001 18:47:48 -0400 (EDT)
Message-ID: <3AF08ECC.E5905B49@bbn.com>
Date: Wed, 02 May 2001 18:48:44 -0400
From: Bill Wright
X-Mailer: Mozilla 4.73 [en] (Windows NT 5.0; I)
X-Accept-Language: en
MIME-Version: 1.0
To: Robert Manning
CC: "cougaar-developers@cougaar.org"
Subject: Re: Predicate question, clarified
References: <3AF05F7E.4D141AFF@trw.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Sometimes you can get this behavior if you predicate tests an attribute that
changes. In your example, it may be that the OrganizationPG changes, making the
subscriptions behave in the way you notice. You might try making your predicate
a DynamicUnaryPredicate instead of a UnaryPredicate and see if it makes a
difference.
Robert Manning wrote:
> Hi,
>
> Thought I'd make this a bit clearer... what is happening, is that
> my subscription to this unary predicate is firing, the unary predicate
> is returning true, but when my execute() method fires in the plugin
> owning the subscription, the Organization that matched the predicate
> isn't in the list of elements (e.g. the java.util.Enumeration class
> contained in IncrementalSubscription). Either I'm doing something
> terribly wrong, or something is broken. Any Ideas?
>
> -------- Original Message --------
> Subject: Predicate question
> Date: Wed, 02 May 2001 11:43:47 -0700
> From: "Robert Manning"
> To: cougaar-developers@cougaar.org
>
> Hi all,
>
> What could make a unary predicate-type class stop matching correctly?
> For example, consider this predicate from Cougaar exercise 4:
>
> class myProgrammersPredicate implements UnaryPredicate{
> public boolean execute(Object o) {
> boolean ret = false;
> if (o instanceof Organization) {
> Organization org = (Organization)o;
> OrganizationPG orgPG = org.getOrganizationPG();
> ret = orgPG.inRoles(Role.getRole("SoftwareDevelopment"));
> }
> return ret;
> }
> }
>
> Say you've put this class into it's own file and made it a public class
> within the same package you're working on. Say also that you're no longer
> looking for SoftwareDevelopment organizations that exist within your own
> cluster - they're coming from another cluster within the same node. For
> some reason, even though the subscription using this predicate is firing,
> it is not correctly matching the role of SoftwareDevelopment to the
> organization. Any ideas? Thanks :)
>
> --
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> Robert Manning TRW S&ITG - TS
> Sr. P/A, Engineer 310.764.9059
> mailto:Robert.Manning@trw.com http://www.trw.com
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> ---------------------------------------------------------------------------
> To unsubscribe from this list please send a message to majordomo@cougaar.org
> with the line "unsubscribe cougaar-developers" in the BODY of the message.
> ---------------------------------------------------------------------------
> To unsubscribe from this list please send a message to majordomo@cougaar.org
> with the line "unsubscribe cougaar-developers" in the BODY of the message.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu May 3 07:51:38 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id HAA25198
for cougaar-developers-outgoing; Thu, 3 May 2001 07:51:38 -0400
Received: from po1.bbn.com (PO1.BBN.COM [192.1.50.38])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id HAA25195
for ; Thu, 3 May 2001 07:51:37 -0400
Received: from bbn.com (ros-dhcp051-176.bbn.com [192.233.51.176])
by po1.bbn.com (8.9.1/8.9.1) with ESMTP id HAA06991;
Thu, 3 May 2001 07:54:36 -0400 (EDT)
Message-ID: <3AF14735.6FAC7B2@bbn.com>
Date: Thu, 03 May 2001 07:55:33 -0400
From: Bill Wright
X-Mailer: Mozilla 4.73 [en] (Windows NT 5.0; I)
X-Accept-Language: en
MIME-Version: 1.0
To: Robert Manning
CC: cougaar-developers@cougaar.org
Subject: Re: [Fwd: Re: Predicate question, clarified]
References: <3AF0A792.A81A6065@trw.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Robert,
The workaround you describe is consistent with what I guessed might be the problem.
I'm confident that using a DynamicUnaryPredicate would fix the problem, too. And I
don't think it would be any less efficient than what you're doing. It's basically
the same amount of work, the only difference is whether you do it in execute() or
the predicate.
Bill
Robert Manning wrote:
> Hi Bill,
>
> Thanks for replying. I will try the Dynamic version tomorrow. While musing
> on this problem, I came up with a different workaround...
>
> The different workaround was this: instead of my UnaryPredicate checking for
> the role "SoftwareDevelopment" in objects of type Organization, I wrote the
> predicate to simply check for objects of type Organization. That means I'm
> getting a lot of extra traffic entering my plugin's execute() method. I tried
> this since it seemed (at the time) that subscriptions to "SoftwareDevelopment"
> weren't firing. Now, in my plugin's execute() method, I have to perform a
> test on the IncrementalSubscription's elements to see if the OrganizationPG
> contains the role "SoftwareDevelopment" (I actually iterate through the whole
> list of roles, looking for this role).
>
> I'm sure this is a much less efficient way to find organizations that fulfill
> the role of SoftwareDevelopment, but it seems to be working.
>
> Also, an additional error message is coming from somewhere in Cougaar, that I
> don't understand:
>
> "DeferredRescind failed [task not found in logplan]: ClusterName/UID"
>
> I suppose it would be difficult to know why that happened offhand, but if by
> some slim chance there is an easy answer to that (and if by chance it has
> something to do with these other difficulties), I'd love to hear what I'm
> doing wrong :)
>
> Thanks Bill!
> -Robert
>
> -------- Original Message --------
> Subject: Re: Predicate question, clarified
> Date: Wed, 02 May 2001 18:48:44 -0400
> From: Bill Wright
> To: Robert Manning
> CC: "cougaar-developers@cougaar.org"
> References: <3AF05F7E.4D141AFF@trw.com>
>
> Sometimes you can get this behavior if you predicate tests an attribute that
> changes. In your example, it may be that the OrganizationPG changes, making the
> subscriptions behave in the way you notice. You might try making your predicate
> a DynamicUnaryPredicate instead of a UnaryPredicate and see if it makes a
> difference.
