moving on.

[Tom Novelli]

On Sat, 12 Dec 1998, RE01 Rice Brian T. EM2 wrote:

> btw, is anyone still following this thread?  does it seem pointless, or
> something?  does anyone see any benefits of the fundamental differences
> between my ideas and the ones you have been working with so far?  i'm just
> wondering when (or if) i'll get any helpful feedback.

Well, I probably can't help much, because I don't really understand what
you're doing. If nobody responds, I guess it is pointless to post all this
stuff. Just write out all yours ideas in your notebook, and just summarize
them to everyone else every week or so. At least, that's what I'd do.

Today I've been working on a concrete "Unified object paradigm" where
everything is an object... That includes variables, functions, and the
object tables (used by the garbage collector). So I guess that makes it
reflective. BTW, I was inspired by the Rekursiv processor.. a lot of ideas
from that project apply to Tunes. Anyone who's interested, look here:
http://www.brouhaha.com/~eric/retrocomputing/rekursiv
http://www.saqnet.co.uk/users/beloff/computing/aieuropa.html

I was also thinking about distributed persistent storage... how I'd
actually implement it. It seems pretty straightforward for a peer-to-peer
network of PC's. But it can get more complicated.. I used the example of a
cluster of massively parallel machines on a LAN connected to the internet.
In a large distributed computing project, there's no way 1 single
processor can keep a list of every individual object. Also, there WILL be
data loss, so redundancy is a must. It's similar to a brain.. I'll have to
study up on brain research.

In the meantime, I think I'll connect my PC's and write a crude prototype
"Persistent Store Manager" that makes the whole thing (both machines' RAM
and disks) work like a single piece of memory.

By the way, Brian, does this stuff have any relevance to what you're
doing? Maybe we could bridge the abstract <-> concrete gap..

--
Tom Novelli