RE: PROJ: Flare

From: Eugene Leitl (
Date: Wed Jul 18 2001 - 04:29:03 MDT

On Wed, 18 Jul 2001, Reason wrote:

> I have to throw in my 2c here; at this present time in the evolution
> of computer languages, I'd say there really are only two choices in
> the fundament of a self-modifying piece of code if you want the

Languages are irrelevant. What you need is an all-purpose framework, which
is expressive, compact, uses a dense encoding, can be executed by current
hardware more or less efficiently, and will scale to future architectures
(that be computronium, and precursors in form of cellular hardware). Oh,
and of course it has to have some magic fairy dust in regards to
information processing, and be intrinsically safe to mutate (the usual
shape of maximum variability within small mutation ball, and the space
permeated with fitness-neutral filaments).

> language to work with you and save you time by its very structure. 1)
> would be Forth, and 2) would be XML.

XML is just verbose, obfuscated SEXPRs. XML is only cool because it's a
standard, and is used in other useful things like XML-RPC.

I don't want to diss XML, but it's only value is that it's a standard.

> Forth is used by the NSA and maybe ten to twenty companies worth knowing.

Forth is used in many strange places. There's even Forth hardware in

> [Deep dark secret: runs on Forth. You didn't hear it from me.
> There's an 8-bit JVM written in Forth for mobile phones too -- now that's
> pretty twisted if you actually know Forth. Forth is pretty cool and very
> easy to pick up...but no-one uses it. A big useful feature is that you can

I wouldn't say no one.

> alter and recompile any and all portions of code while the code is running.
> Fun].
> But enough about Forth. I had a misspent youth, I admit it.

You, too?

> So they're using XML as a basis for Flare and my CTOesque opinion is that
> it's the best choice as an underpinning for what the Institute wants to do.

It's verbose as hell, but Koza uses SEXPRs with some limited success, so I
presume XML will do.

-- Eugen* Leitl <a href="">leitl</a>
ICBMTO : N48 10'07'' E011 33'53''
57F9CFD3: ED90 0433 EB74 E4A9 537F CFF5 86E7 629B 57F9 CFD3

This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:39:49 MDT