>
> Robert Manning wrote:
>
> > Hi,
> >
> > Thought I'd make this a bit clearer... what is happening, is that
> > my subscription to this unary predicate is firing, the unary predicate
> > is returning true, but when my execute() method fires in the plugin
> > owning the subscription, the Organization that matched the predicate
> > isn't in the list of elements (e.g. the java.util.Enumeration class
> > contained in IncrementalSubscription). Either I'm doing something
> > terribly wrong, or something is broken. Any Ideas?
> >
> > -------- Original Message --------
> > Subject: Predicate question
> > Date: Wed, 02 May 2001 11:43:47 -0700
> > From: "Robert Manning"
> > To: cougaar-developers@cougaar.org
> >
> > Hi all,
> >
> > What could make a unary predicate-type class stop matching correctly?
> > For example, consider this predicate from Cougaar exercise 4:
> >
> > class myProgrammersPredicate implements UnaryPredicate{
> > public boolean execute(Object o) {
> > boolean ret = false;
> > if (o instanceof Organization) {
> > Organization org = (Organization)o;
> > OrganizationPG orgPG = org.getOrganizationPG();
> > ret = orgPG.inRoles(Role.getRole("SoftwareDevelopment"));
> > }
> > return ret;
> > }
> > }
> >
> > Say you've put this class into it's own file and made it a public class
> > within the same package you're working on. Say also that you're no longer
> > looking for SoftwareDevelopment organizations that exist within your own
> > cluster - they're coming from another cluster within the same node. For
> > some reason, even though the subscription using this predicate is firing,
> > it is not correctly matching the role of SoftwareDevelopment to the
> > organization. Any ideas? Thanks :)
> >
> > --
> > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> > Robert Manning TRW S&ITG - TS
> > Sr. P/A, Engineer 310.764.9059
> > mailto:Robert.Manning@trw.com http://www.trw.com
> > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> > ---------------------------------------------------------------------------
> > To unsubscribe from this list please send a message to majordomo@cougaar.org
> > with the line "unsubscribe cougaar-developers" in the BODY of the message.
> > ---------------------------------------------------------------------------
> > To unsubscribe from this list please send a message to majordomo@cougaar.org
> > with the line "unsubscribe cougaar-developers" in the BODY of the message.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Thu May 3 23:53:50 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id XAA25599
for cougaar-developers-outgoing; Thu, 3 May 2001 23:53:50 -0400
Received: from pisces.starnet.gov.sg ([203.116.59.242])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with SMTP id XAA25596
for ; Thu, 3 May 2001 23:53:46 -0400
From: nkiabee@starnet.gov.sg
Received: (from nobody@localhost)
by pisces.starnet.gov.sg (8.10.1/8.10.1) id f443whB02335
for cougaar-developers@cougaar.org; Fri, 4 May 2001 11:58:43 +0800
X-Authentication-Warning: pisces.starnet.gov.sg: nobody set sender to nkiabee@starnet.gov.sg using -f
To:
Subject: Re: Allocating a task to multiple supporting clusters?
Message-ID: <988948723.3af228f342b30@www.starnet.gov.sg>
Date: Fri, 04 May 2001 11:58:43 +0800
References: <3AF0A792.A81A6065@trw.com> <3AF14735.6FAC7B2@bbn.com>
In-Reply-To: <3AF14735.6FAC7B2@bbn.com>
MIME-Version: 1.0
Content-Type: text/plain
Content-Transfer-Encoding: 8bit
User-Agent: StarNet Mail System (SMS)
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi all,
I tried with some codes for Cougaar and realised that in order to send a
task to multiple supporting clusters, I need to create a separate instance
of task for each cluster. Is this the behaviour of Cougaar?
I want to send a task to multiple supporting clusters to see who can
perform that task best before I assign the task to it. Are there any
sample codes/implementations for me to better understand it?
Thanks.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri May 4 14:45:23 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id OAA26017
for cougaar-developers-outgoing; Fri, 4 May 2001 14:45:23 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id OAA26014
for ; Fri, 4 May 2001 14:45:22 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA4EBB
for ;
Fri, 4 May 2001 11:51:05 -0700
Message-ID: <3AF2FA19.91370BF8@trw.com>
Date: Fri, 04 May 2001 11:51:05 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: publishAdd outside of execute()
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi All,
Does anyone know if calling SimplePlugIn.publishAdd() outside of its
execute() method will cause an error? I'm doing this inside of an event
handler (actionPerformed), and I'm getting a java.lang.NullPointerException
as a result.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Wed May 9 13:03:39 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id NAA29182
for cougaar-developers-outgoing; Wed, 9 May 2001 13:03:39 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id NAA29179
for ; Wed, 9 May 2001 13:03:37 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA241;
Wed, 9 May 2001 10:09:14 -0700
Message-ID: <3AF979BA.6D0599B3@trw.com>
Date: Wed, 09 May 2001 10:09:14 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Bug or feature?
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi all,
It seems to be the case that tracking of changes in logplan asset
properties does not take place in an expected way, when the same assets'
property(ies) is(are) modified repeatedly in the same execute cycle.
IncrementalSubscription.getChangedList() seems to return a list containing
the same asset, with the number of instances of the asset equal to the
number of times the asset was modified and publishChange'd back to the plan.
If a property in some property group of the asset is changed each time, and
represents different values at each publishChange, these changes are not
stored (or propagated to the IncrementalSubscription); only the *final*
value of the property is returned along with the N-instances of the asset
in IncrementalSubscription.getChangedList().
Example:
Say you have a cluster with two plugins. Both plugins are subscribed to the
same kind of asset. The first plugin is simply watching for additions,
changes, and removals of this object and displaying such events. The second
plugin fetches the asset from the plan, changes something in a property
group, and finally does a publishChange on the asset. This second plugin
might modify this property and publishChange the asset repeatedly in the
same execute cycle. For the sake of example, say this second plugin modifies
and publishChanges the asset 3 times.
