[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Objects, processes, and encapsulation



Hi Tom, Ian, Larry,

I read Tom's paper.

The following statement from his paper, I like very much.

     (Section 2.2.3 Programming Style vs. Language Design)
     The motivating concern of this research is that disciplines,
     being so vital to Software Engineering, should be COMPULSORY
     rather than voluntary - i.e., they should be built into the language.

I think this is an excellent starting point. Furthermore, Tom also doesn't
declare Occam to be holy. He starts from a relatively neutral point
of view. It corresponds to Ian's remark of starting again from the beginning.
I like Tom's approach.

Ian wrote:
> Truth is (and this is a major problem) while in uni, we don't get enough
> hard experience, but while in industry, we get stuck with given tools and
> methods, and don't get enough chance to work out how something should be
> done. UML has arrived (in earnest) after my last time in industry, which
> saw homage to OOP but little real application.

It is the task of research (university & commercial) to develop methods,
standards, tools. In industry, many tools are developed to solve the problems
at hand. However, people in industry are hardly motivated to solve the
problems in a wider scope: they solve their own problems, en communicate
mainly internally about the solutions.

University research should try to enter into that circuit of industrial problem
solvers. Universities could even take the problems out, back to the university,
and try to solve them there.

Statement:
     Industrial Contacts are Crucial to make a Measurable Amount of
     University Research Work being exploited in Practice.



Marcel Boosten           | Philips Medical Systems
System Designer 3DRA          | Room QJ1309
System Design & Image Quality | P.O. Box 10000
Cardio Vascular Development   | 5680 DA  Best
Marcel.Boosten@xxxxxxxxxxx    | The Netherlands
Phone: +31-40-27-69019        | Fax: +31-40-27-65650






Lawrence Dickson <tjoccam@xxxxxxxxxxxxx> on 10/03/2001 18:08:32

To:     ianeast@xxxxxxxxxxxxxxx
        Marcel Boosten/BST/MS/PHILIPS@EMEA2
        occam-com@xxxxxxxxx
cc:
Subject:  Re: Objects, processes, and encapsulation
Classification:



Ian, Marcel, and all,

I heartily agree with Ian. ALL the "fads" have the
characteristic that they bloat almost instantly. Try,
for instance, for a complete definition of Java - how
many thousands of pages?

Only CSP/occam offers the capability of doing everything
(multitasking and multiprocessing) using definitions
that are finite and understandable. I still carry around
my old occam 2 reference manual, plus my Transputer
instruction set (which when used in reverse allows
extension to all processors). Taken together they are
less than 300 pages.

Larry Dickson

>From owner-occam-com-out@xxxxxxxxx Wed Oct  3 01:14:12 2001
Subject: Re: Objects, processes, and encapsulation
Date: Wed, 3 Oct 01 09:14:01 +0100
x-sender: ianeast+akela@xxxxxxxxxxxxxx
x-mailer: Claris Emailer 2.0v2, June 6, 1997
From: ianeast <ianeast@xxxxxxxxxxxxxxx>
To: <marcel.boosten@xxxxxxxxxxx>, <occam-com@xxxxxxxxx>
Mime-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Precedence: bulk

I reserve judgement.

Truth is (and this is a major problem) while in uni, we don't get enough
hard experience, but while in industry, we get stuck with given tools and
methods, and don't get enough chance to work out how something should be
done. UML has arrived (in earnest) after my last time in industry, which
saw homage to OOP but little real application.

I think we need to start again from the beginning, and that means with
CSP. Our problem now is the lack of adequate application. I see this time
as one for regrouping and coming back with new and better tools. It's a
good sign that so many CPA papers got stuck into that. To fiddle with
current fads like Java, UML, etc, I believe is a serious mistake. We'll
simply drown that way. It's too easy to lose the benefits, like security,
and simple abstraction.

If necessary we should be prepared to go all the way into application
ourselves to prove the point (again).

Ian


Dr. Ian Robert East         School of Computing and Mathematical Sciences
ireast@xxxxxxxxxxxxx                            Oxford Brookes University
(44) 1865 483635                                          Turing Building
                                                          Wheatley Campus
                                                          Oxford OX33 1HX
Consultation hours for 2001/2002 Term 1
Mon 09.00..11.00
Fri 11.00..13.00