From: Eugene Leitl (eugene.leitl@lrz.uni-muenchen.de)
Date: Mon Dec 27 1999 - 16:22:49 MST
Bryan Moss writes:
> I offer my Three Laws of Computing for approval,
>
> (1) Release source not standards.
Release definitions in some formal language. Make a reference
implementation in a widely used language on a well-defined system.
> (2) Emphasize hardware.
If we're talking about longevity, we should ignore hardware.
> (3) Avoid human-computer interaction.
>
> The first law is inspired by Squeak; standards can only ever
> provide a *lack* of foresight. The second law is inspired
Source mutates. As long as you keep proper version control, it's
ok. But this makes it unsuitable for standard definition. Better make
it a formal language which doesn't run on anything than paper and
pencil.
> by Danny Hillis; future directions in computing are defined
> by hardware, not software (massively parallel computing,
> FPGAs, quantum computing, ubiquitous computing, etc). The
> third law is inspired by David Tennenhouse, a Darpa computer
> scientist; to quote Tennenhouse, "We need to declare victory
> in the field of interactive computing and move on."
>
> Consider programming. My interface to a program is through
> a language such as C. I write a second interface for the
> user through that interface, and then I write an interface
> to the users interface (an options dialog, for example)
> again through the original interface. That's three layers
> of interface. Now consider a genetic algorithm that mutates
> it's own machine code. Our current direction is to add
> ever-increasing amounts of crud. The breaking point for me
> was the introduction of Wizards; these provide an interface
> to increasingly complex options dialogs. That's four layers
> of interface. Humanity should be ashamed.
Efficiency and longevity/portability are mutually exclusive. There
should be a frontier where people do bitbanging and produce
incompatible systems, but this is not what I was talking about.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:06:13 MST