After the second plugin finishes executing, the execute cycle of the
first plugin runs. The first plugin calls the getChangedList() method of
its subscription, and is returned a list with three assets. As it turns out,
this is the same asset repeated three times. I can understand why that
happens. However, the property which was modified has the same value in
all three instances! That is, a value in a property group for this asset
has taken on the *LAST* value that it had, in the "second" plugin. This
is simply not accurate; this property should contain the value it was
assigned before each publishChange. If the infrastructure is going to give
me three identical instances of the same asset from
IncrementalSubscription.getChangedList, what's the point? I *would* expect
each instance to hold the correct value of the assets' property at the time
it was modified - not simply the LAST value that it was given before the
last publishChange.
Can anyone confirm for me that this is as-designed behavior? If so, it
would seem to me that the infrastructure should return a single instance
of the changed asset, representing its final state, along with a count of
the number of times it was changed. Returning a list of N-assets of the same
type implies that they should each be a snapshot of what this asset looked
like each time it was publishChange'd. Since that isn't the case, the ability
to track the changes from another plugin doesn't exist, and results in extra
code to discard the identical duplicates.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Sat May 26 17:10:44 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id RAA07129
for cougaar-developers-outgoing; Sat, 26 May 2001 17:10:44 -0400
Received: from trafficmagnet.net (IDENT:root@[211.101.236.27])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id RAA07122
for ; Sat, 26 May 2001 17:10:41 -0400
Received: from screencapture1 ([211.101.236.29])
by trafficmagnet.net (8.12.0.Beta7/8.12.0.Beta7) with ESMTP id f4R6H5Wh014308
for ; Sun, 27 May 2001 01:17:05 -0500
Message-Id: <200105270617.f4R6H5Wh014308@trafficmagnet.net>
From: "Christine Hall"
Subject: WWW.COUGAAR.ORG
To: cougaar-developers@cougaar.org
Content-Type: text/html;
Reply-To: "Christine Hall"
Date: Sun, 27 May 2001 05:29:12 +0800
X-Priority: 3
X-Library: Trafficmagnet 8.0
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hello,
I visited
www.cougaar.org
and I
noticed that you are not listed on some search engines. I am sure you can
increase the number of people who visit
www.cougaar.org
. Do you know TrafficMagnet? TrafficMagnet is a unique technology that instantly submits your
web site to over 300,000+ search engines and directories every month. This is a
very low-cost and effective way of advertising your site.
To check our prices and submit
www.cougaar.org
to
300,000+ search engines, go to TrafficMagnet.net
I would love to hear from you.
Best Regards,
Christine Hall
Sales & Marketing
www.TrafficMagnet.net
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Jun 1 15:43:46 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id PAA10932
for cougaar-developers-outgoing; Fri, 1 Jun 2001 15:43:46 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id PAA10929
for ; Fri, 1 Jun 2001 15:43:45 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA3F6C
for ;
Fri, 1 Jun 2001 12:49:29 -0700
Message-ID: <3B17F1C6.327EEC46@trw.com>
Date: Fri, 01 Jun 2001 12:49:28 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Error message
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi all,
Can anyone give me an idea of what this error message means:
Plugins Loaded.
....(65)........+-....(126).Asset is not bound to an
LDM instance!
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Jun 1 16:01:09 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id QAA10951
for cougaar-developers-outgoing; Fri, 1 Jun 2001 16:01:09 -0400
Received: from draught.bbn.com (DRAUGHT.BBN.COM [128.89.28.26])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id QAA10948
for ; Fri, 1 Jun 2001 16:01:08 -0400
Received: (from mthome@localhost)
by draught.bbn.com (8.11.0/8.11.0) id f51K1UR06826;
Fri, 1 Jun 2001 16:01:30 -0400
From: Michael Thome
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <15127.62617.964170.584031@draught.bbn.com>
Date: Fri, 1 Jun 2001 16:01:29 -0400
To: "Robert Manning"
Cc: "cougaar-developers@cougaar.org"
Subject: Re: Error message
In-Reply-To: <3B17F1C6.327EEC46@trw.com>
References: <3B17F1C6.327EEC46@trw.com>
X-Mailer: VM 6.90 under 21.1 (patch 14) "Cuyahoga Valley" XEmacs Lucid
Reply-To: mthome@bbn.com
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
>>>>> "Robert" == Robert Manning writes:
> Hi all,
> Can anyone give me an idea of what this error message means:
> Plugins Loaded.
> ....(65)........+-....(126).Asset is not bound to an
> LDM instance!
This generally means that someone has created the Asset instance
without the assistance of the asset factory, or is doing something
like trying to actually use (in a log plan, for instance) an asset
which was deserialized from some non-standard form (http, file, xml).
--
Michael Thome (mthome@bbn.com)
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Jun 1 16:32:21 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id QAA10970
for cougaar-developers-outgoing; Fri, 1 Jun 2001 16:32:21 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id QAA10967
for ; Fri, 1 Jun 2001 16:32:21 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAAD22
for ;
Fri, 1 Jun 2001 13:38:17 -0700
Message-ID: <3B17FD37.F8690F48@trw.com>
Date: Fri, 01 Jun 2001 13:38:15 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Identical subscriptions from multiple plugins
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi all,
Suppose I have a cluster with two plugins, where both of these plugins
have identical IncrementalSubscription's to the cougaar
AllocationPredicate(). When the IncrementalSubscription fires for some
Allocation in this cluster, one or the other of these two plugins will
catch the event, but they *will not both* get the event. There seems to
be no way to tell which plugin will get the event in what order; if there
are multiple Allocations, any number of them may go to either plugin in
any order. Apparently Cougaar will give the IncrementalSubscription
event to any plugin in any random order.
May I assume from this behavior that if I put two plugins into a
cluster, that the two plugins should not have identical subscriptions
to any plan objects? Thanks!
