[Modeling] Modeling an Agent Class
James Odell
email@jamesodell.com
Tue, 10 Jun 2003 12:55:05 -0400
> 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.
--Boundary_(ID_vm3VgsV61yV4xIkPF/7rpQ)
Content-type: text/plain; charset=ISO-8859-1
Content-transfer-encoding: quoted-printable
Hi Giovanni ,
We all appreciate all the good work you and Paola are doing in the
AgentClass area. However, I would be very careful about making making
comments that might be inflammatory (e.g., someone being =B3careless=B2). I
think it is very important that we all remain civil to one another. We are
all working on a difficult and important problem =8B and it is very important
that we solve this together as a team. So please, let us be constructive i=
n
our comments to one another.
So, on to your question: In UML 2.0, an Object is defined as an instance of
Class. So, Classes have objects, and Classifiers have members. So, would
it be useful to use the following, instead:
UML members =3D UML objects + AUML agents + UML Actors + UML Signals +
UML Activity + ....
Would this work?
Ciao,
Jim
On 6/10/03 11:51 AM, "Giovanni Rimassa" indited:
> Wagner, G.R. wrote:
>>> please find attached a short document (four pages) where Paola and I
>>> have described and motivated our proposal for modeling agent classes.
>>=20
>> I think there is a conceptual problem with your approach.
>>=20
>> Yes, we all agree that agents are not "objects", but we have to
>> be careful with the term "object". In the UML, everything in the
>> world is an "object", including agents. This means that there are
>> two ways how to distinguish agents and objects (corresponding
>> to two different terminologies):
> Gerd,
> you seem competent to me, but also a little careless. On what ground do y=
ou
> claim that
> 'In the UML everything in the world is an object'? You should refer to pr=
ecise
> sections of
> an official UML document (my favorite one is UML 2.0 Superstructure,
> ad/2003-03-02)
> to support your view.
>=20
> That said, I've been careless, too, in citing 'object' only to discover t=
hat
> though the word 'object' with
> a lowercase 'o' is widely used throughout the document, there doesn't app=
ear
> to be an 'Object' with an
> uppercase 'O' that is an instance of a 'Class' with a capital 'C'. Can an=
ybody
> help me? Jim, do you know
> whether 'Object' is defined somewhere in UML 2.0?
>=20
> I used 'object' to mean 'an instance of a class', inspired by the definit=
ion
> of a Class at page 63 of the UML 2.0 Superstructure:
> "A class describes a set of objects that share the same specification=
s of
> features, constraints, and semantics"
>=20
> So, I think that by replacing the word 'object' with the sentence
> 'InstanceSpecification such that its classifier is a Class', my
> previous reasoning still holds and is now fully defined and compliant wit=
h UML
> 2.0 metamodel.
>=20
>=20
> I'm reporting here where are in the UML 2.0 Superstructure the various
> definitions of the concepts I used
> in the previous paragraph.
>=20
> Classifier: page 56
> InstanceSpecification: page 34
> InstanceSpecification-Classifier association: page 35
>=20
>>=20
>> 1) The UML:object concept subsumes non-agentive entities
>> (your "objects") and agents, so
>>=20
>> UML:objects =3D "AUML:objects" + agents
>>=20
>> 2) In AORML, there are <<object>> and <<agent>> class stereotypes,
>> so
>> UML:objects =3D AORML:entities
>> =3D AORML:objects + AORML:agents + AORML:eve=
nts +
>> ...
>>=20
> Does your AORML language take into account UML 1.4 or also UML 2.0? If yo=
u
> consider UML 2.0,
> it seems to me that AORML:entities are UML2.0:InstanceSpecifiers...
>> Your approach is based on the misunderstanding that AUML:objetcs =3D
>> UML:objects. But UML:objects do not correspond to AUML:objects,
>> but rather to AUML:entities (like in AORML).
> Again, I think you should point at where in an official UML document of y=
our
> choice, the above equation is stated.
