Re: [cosmacelf] Re: Twitter Experiment
- Dear Lee,
You might be interested in a book I came across recently:
"Matrix Thinking" by David Noel, BFC Press, 1997. It's available for
free download over the internet. It's a little out there but has some
interesting ideas along the same lines.
On 19-May-09, at 11:53 AM, Lee Hart wrote:
Dave Ruske wrote:
>> ...It's no substitute for this group, it's more just a way
>> to shout a quick "eureka" to a bunch of like-minded people when you
>> finally get that TCP/IP stack working on your battered old RCA VIP,
>> or build an 1861 out of triodes.
Al W wrote:
> an 1861 out of triodes?... as in vaccuum tubes? No problem here
> because I take medications to surpress such ideas!
Brute force solutions don't require much thought -- just big hammers.
Today, such solutions are popular, because technology is cheap and easy.
Turn on a light with a switch? Naa, use a microcomputer with a million
transistors; it's easier.
The problem with such solutions is that they have negative consequences
in the long run. They don't last; complexity leads to poor reliability.
They use more resources, both in energy and raw materials. They are very
hard to fix; you generally have to throw them away and replace them.
And, they tend to be special-case solution; they can't be applied or
extended to other problems, so you need to start all over.
My 30-year-old RCA VIP still works. Do you think your PC, with hundreds
of times more parts, will still be working in even 3 years?
The VIP uses less than 1 watt. Even your laptop uses 100 times more
power. Is that wise in a world that's running out of energy?
The VIP has socketed ICs, and is easy to fix. Pretty much all the parts
are still available, and it is fully documented and easy enough to
understand so you *can* fix or modify it as the need arises. None of
this is true about your PCs; when it breaks, you have to replace it (or
at least, replace entire subassemblies). More junk for the landfills.
The purpose for thinking about TCP/IP stacks on a VIP, or an 1861 with
triodes is not that you would literally solve the problem that way... It
is to force you to *think* about the problem. How do you accomplish the
*function* of the TCP/IP stack or 1861 on a completely different
minimalist platform? Thinking about this will force you to decide what
is important, and what is just mindless overkill.
Example 1: You could certainly connect the VIP's audio cassette port to
the phone line, and exchange data with remote computers. The data rate
is low, but the amount of data you send it similarly low. The VIP is
obviously capable of efficiently handling 140-character Twitter
messages! The challenge is to figure out what parts of TCP/IP are
needed, and which are fluff.
Example 2: A good old CRT is basically a single triode vacuum tube. You
wouldn't build the functional equivalent of an 1861 with a thousand
triodes; you'd build a special CRT that converts digital data into video
on a screen. The challenge to figure out how such a tube might work, and
whether it is in fact feasible to build. (You may laugh... but the
blue-green dot matrix fluorescent display on the dashboard of some cars
*is* a vacuum tube display)!
Our high level of technology gives us *many* solutions to most problems.
But, it is human nature to simply pick the first solution that pops into
our heads, and relegate all the other solutions to insignificant niche
markets, or ignore them completely as if they don't exist.
"The design is not finished when you can't find anything more to add.
It is finished when you cannot find anything more to remove."
(see http://www.jbox.dk/quotations.htm for more quotes on simplicity)
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
- - - - - - - - - -
William A. McKee <william@...>
Asia Communication Quebec Inc.