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Jun 4 09:00:16 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id JAA12589
for cougaar-developers-outgoing; Mon, 4 Jun 2001 09:00:16 -0400
Received: from draught.bbn.com (DRAUGHT.BBN.COM [128.89.28.26])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id JAA12586
for ; Mon, 4 Jun 2001 09:00:15 -0400
Received: (from mthome@localhost)
by draught.bbn.com (8.11.0/8.11.0) id f54D5hc31463;
Mon, 4 Jun 2001 09:05:43 -0400
From: Michael Thome
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <15131.34727.228282.179574@draught.bbn.com>
Date: Mon, 4 Jun 2001 09:05:43 -0400
To: "Robert Manning"
Cc: "cougaar-developers@cougaar.org"
Subject: Re: Identical subscriptions from multiple plugins
In-Reply-To: <3B17FD37.F8690F48@trw.com>
References: <3B17FD37.F8690F48@trw.com>
X-Mailer: VM 6.90 under 21.1 (patch 14) "Cuyahoga Valley" XEmacs Lucid
Reply-To: mthome@bbn.com
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
It is difficult to tell without more details (your code, cougaar/alp
version, etc), but:
>>>>> "Robert" == Robert Manning writes:
> Hi all,
> Suppose I have a cluster with two plugins, where both of these plugins
> have identical IncrementalSubscription's to the cougaar
> AllocationPredicate().
First, define "identical". If the two subscriptions are somehow ==,
then the rest of this makes sense - you've done something that the
infrastructure has not expected... Essentially, you've got *one*
subscription. If the infrastructure has allowed you to do this, we
need to add some more error checking. If this isn't the situation,
I'm at a loss, as there are *many* examples of plugins with
subscriptions to the same objects which work perfectly fine - none of
the demonstrations would have worked if this did not function correctly.
> When the IncrementalSubscription fires for some
> Allocation in this cluster, one or the other of these two plugins will
> catch the event, but they *will not both* get the event. There seems to
> be no way to tell which plugin will get the event in what order; if there
> are multiple Allocations, any number of them may go to either plugin in
> any order. Apparently Cougaar will give the IncrementalSubscription
> event to any plugin in any random order.
> May I assume from this behavior that if I put two plugins into a
> cluster, that the two plugins should not have identical subscriptions
> to any plan objects? Thanks!
Two plugins may subscribe to the same objects, but may not usefully
share a subscription object. The subscription object maintains
the transactional state with respect to the plugin it serves, so the
behavior of two plugins attempting to share state is undefined.
--
Michael Thome (mthome@bbn.com)
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Jun 4 17:20:46 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id RAA12804
for cougaar-developers-outgoing; Mon, 4 Jun 2001 17:20:46 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id RAA12801
for ; Mon, 4 Jun 2001 17:20:45 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA4104
for ;
Mon, 4 Jun 2001 14:26:40 -0700
Message-ID: <3B1BFD0F.524C428A@trw.com>
Date: Mon, 04 Jun 2001 14:26:39 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Re: Identical subscriptions from multiple plugins
References: <3B17FD37.F8690F48@trw.com> <15131.34727.228282.179574@draught.bbn.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi Michael,
Thanks for responding. I didn't mean to imply that I had a reference
to the same subscription from two different plugins; simply that each
plugin had an IncrementalSubscription that used the same predicate
(org.cougaar.domain.planning.ldm.predicate.AllocationPredicate). AFAIK,
your infrastructure is intact; I wouldn't know how to get a pointer to
the same subscription from within two different plugins :)
Actually, it appears this is a case of brain-fade on my part. Remind me
not to send such emails on a Friday afternoon! My thanks go to another
list member (name withheld because he emailed me privately) for pointing
out that if one of the plugins were to modify the Allocation, then the
other plugin might not see it because of that change. In my case, both
plugins are changing the EstimatedResult to equal the ReportedResult,
iff the EstimatedReport is null and the ReportedResult is not null. The
clue I overlooked was always seeing exactly the right total number of
Allocations between the two plugins. Consider the effect of the following
code snippet, found in both of the plugins:
//myAllocations is an IncrementalSubscription to AllocationPredicate
Enumeration alloc_enum = myAllocations.getChangedList();
while (alloc_enum.hasMoreElements())
{
Allocation alloc = (Allocation) alloc_enum.nextElement();
if ((alloc.getReportedResult() != null) &&
(alloc.getEstimatedResult() == null))
{
if (alloc.getReportedResult().isSuccess() == true)
System.out.println("Task Successfully Completed.");
else
System.out.println("Task Failed, Not Completed.");
alloc.setEstimatedResult(alloc.getReportedResult()); //there's the change
publishChange(alloc);
}
}
I solved my problem by creating two different predicates, one for
each plugin. Since the task verbs being looked for were different, the
new predicates extract the Task from the Allocation and test for the
appropriate Verb (I should have done that in the first place). It now
works as expected. Thanks for the assistance!
-Robert
Michael Thome wrote:
>
> It is difficult to tell without more details (your code, cougaar/alp
> version, etc), but:
>
> >>>>> "Robert" == Robert Manning writes:
> > Hi all,
> > Suppose I have a cluster with two plugins, where both of these plugins
> > have identical IncrementalSubscription's to the cougaar
> > AllocationPredicate().
> First, define "identical". If the two subscriptions are somehow ==,
> then the rest of this makes sense - you've done something that the
> infrastructure has not expected... Essentially, you've got *one*
> subscription. If the infrastructure has allowed you to do this, we
> need to add some more error checking. If this isn't the situation,
> I'm at a loss, as there are *many* examples of plugins with
> subscriptions to the same objects which work perfectly fine - none of
> the demonstrations would have worked if this did not function correctly.
>
> > When the IncrementalSubscription fires for some
> > Allocation in this cluster, one or the other of these two plugins will
> > catch the event, but they *will not both* get the event. There seems to
> > be no way to tell which plugin will get the event in what order; if there
> > are multiple Allocations, any number of them may go to either plugin in
> > any order. Apparently Cougaar will give the IncrementalSubscription
> > event to any plugin in any random order.
>
> > May I assume from this behavior that if I put two plugins into a
> > cluster, that the two plugins should not have identical subscriptions
> > to any plan objects? Thanks!
>
> Two plugins may subscribe to the same objects, but may not usefully
> share a subscription object. The subscription object maintains
> the transactional state with respect to the plugin it serves, so the
> behavior of two plugins attempting to share state is undefined.
