CMUCL and LispOS
Kelly Murray
kem@Franz.COM
Tue, 29 Apr 1997 12:32:43 -0700
> From: "Vassili Bykov" <vbykov@cam.org>
> A while ago I used to have a Windoze system with Smalltalk/V set up as
> Progman replacement. For all I cared, that was pretty much a Smalltalk
> machine. Something along these lines, of course with a deeper
> penetration into the base system could be a reasonable evolutionary way
> to go.
I agree this is most definitely the place to start.
Call it a Virtual SilkOS if you want.
I think it also warrants some consideration
of making this system run on top of existing Lisp systems as well.
Certainly Common Lisp ones, but maybe even Scheme too if possible.
Why? Because then we can also take advantage of the efforts of
the commercial Lisp vendors (are there any significant commercial
Scheme vendors?)
Do we want "replace" their commercial products with our free ones?
Or can we also support and leverage their efforts as well?
I have much more to say about this, but will leave it here for now.
I mostly agree with Mike McDonalds two-prong strategy,
we can get started today with ACL on Linux.
Start working on replacing the operating system user interfaces
with web browser interfaces.
For example, a mail interface like http://www.hotmail.com
should be implemented in Silk. As well as a SMTP server written in Lisp.
They can both be completely persistent object-oriented,
and eliminate all
This is a good start because email is central to our lives ;)
and so we can't just play around -- we must can get this to be really,
honestly work and be useable. We will be forced to walk-the-walk
and not just talk-the-talk. Now we can wait till it is debugged
before switch over to it fulltime.
At this point, I can say that I have already written a
persistent-object web server in SilkScript, and this web server has
been the public web server for www.franz.com for a while now.
The latest non-interuppted-24-hours-7-days-a-week stretch is 37 days.
This software does not run under Linux however, and uses AllegroStore,
which doesn't have the low-level implementation I believe we need.
-Kelly Edward Murray