Testing the waters.

Drew R Whitehouse Drew.Whitehouse@anu.edu.au
08 May 1997 08:47:12 +1000

>>>>> "ET" == ET  <emergent@eval-apply.com> writes:

    ET> Suppose my company were interested in developing LispOS in
    ET> this manner: Native i386 code, no underlying unix (lisp down
    ET> to the metal), start with Scheme as basis language, design
    ET> with support for persistent object system, implementation
    ET> tools running on '95/NT, free software.

    ET> How many people would be interested in actually investing time
    ET> and energy on this?  If there is a critical mass, I'll start
    ET> the project.

	Yay, this is what I've been waiting for. I'm in favour of
scheme as a base. My hunch is that common lisp has altogether too much
historical baggage to make it useful in a new clean efficient OS. I
wouldn't worry about the followup post that questions the lack of
scheme standard libraries. If we were to wait for scheme standard
*anythings* to go through the language lawyer mill before acting we'd
be waiting for the next few centuries before we saw the extensions
that will no doubt be necessary for an entire "scheme machine".

	One big worry is that there is a heck of a lot of devices
floating around out there. Is it your idea to somehow harness 95/NT
device drivers ? Or maybe we could all agree on getting together and
using similar systems that have the same common (cheap) hardware
devices (eg IDE disk + S3 Trio) ?

	I notice Tom Lord popped with one of his cryptic sporadic
messages on the guile list recently with a proposal something like
this (subject was something like "cheap cheap cheap"). 

	Are there any decent native i386 scheme compilers at the
moment ? 

	Some guiding thoughts for a scheme OS could be - new, small,
clean, quick, pragmatic.


;; mailto:Drew.Whitehouse@anu.edu.au   http://anusf.anu.edu.au/~drw900/
;; Viz programmer, Australian National University Supercomputer Facility
;; Unsolicited commercial e-mail will be billed at $250/message.