>
> --
> Michael Thome (mthome@bbn.com)
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Jun 4 19:22:43 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id TAA12856
for cougaar-developers-outgoing; Mon, 4 Jun 2001 19:22:43 -0400
Received: from avocet.mail.pas.earthlink.net (avocet.mail.pas.earthlink.net [207.217.121.50])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id TAA12853
for ; Mon, 4 Jun 2001 19:22:42 -0400
Received: from patdesktop (ip251.dayton11.oh.pub-ip.psi.net [38.31.203.251])
by avocet.mail.pas.earthlink.net (EL-8_9_3_3/8.9.3) with SMTP id QAA12313
for ; Mon, 4 Jun 2001 16:28:59 -0700 (PDT)
Message-ID: <001f01c0ee17$8ecf9080$3800000a@patdesktop>
From: "Pat Clark"
To:
Subject: Fw: Identical subscriptions from multiple plugins
Date: Tue, 5 Jun 2001 19:30:47 -0400
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4133.2400
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Robert,
Sorry for the double copy of this. If my response was the one that
helped, I figured I better correct MY Friday night brain-dead mistake of
sending this to you privately instead of to the group (another group I
belong to automatically puts the group name in the "to" field when I reply).
Maybe it'll help someone else, I know a lot of this is tricky to think
through the first couple times through it.
Pat Clark
----- Original Message -----
From: "Pat Clark"
To: "Robert Manning"
Sent: Saturday, June 02, 2001 10:38 PM
Subject: Re: Identical subscriptions from multiple plugins
> Robert,
>
> Pretty sure if you have two plugins with identical subscriptions (but
> different instances -- they don't point to the same subscription object,
but
> you're subscribing to the same conditions), they will both be woken up and
> their execute methods will be run. We have done this only by accident,
but
> spent hours finding out that in our case it was a problem. But I can
> envision where you would actually want to do this. As a contrived
example,
> in our BooksOnLine society, here's what we could have done. We could have
> had two plugins in the Warehouse, each subscribing to the
BooksFromWarehouse
> task sent by the OrderManager cluster. Each has subscribes to the same
> condition (tasks where verb = BooksFromWarehouse), but a different execute
> method. One plugin would do the logic of allocating the packer for the
> books, the other would examine the inventory in light of the current order
> and recent orders, and determine whether the publisher should be sent a
task
> to order more books.
>
> If you're not seeing this behavior, it's possible that once the
servicing
> is done with the first plugin, the second one "falls through" because
you're
> testing for a condition you've already serviced. Put a println at the
> beginning of the execute method and I'm pretty sure you'll see it print
> twice for each item.
>
> Pat Clark
> Clark Software Engineering,Ltd.
> 5100 Springfield St. Suite 308
> Dayton, OH 45431
> 937.256.7794
> fax: 937.256.7842
> patkclark@earthlink.net
>
> ----- Original Message -----
> From: "Robert Manning"
> To:
> Sent: Friday, June 01, 2001 4:38 PM
> Subject: Identical subscriptions from multiple plugins
>
>
> > Hi all,
> >
> > Suppose I have a cluster with two plugins, where both of these plugins
> > have identical IncrementalSubscription's to the cougaar
> > AllocationPredicate(). When the IncrementalSubscription fires for some
> > Allocation in this cluster, one or the other of these two plugins will
> > catch the event, but they *will not both* get the event. There seems to
> > be no way to tell which plugin will get the event in what order; if
there
> > are multiple Allocations, any number of them may go to either plugin in
> > any order. Apparently Cougaar will give the IncrementalSubscription
> > event to any plugin in any random order.
> >
> > May I assume from this behavior that if I put two plugins into a
> > cluster, that the two plugins should not have identical subscriptions
> > to any plan objects? Thanks!
> >
> > --
> > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> > Robert Manning TRW S&ITG - TS
> > Software Engineer 310.764.9059
> > mailto:Robert.Manning@trw.com http://www.trw.com
> > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
>
> --------------------------------------------------------------------------
> -
> > To unsubscribe from this list please send a message to
> majordomo@cougaar.org
> > with the line "unsubscribe cougaar-developers" in the BODY of the
message.
>
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Tue Jun 5 13:57:53 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id NAA13327
for cougaar-developers-outgoing; Tue, 5 Jun 2001 13:57:53 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id NAA13324
for ; Tue, 5 Jun 2001 13:57:52 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA2AD6
for ;
Tue, 5 Jun 2001 11:03:49 -0700
Message-ID: <3B1D1F04.B6076B00@trw.com>
Date: Tue, 05 Jun 2001 11:03:48 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Re: Fw: Identical subscriptions from multiple plugins
References: <001f01c0ee17$8ecf9080$3800000a@patdesktop>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi Pat,
Yes, it was you :) If someone emails me off-list, I usually respond in
like fashion. Interesting you should mention this... I checked the header
of your email, and there's no "Reply-To:" field in it. It looks like
they're using Majordomo, so I'd think that a Reply-To field could be
added, no? -Robert
Pat Clark wrote:
>
> Robert,
> Sorry for the double copy of this. If my response was the one that
> helped, I figured I better correct MY Friday night brain-dead mistake of
> sending this to you privately instead of to the group (another group I
> belong to automatically puts the group name in the "to" field when I reply).
> Maybe it'll help someone else, I know a lot of this is tricky to think
> through the first couple times through it.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Wed Jun 6 23:37:19 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id XAA14413
for cougaar-developers-outgoing; Wed, 6 Jun 2001 23:37:19 -0400
Received: from pisces.starnet.gov.sg ([203.116.59.242])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with SMTP id XAA14410
for ; Wed, 6 Jun 2001 23:37:15 -0400
From: nkiabee@starnet.gov.sg
Received: (from nobody@localhost)
by pisces.starnet.gov.sg (8.10.1/8.10.1) id f573nvF11240
for cougaar-developers@cougaar.org; Thu, 7 Jun 2001 11:49:57 +0800
X-Authentication-Warning: pisces.starnet.gov.sg: nobody set sender to nkiabee@starnet.gov.sg using -f
To:
Subject: Re: RMI marshal exception?
