NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/comp/object/5030

Internet Message Format | 1993-01-24 | 3.6 KB

Path: sparky!uunet!pipex!warwick!uknet!gdt!aber!fronta.aber.ac.uk!pcg
From: [e-mail] (Piercarlo Grandi)
Newsgroups: comp.object
Subject: Re: Rethinking the Object Paradigm (Contradictions)
Message-ID: <[e-mail]>
Date: 23 Jan 93 17:49:48 GMT
References: <[e-mail]>
Sender: [e-mail] (USENET news service)
Reply-To: [e-mail] (Piercarlo Grandi)
Organization: Prifysgol Cymru, Aberystwyth
Lines: 57
In-Reply-To: [e-mail]'s message of 19 Jan 93 00: 01:20 GMT
Nntp-Posting-Host: decb.aber.ac.uk

>>> On 19 Jan 93 00:01:20 GMT, [e-mail] (Bob Marcus) said:

Marcus> The purpose of the original posting on "Rethinking the Object
Marcus> Paradigm" was to expose contradictions such as the two responses
Marcus> below. The author of the article was well aware of the research,
Marcus> commercial and standardization efforts in this area.

I am amazed by the claim that such awareness existed. Rereading the
original article:

    At the beginning of a New Year it might be timely to take a look at
 the current object paradigm to see if it needs to be amended or
 extended. In particular, many of the applications and distributed
 systems that are being developed today seem to fall outside the current
 paradigm and the OMG standardization process. [ ... ]

   Should this type of systems be incorporated into an extension of the
 object paradigm? Below are some related questions. [ ... ]

To me this gives the impression of somebody who has noticed that there
are several cases in which things are not quite reducible to her frame
of reference, and thinks that maybe nobody else has noticed it... In
 particular somebody that thinks that "current paradigm" is more or less
on the same level semantic level as "OMG standardization".

As to the contradictions, they are between a positive and a normative
viiew of what OO is about; my observation that actor based OO *is in
fact* a little known but ancient branch of OO research is positive;
Meyer's observation that actor computing *should not* be taken seriously
is normative instead (and maybe he *was* unaware of actor OO research
too, which would not be suprising -- it is quite obscure, and in any
case I surmise he would just dismiss it just as some people dismiss
neural network research).

I made a comment on what's there and not on what should be there simply
because that was the tone of the original article; if I were to express
my judgement, I would (surprise!) be sympathetic to Meyer's position: I
am not particularly fond of actor based OO; for me it's hard to see
actor systems as a form of closure based OO with particular restrictions
(have always one, and only one, thread associated with a closure, or, in
other words, serially reusable continuations instead of closures). But
then a lot of people seem to think that client-server computing is a
concept and not an implementation incident...

Marcus> In particular, the lack of standards for asynchronous messaging
Marcus> among active objects and/or actors.

It would be interesting to be able to have a messaging standard for
actor based languages. But I am not optimistic; this cannot be easily
done without some serious restrictions, for much the same reason for
which it is hard to define a standard method invocation call sequence
for non actor languages: there is enormous potential variety in the
design spectrum, and standard like the OMG ones really target only the
minimum facilities that everybody is likely to support.
--
Piercarlo Grandi <[e-mail]> c/o Dept of CS
University of Wales, Penglais, Aberystwyth SY23 3BZ, UK