Board index » delphi » Interesting article on concurrent programming

Interesting article on concurrent programming


2006-07-21 03:28:57 AM
delphi61
tinyurl.com/l795g
"We must and can build concurrent computation models that are far more
deterministic, and we must judiciously and carefully introduce
nondeterminism where needed. Nondeterminism should be explicitly added
to programs, and only where needed, as it is in sequential programming.
Threads take the opposite approach. They make programs absurdly
nondeterministic and rely on programming style to constrain that
nondeterminism to achieve deterministic aims."
Perhaps DevCo is looking at stuff like this when they mentioned they are
looking at doing some things for concurrency.
--
Brian Moelk
Brain Endeavor LLC
XXXX@XXXXX.COM
 
 

Re:Interesting article on concurrent programming

"Brian Moelk" <XXXX@XXXXX.COM>wrote
Quote
"We must and can build concurrent computation models that are far
more deterministic, ... Nondeterminism should be explicitly added
to programs, and only where needed, as it is in sequential programming.
Threads take the opposite approach. They make programs absurdly
nondeterministic and rely on programming style to constrain ...
Reminds me of the debate about the subject of
Milton and Rose Friedman's book titled "Free to Choose":
Which is better lots of little decision makers doing their
own optimization or one bureaucratic central management
organization trying to do what they think is best for all.
--JohnH
 

Re:Interesting article on concurrent programming

John Herbster writes:
Quote
Which is better lots of little decision makers doing their
own optimization or one bureaucratic central management
organization trying to do what they think is best for all.
Reading the article, I wasn't reminded of that at all. ;)
--
Brian Moelk
Brain Endeavor LLC
XXXX@XXXXX.COM
 

Re:Interesting article on concurrent programming

Quote
Reminds me of the debate about the subject of
Milton and Rose Friedman's book titled "Free to Choose":
Which is better lots of little decision makers doing their
own optimization or one bureaucratic central management
organization trying to do what they think is best for all.
One general management that fits common needs with the possibility to adapt for
personal needs.
A minimal management should be kept under the control of a supervisor that
ensures crucials tasks get done and elementary guidelines are respected.
 

Re:Interesting article on concurrent programming

John Herbster <herb-sci1_at_sbcglobal.net>writes
<44bfd9e7$XXXX@XXXXX.COM>
Quote
Reminds me of the debate about the subject of
Milton and Rose Friedman's book titled "Free to Choose":
Which is better lots of little decision makers doing their
own optimization or one bureaucratic central management
organization trying to do what they think is best for all.
Hmmm...interesting. That would suggest that fat clients are the way to go, in
macro-architecture, and event-driven architectures are best in micro-
architectures (application level and below).
--
***Free Your Mind***
Posted with JSNewsreader Preview 0.9.4.2549
 

Re:Interesting article on concurrent programming

Another persepctive, suggesting the fundamental problem is plan
interference:
www.drjava.de/e-presentation/html-english/img0.html
 

Re:Interesting article on concurrent programming

If I were to remove threads from my commercial app it would appear to be
horribly unresponsive. I haven't had any problems at all with threads since
I moved to WideString instead of String.
 

Re:Interesting article on concurrent programming

Peter Morris [Droopy eyes software] writes:
Quote
If I were to remove threads from my commercial app it would appear to be
horribly unresponsive.
Of course. I don't think that was the point of the paper. ;)
Quote
I haven't had any problems at all with threads since
I moved to WideString instead of String.
Interesting, sounds like a bug in string refcounting or it reveals some
problems in how those strings are being used.
--
Brian Moelk
Brain Endeavor LLC
XXXX@XXXXX.COM
 

Re:Interesting article on concurrent programming

Quote
Perhaps DevCo is looking at stuff like this when they mentioned they are
looking at doing some things for concurrency.
I doubt that they will use anything described in the article. Presumably
they might work within the current framework of common windows approaches
and simplify threading and the like, but I don't thin they have the
resources or inclination to go for the kind of language
redesign/modification described in the article in order to make Delphi more
concurrent.
Oliver Townshend
 

Re:Interesting article on concurrent programming

Oliver Townshend writes:
Quote
>Perhaps DevCo is looking at stuff like this when they mentioned they are
>looking at doing some things for concurrency.

I doubt that they will use anything described in the article. Presumably
they might work within the current framework of common windows approaches
and simplify threading and the like, but I don't thin they have the
resources or inclination to go for the kind of language
redesign/modification described in the article in order to make Delphi more
concurrent.
I tend to agree, although if they want a significant competitive
advantage IMO, it is worth looking at rather than just providing the
common thread control language elements that other languages already
provide.
--
Brian Moelk
Brain Endeavor LLC
XXXX@XXXXX.COM