Message-ID: <991885796.3b1ef9e4e8110@www.starnet.gov.sg>
Date: Thu, 07 Jun 2001 11:49:56 +0800
References: <3AF0A792.A81A6065@trw.com> <3AF14735.6FAC7B2@bbn.com> <988948723.3af228f342b30@www.starnet.gov.sg>
In-Reply-To: <988948723.3af228f342b30@www.starnet.gov.sg>
MIME-Version: 1.0
Content-Type: text/plain
Content-Transfer-Encoding: 8bit
User-Agent: StarNet Mail System (SMS)
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Have anyone got any ideas what causes this RMI Marshal exception?
java.rmi.MarshalException: error marshalling arguments; nested exception
is:
java.net.SocketException: socket write error (code=10053)
java.net.SocketException: socket write error (code=10053)
at java.net.SocketOutputStream.socketWrite(Native Method)
at java.net.SocketOutputStream.write(SocketOutputStream.java:87)
at java.io.BufferedOutputStream.write
(BufferedOutputStream.java:116)
at java.io.ObjectOutputStream.drain(ObjectOutputStream.java:1238)
at java.io.ObjectOutputStream.setBlockData
(ObjectOutputStream.java:1260)
at java.io.ObjectOutputStream.writeObject
(ObjectOutputStream.java:383)
at sun.rmi.server.UnicastRef.marshalValue(UnicastRef.java:272)
at sun.rmi.server.UnicastRef.invoke(UnicastRef.java:110)
at org.cougaar.core.society.rmi.MTImpl_Stub.receiveMessage
(MTImpl_Stub.j
ava:89)
at
org.cougaar.core.society.rmi.RMIMessageTransport$KeyedMT.receiveMessa
ge(RMIMessageTransport.java:451)
at
org.cougaar.core.society.rmi.RMIMessageTransport$DestinationQueue.run
(RMIMessageTransport.java:788)
at java.lang.Thread.run(Thread.java:479)
Regards.
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Jun 8 05:47:47 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id FAA15218
for cougaar-developers-outgoing; Fri, 8 Jun 2001 05:47:47 -0400
Received: from pisces.starnet.gov.sg ([203.116.59.242])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with SMTP id FAA15215
for ; Fri, 8 Jun 2001 05:47:45 -0400
From: nkiabee@starnet.gov.sg
Received: (from nobody@localhost)
by pisces.starnet.gov.sg (8.10.1/8.10.1) id f58A0ZZ15099
for cougaar-developers@cougaar.org; Fri, 8 Jun 2001 18:00:35 +0800
X-Authentication-Warning: pisces.starnet.gov.sg: nobody set sender to nkiabee@starnet.gov.sg using -f
To:
Subject: Re: RMI marshal exception?
Message-ID: <991994435.3b20a243d50ad@www.starnet.gov.sg>
Date: Fri, 08 Jun 2001 18:00:35 +0800
References: <3AF0A792.A81A6065@trw.com> <3AF14735.6FAC7B2@bbn.com> <988948723.3af228f342b30@www.starnet.gov.sg> <991885796.3b1ef9e4e8110@www.starnet.gov.sg>
In-Reply-To: <991885796.3b1ef9e4e8110@www.starnet.gov.sg>
MIME-Version: 1.0
Content-Type: text/plain
Content-Transfer-Encoding: 8bit
User-Agent: StarNet Mail System (SMS)
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Maybe I should give more explanation to my problems:
I have setup a few PCs running various nodes supporting each others. I
noticed that some of the PCs can receive tasks from other nodes (in
different PCs) but they simply cannot publish tasks to others (in
different PCs). Only some PCs have the problems sending tasks out and they
give the error.
Are there any setting needs to be done. I don't recall doing that for
those PCs that can publish tasks out.
Regards.
----------------------------------------------------------------------
Quoting nkiabee@starnet.gov.sg:
> Have anyone got any ideas what causes this RMI Marshal exception?
>
> java.rmi.MarshalException: error marshalling arguments; nested
> exception
> is:
> java.net.SocketException: socket write error (code=10053)
> java.net.SocketException: socket write error (code=10053)
> at java.net.SocketOutputStream.socketWrite(Native Method)
> at
> java.net.SocketOutputStream.write(SocketOutputStream.java:87)
> at java.io.BufferedOutputStream.write
> (BufferedOutputStream.java:116)
> at
> java.io.ObjectOutputStream.drain(ObjectOutputStream.java:1238)
> at java.io.ObjectOutputStream.setBlockData
> (ObjectOutputStream.java:1260)
>
> at java.io.ObjectOutputStream.writeObject
> (ObjectOutputStream.java:383)
> at
> sun.rmi.server.UnicastRef.marshalValue(UnicastRef.java:272)
> at sun.rmi.server.UnicastRef.invoke(UnicastRef.java:110)
> at org.cougaar.core.society.rmi.MTImpl_Stub.receiveMessage
> (MTImpl_Stub.j
> ava:89)
> at
> org.cougaar.core.society.rmi.RMIMessageTransport$KeyedMT.receiveMessa
> ge(RMIMessageTransport.java:451)
> at
> org.cougaar.core.society.rmi.RMIMessageTransport$DestinationQueue.run
> (RMIMessageTransport.java:788)
> at java.lang.Thread.run(Thread.java:479)
>
>
> Regards.
> -------------------------------------------------------------------------
--
> To unsubscribe from this list please send a message to
> majordomo@cougaar.org
> with the line "unsubscribe cougaar-developers" in the BODY of the
> message.
>
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Jun 8 17:14:43 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id RAA15570
for cougaar-developers-outgoing; Fri, 8 Jun 2001 17:14:43 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id RAA15567
for ; Fri, 8 Jun 2001 17:14:42 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA54F7;
Fri, 8 Jun 2001 14:20:29 -0700
Message-ID: <3B21419B.3774ECED@trw.com>
Date: Fri, 08 Jun 2001 14:20:27 -0700
From: "Robert Manning"
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Usage for Results in an Allocation
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi All,
In interface org.cougaar.domain.planning.ldm.plan.Allocation, there are
several types of AllocationResult's (through class PlanElement):
EstimatedResult, ReportedResult, ReceivedResult, ObservedResult
According to the javadocs, I can obtain an AllocationResult instance for
any of these four types of results. I can also set the AllocationResult for
two types of results: EstimatedResult and ObservedResult (PlanElementImpl
javadoc states that only the infrastructure is to use setReceivedResult,
and that setReportedResult is deprecated).
