On Sun, Jul 6, 2008 at 10:22 AM, K. K. Subramaniam <subbukk at gmail.com>
wrote:

> You do have a point about Intel/MS ignoring past work in this area. Their
> research director could have presented the capability as a revival of prior
> research instead of making it appear like a new challenge. CST and Actalk
> are
> over two decades old and cited in many ACM papers.  Their collective
> amnesia
> is inexcusable.


Do you think it is intentional, or merely a sad commentary on the general
state of our current scientific infrastructure?  I do process improvement
(Six Sigma) work at a hospital, which *assumes* an underlying foundation of,
say, deductive and inductive reasoning, logic, math, science, even
philosophy to some extent.  I think industry isn't so different from
medicine, where vast specialization has pushed the generalist to the side so
that not too many people see "the big picture" anymore.

I'm not saying it's an excuse, just maybe a reason--and yet another reason
why the world NEEDS people like you having this conversation.  I think of
the Squeak community kind of like Science Fiction writers--invariably what
they dream up comes to pass...


> If processing elements could be wrapped in an object, then 1024 looks like
> a
> small number....


I know I'm not thinking at the level you guys are right now, but this is
interesting...can you explain what you mean by "wrapping processing elements
in an object" so I could get a picture of what that might look like?

Rob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20080706/2a6b5a0f/attachment.htm