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

Re: Making CSP into a success


What an interesting and constructive set of ideas initiated by
Marcel. Let me brainstorm a bit.

(1) C was a product of academia too. Why did it succeed?
Interlinked factors:
  (a) It reached critical mass, and everybody learned it.
  (b) There were no legal fences around it, free software
development helped it take off.
  (c) Soon, there was a C compiler for everything.

(2) The industrialist must be convinced of a need. Moore's Law
has thus far hobbled parallel processing in that respect.

(3) The industrialist works from ignorance, i.e. makes decisions
based on what experts tell him. This tremendously hurts an
"outsider". Especially if it has a record of failure (from a
business perspective) like the Transputer.

Before I get run out of occam-com on a rail, let me add:

(4) Government (education, space and especially military) can
respond more creatively (longer-range vision) to technically
offbeat ideas. Don't forget them!

(5) The bias toward massive complexity caused by industry's
blind side is getting so extreme that "revolutionary" rethinking
may be possible. We need to prepare now to STRIKE LIKE A SNAKE
when the time is ripe.

Specific responses:

>From: M_Boosten <mboosten@xxxxxxxxxxxxxxxxxxx>
>Date: Fri, 29 Sep 2000 09:23:50 +0200
>Message-Id: <200009290723.JAA21255@tigran>
>To: java-threads@xxxxxxxxx, occam-com@xxxxxxxxx
>Subject: Making CSP into a success
>4. Acadamics should realize the industrialists (generally people who
>   are eager for both success and money) will pick up their ideas
>   if they produce (1) readable, well-understandable papers, that (2)
>   solve real-life problems, and (3) show which problems will not be
>   solved, and which category of techniques complements the new ideas.
>   For serious industry, money is peanuts compared to a practically
>   useable idea.

Also very important: You must offer the industrialist usable tools,
because a tool under development looks like uncertainty and failure
to him. MAYBE YOU ALREADY HAVE A USABLE TOOL, if a "recipe" for using
it can be put together and demonstrated.

>From: M_Boosten <mboosten@xxxxxxxxxxxxxxxxxxx>
>Date: Fri, 29 Sep 2000 14:59:15 +0200
>Message-Id: <200009291259.OAA25587@tigran>
>To: java-threads@xxxxxxxxx, occam-com@xxxxxxxxx, Oyvind.Teig@xxxxxxxxxxxx
>Subject: Re: Making CSP into a success
>Hi every one, and Oyvind, and Stephen in particular,
>> Marcel does not mention what industrialists should do!
>> I'm excited to hear!
>Ok, I'll complete my view.  Stephen's input has been used.
>pre-s. no intention to insult anyone... of course... just trying to
>       make my point clear.
>I think industrialists can be divided into two groups:
>1. USER
>USERs are characterized by:
>- They have got a problem.
>- If the solution is ever accepted, it will be partially implemented,
>  just enough to solve the problem at hand.  To get more than the
>  basic concepts accepted/implemented is pure luck.
>- I'm in this situation. Yes: I'm frustrated about quite a few
>  things...

Marcel, isn't this situation the result of the difficulty in presenting
a "complete package" (what the jargon calls a "solution")? If all the
vendors aren't doing CSP...

That creates a conservative bias and herd instinct which is disguised
by all the multicolored marketing fluff. There's a gap (you can't get
there because you're not already there). Since this gap would imply
you can never introduce a revolutionary technology, and we know that
is not true, there must be a way around it. You propose...

>- Want convincing proof that MONEY can be made by solving
>  USERs problems

I do NOT think this is the whole solution. This may sound weird, but I
think you also get help from


the kind of people who have fueled GPL and academic creations - a more
"primary source" than money, which also has a conservative bias (market

The ideal funding source is therefore an individual who likes the
idea enough so he will see it through to completion for its beauty
alone. Paradoxically, he is the most likely to make a massive amount
of money - because of taking over an unpopulated niche - because he
doesn't really care that much about the money.

>From: M_Boosten <mboosten@xxxxxxxxxxxxxxxxxxx>
>Date: Fri, 29 Sep 2000 15:52:07 +0200
>Message-Id: <200009291352.PAA25900@tigran>
>To: java-threads@xxxxxxxxx, occam-com@xxxxxxxxx, Oyvind.Teig@xxxxxxxxxxxx
>Subject: Re: Making CSP into a success
>> I did say something about the industrial/academia problem in my
>> presentation at CPA 2000:
>> http://www.autronica-maritime.com/pub/tech/rd/CPA2000/frame.html
>> See "Interesting", "Dull" and "Why not?".
>Yes, I loved that presentation.
>I think its value was under estimated.
>You deserved a conference price, I think...
>Realizing the situation in which we are stuck is essential
>to realize how to get out of it.
>> I think occam is beautiful.
>Occam is indeed beautiful, but far from sexy.
>Sex and money makes the world go around.

That use of "sexy" is a hackneyed marketing metaphor which
stifles creative thought.

Oyvind is right, but it is hard to show the world. We need to
find a small niche and completely "occamize" it. Reconfigurable

>>   According to Marcel I should be ugly.
>I HOPE you mean "It" instead of "I", otherwise, I'm unable
>to understand your conclusion.
>Occam is ugly, but indeed, it is beautiful if you know
>what it is all about...
> Marcel

CSP and occam begin to look very beautiful indeed when you are
trying to get a complex system of independent parts to work. The
OO world just chokes and dies on this... We need to find a
horrible practical problem that is a pushover with CSP.