My questions have to do with ambiguity and apparent usage.
Issue #1): Why Set Different Results?
The javadoc for PlanElement.setEstimatedResult() states "Set the estimated
allocation result so that a notification will propagate up another level."
According to the Cougaar Plugin Developers' Guide 7.0, para. 5.4.3.2, titled
"Notification (AllocationResult) Propagation", when an estimated
AllocationResult is set on a plan element (something that can happens with a
call to theLDMF.createAllocation that includes a non-null AllocationResult)
the "Local infrastructure LP propagates change to next upstream plan element
by copying this information to the reportedAllocationResult slot and performs
publishChange(PlanElement)." In other words, the Cougaar infrastructure copies
the EstimatedResult from the plugin/cluster performing a task, to the
ReportedResult "slot" in the plugin/cluster that issued the task in the first
place (see Figure 26, pg. 77).
I can understand (and greatly appreciate) the utility of automatically
communicating the AllocationResult in this fashion. But why is it necessary
to change the meaning of that result? In other words, in the plugin/cluster
that Allocates the Task, the inclusion of a non-null AllocationResult inside
a call to theLDMF.createAllocation() results in the apparent setting of the
EstimatedResult in that plugin, and the automatic setting of the
ReportedResult in the tasking plugin/cluster. Why isn't the EstimatedResult
set instead? Apparently, the setting of EstimatedResult is being used to
trigger the sending of the AllocationResult to the tasking plugin/cluster;
so by automatically setting the EstimatedResult in the original plugin/cluster,
you'd create an automatic chain of events throughout the infrastructure, for
as many plugins as were involved in this task. It almost appears as if this
is a workaround used to prevent this effect.
IMO, it would be simpler and clearer to have a method to trigger the
"upstream" reporting of all AllocationResults, and keep the meanings of the
various *Result types consistent across plugins. Which brings me to my next
question:
Issue #2): What are ObservedResult and ReceivedResult Used For?
Since EstimatedResult and ReportedResult seem to be the only two result types
I really need to worry about, what are these other two for? And, is there any
automatic functionality hidden behind them (such as for EstimatedResult)?
Can they be used to report the result of an allocation, the result of
performing a task, something else, etc? Which brings me to my third question:
Issue #3): What's The AllocationResult Used For?
IMO, the automatic communication of an AllocationResult on a task might come
in very handy for several things: it can be used to notify a plugin that
created a task that the task has been received and Allocated to some Asset
class. It could also be used to signify that a task *has been completed*.
Since the AllocationResult holds info relating to the confidence, success,
aspect types and results on *a task*, it seems quite natural to me, at first
glance, that AllocationResults and the automatic infrastructure communication
offered by setting the EstimatedResult, refer to the result of completing a
task (not simply allocating a task).
If the AllocationResult refers only to the result of an Allocation (e.g. the
association of an Asset and a Task), and the confidence/success/aspect type
array/results array information altogether only refer to the act of making an
Allocation, then what shall we use to report the result(s) of the task itself?
It would seem wasteful to create another task, just to report the results of
the original task. Then we'd have to worry about reporting the result of the
task that reported the result of a task that reported the result of a task
... (recursively ad infinitum).
So, what is the AllocationResult really used for? Can I use it for anything
I want to, as long as I use it consistently? Or must I consider that it is to
only be used for reporting results of allocations? If so, what would be the
preferred way to report on task completion (which to me is the more important
issue anyway)?
Thanks,
Robert
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Fri Jun 15 15:00:58 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id PAA20333
for cougaar-developers-outgoing; Fri, 15 Jun 2001 15:00:58 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id PAA20330
for ; Fri, 15 Jun 2001 15:00:56 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA1DD8
for ;
Fri, 15 Jun 2001 12:06:50 -0700
Message-ID: <3B2A5CC9.75F6BFE4@trw.com>
Date: Fri, 15 Jun 2001 12:06:49 -0700
From: "Robert Manning"
Organization: TRW
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: RoleSchedule?
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi All,
How do I access and manipulate the RoleSchedule and the schedules
for the various plan elements that are supposedly stored in it?
Does a RoleSchedule pertain to the roles between organizations, or
to the roles an organization plays while performing some task?
Correct me if I'm wrong, but it would appear that when I have an
asset of type Organization in my cluster (which includes myself
as an asset), that whenever I allocate a task to such an asset, an
entry should be made into the RoleSchedule for that asset. What
I can find in the source code indicates that this takes place inside
SimplePlugIn.publishAdd(). However, according to TASKS.PSP, there is
some kind of default RoleSchedule inside my Organizational assets,
that begins in 1990 and ends in 2010 (regardless of how many other
tasks I've allocated to that organization). This seems to be true
whether I'm allocating a task to my own cluster or some other
cluster.
If this isn't the right class to use for scheduling, can someone
point me in the right direction? Thanks.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Jun 18 13:46:14 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id NAA22265
for cougaar-developers-outgoing; Mon, 18 Jun 2001 13:46:14 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id NAA22262
for ; Mon, 18 Jun 2001 13:46:12 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA5D19;
Mon, 18 Jun 2001 10:52:04 -0700
Message-ID: <3B2E3FC3.317B4E47@trw.com>
Date: Mon, 18 Jun 2001 10:52:03 -0700
From: "Robert Manning"
Organization: TRW
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Allocation to self doesn't work?
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi All,
I am trying to allocate a task to my own cluster, but for some
reason the allocation does not appear in my logplan (in TASKS.PSP).
I obtained a reference to my cluster as an Asset, by using a very
general predicate that looks for all objects of type Asset, then
searches for matches between ItemIdentificationPG.ItemIdentification
(sans "UIC/") and getDelegate().getMetricsSnapshot().clusterName.
Once I have a reference to my own cluster as an Asset, I can then
create an Allocation using theLDMF.createAllocation() and publishAdd
it to my logplan.
This all seems to work without error, but the allocation never shows
up. Why?