>=20
> Bye,
>=20
> Giovanni
>=20
>>=20
>> -Gerd
>>=20
>> =20
>>=20
>> _______________________________________________
>> Modeling mailing list
>> Modeling@www.fipa.org
>> http://fipa.org/mailman/listinfo/modeling
>>=20
>=20
--Boundary_(ID_vm3VgsV61yV4xIkPF/7rpQ)
Content-type: text/html; charset=ISO-8859-1
Content-transfer-encoding: 7BIT
<HTML>
<HEAD>
<TITLE>Re: [Modeling] Modeling an Agent Class</TITLE>
</HEAD>
<BODY>
<FONT FACE="Arial">Hi Giovanni ,<BR>
<BR>
We all appreciate all the good work you and Paola are doing in the AgentClass area. However, I would be very careful about making making comments that might be inflammatory (e.g., someone being “careless”). I think it is very important that we all remain civil to one another. We are all working on a difficult and important problem — and it is very important that we solve this together as a team. So please, let us be constructive in our comments to one another.<BR>
<BR>
So, on to your question: In UML 2.0, an Object is defined as an instance of Class. So, Classes have objects, and Classifiers have members. So, would it be useful to use the following, instead:<BR>
UML members = UML objects + AUML agents + UML Actors + UML Signals + UML Activity + ....<BR>
<BR>
Would this work?<BR>
<BR>
<BR>
Ciao,<BR>
<BR>
Jim<BR>
<BR>
<BR>
<BR>
On 6/10/03 11:51 AM, "Giovanni Rimassa" indited:<BR>
<BR>
</FONT><BLOCKQUOTE><FONT FACE="Arial">Wagner, G.R. wrote:<BR>
</FONT><BLOCKQUOTE><BLOCKQUOTE><FONT FACE="Arial">please find attached a short document (four pages) where Paola and I <BR>
have described and motivated our proposal for modeling agent classes.<BR>
</FONT></BLOCKQUOTE><FONT FACE="Arial"><BR>
I think there is a conceptual problem with your approach. <BR>
<BR>
Yes, we all agree that agents are not "objects", but we have to<BR>
be careful with the term "object". In the UML, everything in the <BR>
world is an "object", including agents. This means that there are <BR>
two ways how to distinguish agents and objects (corresponding <BR>
to two different terminologies):<BR>
</FONT></BLOCKQUOTE><FONT FACE="Arial">Gerd,<BR>
you seem competent to me, but also a little careless. On what ground do you claim that<BR>
'In the UML everything in the world is an object'? You should refer to precise sections of<BR>
an official UML document (my favorite one is UML 2.0 Superstructure, ad/2003-03-02)<BR>
to support your view.<BR>
<BR>
That said, I've been careless, too, in citing 'object' only to discover that though the word 'object' with<BR>
a lowercase 'o' is widely used throughout the document, there doesn't appear to be an 'Object' with an<BR>
uppercase 'O' that is an instance of a 'Class' with a capital 'C'. Can anybody help me? Jim, do you know<BR>
whether 'Object' is defined somewhere in UML 2.0?<BR>
<BR>
I used 'object' to mean 'an instance of a class', inspired by the definition of a Class at page 63 of the UML 2.0 Superstructure:<BR>
"A class describes a set of objects that share the same specifications of features, constraints, and semantics"<BR>
<BR>
So, I think that by replacing the word 'object' with the sentence 'InstanceSpecification such that its classifier is a Class', my<BR>
previous reasoning still holds and is now fully defined and compliant with UML 2.0 metamodel.<BR>
<BR>
<BR>
I'm reporting here where are in the UML 2.0 Superstructure the various definitions of the concepts I used<BR>
in the previous paragraph.<BR>
<BR>
Classifier: page 56<BR>
InstanceSpecification: page 34<BR>
InstanceSpecification-Classifier association: page 35<BR>
<BR>
</FONT><BLOCKQUOTE><FONT FACE="Arial"><BR>
1) The UML:object concept subsumes non-agentive entities<BR>
(your "objects") and agents, so<BR>
<BR>
UML:objects = "AUML:objects" + agents<BR>
<BR>
2) In AORML, there are <<object>> and <<agent>> class stereotypes,<BR>
so<BR>
UML:objects = AORML:entities <BR>
= AORML:objects + AORML:agents + AORML:events + ...<BR>
<BR>
</FONT></BLOCKQUOTE><FONT FACE="Arial">Does your AORML language take into account UML 1.4 or also UML 2.0? If you consider UML 2.0,<BR>
it seems to me that AORML:entities are UML2.0:InstanceSpecifiers...<BR>
</FONT><BLOCKQUOTE><FONT FACE="Arial">Your approach is based on the misunderstanding that AUML:objetcs =<BR>
UML:objects. But UML:objects do not correspond to AUML:objects, <BR>
but rather to AUML:entities (like in AORML).<BR>
</FONT></BLOCKQUOTE><FONT FACE="Arial">Again, I think you should point at where in an official UML document of your choice, the above equation is stated.<BR>
<BR>
Bye,<BR>
<BR>
Giovanni<BR>
<BR>
</FONT><BLOCKQUOTE><FONT FACE="Arial"><BR>
-Gerd<BR>
<BR>
<BR>
<BR>
_______________________________________________<BR>
Modeling mailing list<BR>
Modeling@www.fipa.org<BR>
http://fipa.org/mailman/listinfo/modeling<BR>
<BR>
</FONT></BLOCKQUOTE><FONT FACE="Arial"><BR>
</FONT></BLOCKQUOTE><FONT FACE="Arial"><BR>
</FONT>
</BODY>
</HTML>
--Boundary_(ID_vm3VgsV61yV4xIkPF/7rpQ)--