[ColorForth] Chuck Moore to do presentation on OKAD II in Color Forth 4/14
- Subject: [ColorForth] Chuck Moore to do presentation on OKAD II in Color Forth 4/14
- From: Jeff Fox <fox@xxxxxxxxxxxxxxxxxxx>
- Date: Fri, 13 Apr 2001 22:00:39 -0700
- Organization: UltraTechnology
Dear ColorForth list readers:
Mike wrote:
> I hope you get some good tapes. I just checked out http://www.enumera.com/
> to see what they are saying, because I never heard of them before. It was
> pretty clear that they were working on productizing a MISC chip, without
> giving away many details. They seem strongly attached to FreeBSD... would
John Sokol is the person starting Enumera, he was also the person
who put BSD into the public domain way back when. He was already
involved in the other parallel computing projects listed when
I brought Chuck down to the Silicon Valley Entrepenuer's Connection
(formerly the Parallel Processing Connection) to do a presentation
last summer. The specifics of the chip work Chuck has been doing
has been proprietary so I am interested to see how much Chuck will
make public. I generally have to wait for him to talk about
something in public before I can discuss it in public.
Over the years we have frightened and upset a number of people
by talking about chips that look like the ones they use (same
size, same cost, same package, and some of the same built in
I/O devices) but instead of needing 100 cycles of a 1-10Mhz
clock for a typical Forth word (.1 to .01 Forth mips) Chuck
has been talking 100 to 500 Forth mips. So having a CPU
and I/O coprocessors that are 1000x faster than what people
are used to has been a problem. The other people using CPU
that are hundreds or thousands of times bigger, more complex,
more expensive, more power hungry, and harder to program have
also been upset about Chuck making fast cheap chips and
small simple Forths. Chuck has upset and frightened a lot
of people.
At times people have complained that I "use too many zeros" and
like to remove two or three of them from my numbers. So while
the numbers are proprietary at this time I will say that Chuck
is going to make this situation much worse by adding a lot more
zeros. It frightens me sometimes and I am currious to see
what other people's reactions will be when the numbers are
released.
> they try to run FreeBSD on MISC? Perhaps NetBSD would be easier to port...
> but either seems too big to really fit well on MISC.
C and BSD fit into the plans but there is no plan that I
know of to host BSD on a single MISC node.
> Is the Enumera work being done outside iTVC? Or is iTVC dead at this point?
Enumera is working on aquiring rights from iTV, Chuck Moore and
Computer Cowboys, Dr. Ting and Offete Enterprises, and Jeff Fox
and UltraTechnology. iTV is alive but has been focues on other
projects.
> Back to ColorForth: I am currious about how Chuck's latest Color Forth
> handles system-level things like disk I/O,
Reading and writing of disk tracks. Similar to the old block
interface of reading and writing sectors.
> RAM allocation,
Static allocation of some things including Blocks which are just
memory references. Some dynamic allocation I imagine in OKAD,
but I don't think it is abstracted in a conventional way. The
idea is that Chuck is using about 20K of code and has 128M
for data at the present time.
> and interfacing
> to the 3D accellerator.
On his desktop he used the video card 3D accelerator, on the
version running on the laptop there is a software layer to
render the graphics. The machine redraws the screen as a
background task running at about 30 to 70 times a second,
compared to windows that can take several seconds to
redraw the screen.
> How "generic" are these interface routines? I know
> he tends to avoid solving the general problem, but I wonder how flexible
> ColorForth is today for a variety of applications.
Chuck's Color Forth is pretty specialized. It has boot, video,
disk, and keyboard services, and will get sound and network card
interfaces since Chuck has documentation and plans.
He asked to look at a book I have on modern subsonic aerodynamic
simulations. I can imagine that his Color Forth would make a
nice environment to get high speed simulations as the problem
resembles his CAD problems in some ways. But I doubt if it
would be seen as very generic by too many people.
Of course Chuck sees it as completely generic. He says that
not knowing what task he was going to solve he would arm
himself with a couple of K of generic code, known as Color
Forth, and would extend it into an application.
> I don't know if you will
> see this email before the presentation, but if these topics do not come up,
> perhaps you can try to ask a question or two on my behalf. I think others
> may be interested, like the author of Enth, Sean Pringle. Thanks.
I have to get going pretty soon. More later.
Jeff
------------------------
To Unsubscribe from this list, send mail to Mdaemon@xxxxxxxxxxxxxxxxxx with:
unsubscribe ColorForth
as the first and only line within the message body
Problems - List-Admin@xxxxxxxxxxxxxxxxxx
Main ColorForth site - http://www.ultratechnology.com