Thanks! -Robert
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Jun 18 14:32:45 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id OAA22295
for cougaar-developers-outgoing; Mon, 18 Jun 2001 14:32:45 -0400
Received: from mail1.dh.trw.com (mail1.dh.trw.com [129.193.109.1])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id OAA22292
for ; Mon, 18 Jun 2001 14:32:44 -0400
Received: from trw.com ([129.193.108.38]) by mail1.dh.trw.com
(Netscape Messaging Server 3.6) with ESMTP id AAA316C;
Mon, 18 Jun 2001 11:38:40 -0700
Message-ID: <3B2E4AAF.A868B7FE@trw.com>
Date: Mon, 18 Jun 2001 11:38:39 -0700
From: "Robert Manning"
Organization: TRW
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: "cougaar-developers@cougaar.org"
Subject: Usage for Results in an Allocation (x2)
References: <3B21419B.3774ECED@trw.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-cougaar-developers@cougaar.org
Precedence: bulk
Hi All,
The following email was sent to the list 10 days ago. In the hopes
that it was forgotten and not ignored, I'm sending the question again.
Thanks :)
Robert Manning wrote:
>
> Hi All,
>
> In interface org.cougaar.domain.planning.ldm.plan.Allocation, there are
> several types of AllocationResult's (through class PlanElement):
>
> EstimatedResult, ReportedResult, ReceivedResult, ObservedResult
>
> According to the javadocs, I can obtain an AllocationResult instance for
> any of these four types of results. I can also set the AllocationResult for
> two types of results: EstimatedResult and ObservedResult (PlanElementImpl
> javadoc states that only the infrastructure is to use setReceivedResult,
> and that setReportedResult is deprecated).
>
> My questions have to do with ambiguity and apparent usage.
>
> Issue #1): Why Set Different Results?
>
> The javadoc for PlanElement.setEstimatedResult() states "Set the estimated
> allocation result so that a notification will propagate up another level."
> According to the Cougaar Plugin Developers' Guide 7.0, para. 5.4.3.2, titled
> "Notification (AllocationResult) Propagation", when an estimated
> AllocationResult is set on a plan element (something that can happens with a
> call to theLDMF.createAllocation that includes a non-null AllocationResult)
> the "Local infrastructure LP propagates change to next upstream plan element
> by copying this information to the reportedAllocationResult slot and performs
> publishChange(PlanElement)." In other words, the Cougaar infrastructure copies
> the EstimatedResult from the plugin/cluster performing a task, to the
> ReportedResult "slot" in the plugin/cluster that issued the task in the first
> place (see Figure 26, pg. 77).
>
> I can understand (and greatly appreciate) the utility of automatically
> communicating the AllocationResult in this fashion. But why is it necessary
> to change the meaning of that result? In other words, in the plugin/cluster
> that Allocates the Task, the inclusion of a non-null AllocationResult inside
> a call to theLDMF.createAllocation() results in the apparent setting of the
> EstimatedResult in that plugin, and the automatic setting of the
> ReportedResult in the tasking plugin/cluster. Why isn't the EstimatedResult
> set instead? Apparently, the setting of EstimatedResult is being used to
> trigger the sending of the AllocationResult to the tasking plugin/cluster;
> so by automatically setting the EstimatedResult in the original plugin/cluster,
> you'd create an automatic chain of events throughout the infrastructure, for
> as many plugins as were involved in this task. It almost appears as if this
> is a workaround used to prevent this effect.
>
> IMO, it would be simpler and clearer to have a method to trigger the
> "upstream" reporting of all AllocationResults, and keep the meanings of the
> various *Result types consistent across plugins. Which brings me to my next
> question:
>
> Issue #2): What are ObservedResult and ReceivedResult Used For?
>
> Since EstimatedResult and ReportedResult seem to be the only two result types
> I really need to worry about, what are these other two for? And, is there any
> automatic functionality hidden behind them (such as for EstimatedResult)?
> Can they be used to report the result of an allocation, the result of
> performing a task, something else, etc? Which brings me to my third question:
>
> Issue #3): What's The AllocationResult Used For?
>
> IMO, the automatic communication of an AllocationResult on a task might come
> in very handy for several things: it can be used to notify a plugin that
> created a task that the task has been received and Allocated to some Asset
> class. It could also be used to signify that a task *has been completed*.
> Since the AllocationResult holds info relating to the confidence, success,
> aspect types and results on *a task*, it seems quite natural to me, at first
> glance, that AllocationResults and the automatic infrastructure communication
> offered by setting the EstimatedResult, refer to the result of completing a
> task (not simply allocating a task).
>
> If the AllocationResult refers only to the result of an Allocation (e.g. the
> association of an Asset and a Task), and the confidence/success/aspect type
> array/results array information altogether only refer to the act of making an
> Allocation, then what shall we use to report the result(s) of the task itself?
> It would seem wasteful to create another task, just to report the results of
> the original task. Then we'd have to worry about reporting the result of the
> task that reported the result of a task that reported the result of a task
> ... (recursively ad infinitum).
>
> So, what is the AllocationResult really used for? Can I use it for anything
> I want to, as long as I use it consistently? Or must I consider that it is to
> only be used for reporting results of allocations? If so, what would be the
> preferred way to report on task completion (which to me is the more important
> issue anyway)?
>
> Thanks,
> Robert
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Robert Manning TRW S&ITG - TS
Software Engineer 310.764.9059
mailto:Robert.Manning@trw.com http://www.trw.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
---------------------------------------------------------------------------
To unsubscribe from this list please send a message to majordomo@cougaar.org
with the line "unsubscribe cougaar-developers" in the BODY of the message.
From owner-cougaar-developers@cougaar.org Mon Jun 18 15:18:47 2001
Return-Path:
Received: (from majordom@localhost)
by alp-61.alp.isotic.org (8.9.3/8.9.3) id PAA22319
for cougaar-developers-outgoing; Mon, 18 Jun 2001 15:18:47 -0400
Received: from mailgate1.darpa.mil (mailgate1.darpa.mil [192.5.18.61])
by alp-61.alp.isotic.org (8.9.3/8.9.3) with ESMTP id PAA22316
for ; Mon, 18 Jun 2001 15:18:46 -0400
Received: from mailhub1.darp