The LispOS Project: a position paper, part 1

[Dave Mason]

Sorry I haven't read all 1500 previous postings, so I hope I'm not
repeating too much.

Scott L. Burson writes:
> I guess you must mean that the specific application of Web browser applets
> wasn't thought of until later.  I'm fairly certain that the idea of sending
> executable code across the network was part of the design from the beginning
> (that being something Gosling, who previously created NeWS, has been into for
> a long time).  So it was a very small jump from there to applets.  I would say
> that really, he did identify the general region within which the "killer app"
> would fall, even if its exact form wasn't immediately clear.

Well, I can tell you the general region within which the killer app
will fall: an operating system that allows secure access to its guts
so that you can get almost the full performance out of the hardware -
which is going to be necessary to get flexible multimedia and speech
out of reasonably priced hardware within the next 5 years.

This is not a new idea.  We should be looking at:
  Henry Massalin's work on the Synthesis Kernel,
     http://www.cs.arizona.edu/people/bridges/os/research.html 
  Gregor Kiczales work on MOPs
     http://www.parc.xerox.com/spl/projects/oi/towards-talk/transcript.html
  QNX's Photon display system
     http://www.qnx.com/product/photon/photonindex.html

And here's a good general resource page on much of the area:
     http://www.parc.xerox.com/spl/projects/oi/oi-bib.html

Personally, I'd go for a system written in CAML at the lowest level
that included a complete reflective system supporting Scheme.

But I won't have time to work on it much, so what I'd go for probably
isn't very relevant.  :-)

../Dave