Re: MISC-d Digest V99 #97

["GARY B. LAWRENCE" <garyl@xxxxxxxxxxx>]

MISC-d-request@pisa.rockefeller.edu wrote:
> 
> Subject:
> 
> MISC-d Digest                           Volume 99 : Issue 97
> 
> Today's Topics:
>          Re: That alternative web site
>          Re: MISC-d Digest V99 #96
> 
>     ---------------------------------------------------------------
> 
> Subject: Re: That alternative web site
> Date: Wed, 03 Nov 1999 10:40:39 +0000
> From: Luis Commins <Luis@ucea.ac.uk>
> To: misc
> 
> It's at www.m-i-s-c.net. Send your info to me at this address and I will add the data to what I am setting up.
> 
> Luis.
> 
> >>> "Wayne Morellini" <waynemm@hotmail.com> 03/11/99 03:43:02 >>>
> I remember something about an alternative web site to send messages to being
> set up, does anybody have the address.  I think some people in the list
> might have gone south, somehow.  At the same time I don't know whats
> happened with Jeff or ITVC (Just seen the changes to their site, but the
> last news was from 1999).
> 
> Please advice.
> 
> Thanks.
> 
> Wayne.
> 
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
> 
>     ---------------------------------------------------------------
> 
> Subject: Re: MISC-d Digest V99 #96
> Date: Wed, 3 Nov 1999 07:35:56 -0800 (PST)
> From: ut@value.net (Jeff Fox)
> To: MISC
> 
> Hi MISC readers:
> 
> UltraTechnology moved the web site from dnai.com to value.net last month.
> There still seem to be some problems with some domain name servers
> still caching the old address and directing traffic to a non-existent site.
> Until the domain name servers get properly updated you will need to use
> 
> http://www.value.net/~ut for http://www.ultratechnology.com    and
> ut@value.net for fox@ultratechnology.com
> 
> The registered domain name may work from your location.  if there is
> still a problem use the value.net addresses.
> 
> >> It's a shame the UltraTechnology site is down, it would have been nice to
> 
> see above: trace your name servers to see why you are not directed to
> value.net from ultratechnology.com
> 
> >At the same time I don't know whats
> >happened with Jeff or ITVC (Just seen the changes to their site, but the
> >last news was from 1999).
> 
> I think you have a Y2K problem, it still is 1999.
> 
> Jeff Fox
 It is great to see that Jeff Fox has a new website! My link to the 0ld
one worked 
until about 3 weeks ago and I was afraid he had given up. I went to the
new site at value.net/~ut and found all the familiar links. I also
noticed that Jeff has entered some of the text from the first two
videos. I have watched these two videos many times and appreciate the
ability to look at a printed listing of the scource code
while watching the video. I would very much recommend buying the one
with  the demo of Chuck Moore's chip design system. This demo shows a
system that Chuck did down to the assembly level on a clone pc and using
his own code he has made a simple tool structure
that does the design for the masks of his chips. I showed this demo to
members of my local computer group - non forth users - and none of them
understood the great advantage
one can have in terms of flexibility, understanding and modeling when
you can do a chip 
design with your own software. I see a problem that companies like
Netscape have when
developing a browser to run on someone else's  partially hidden software
- Microsoft .
Developers at netscape have a problem finding software errors in the
software if they do
not know if it is caused by logic problems in their scource code or
problems in Windows
or lack of correct information in interfacing to Windows. The recent
finding by the antitrust judge in the Microsoft case bring to light the
problems small software companies have in dealing with large companies
that control the operation system. I believe that Chunk's idea of making
custom silicon systems that handle jobs that can be
outsourced from the large system. This idea if it caught on would
allowed small fast moving companies to link in new ideas without control
of Microsoft or other giants. We
already see something similar on a world wide scale in that the internet
is a link of 
different types of servers using a open and well defined protocall -
tcp/ip. Microsoft
is trying to take as much of this market as possible, but servers run on
open systems such as Linux and Bsd Unix are being implemented, upgraded,
and debugged at a faster rate than even a big company like Microsoft can
match. 
     I can see a situation were custom systems based on Chuck's chips
could do a local net arrangement. They would be hooked to the main pc by
ethernet, or usb or even slow
serial and send commands and recieve data through the same tcp/ip used
externally. Other
people are developing systems like this already, but for instance a web
site on the pic
chip - one of the examples on the web - does not have the ability to do
a full mutitasking graphics display operating system in a compressed rom
of 128-256k . Itv has
stated that their browser-operating system fits into a 128k rom. The F21
chip has the
power to do the graphics and handle some analog inputs and outputs at
the same time. 
In the electronics industry a person could use the F21 with a flash
drive to do jobs done by things like organizers and programmable
caculators and also have inputs and outputs that can be used for a
simple oscilloscope or dvm or function generator. I have seen a pic
based 5mghz oscilloscope in Tech America for about $200.00 . I could see
a F21 doing a much better job with a flash drive logging mode, internet
link, flexible programming modes and other functions like organizers,
caculators, spreadsheets etc.   
    I could even see a case in which F21 systems could be used as
internet servers or
internet custom robots that do particular internet search jobs for
people. Once the hardware became available and cheap, companies using
F21 systems might start to build 
up application wordsets and make some of them open to subcontractors or
open scource.
Companies like Sun and IBM are learning that if they support open
scource software they
may lose some lockin customers but gain a larger base of value added
customers - more
companies that are on the internet who want the large company marketing
and accounting 
systems. We might be able to see a smaller but similar system with F21
systems. 
   One nice advantage with the low fat computers is that they are low
fat computers! I have been learning Linux commands to get a better
understanding of the internet. I am amazed at the complexity that a
person has to conquer to even setup a Linux workstation. The F21 by
having the display processor, the memory manager, some I/O pins, the net
processor and analog I/O processors on chip has alot of the address
decoding on chip. I shudder when I see a Linux box being setup just to
be a router and in order to do this setup it took a good programmer a
few weeks of kernel rebuild to run this system on a 386 with a 8 meg ram
disk. Once a F21 design was developed to do an internet node - rebuilds
to customize should be much easier. In Chuck's chip design he loads the
whole
system into memory as one load, avoiding needing a file system at all.
Since Linux is 
based on Unix it uses the mainframe structure where all actions are done
from a file. 
This system appears to have a better structure than Windows, but you
spend a tremendous amount of time just learning which directories hold
root boot kernel, loading scripts
 device drivers, logs, security bits and so forth. Even in Dos I saw a
tremendous number
 of files holding very small executables like print or format. In a
forth system - even 
the primitive fig 6502 forth 80k floppy drive system I had - you could
have words in memory that would do simple print commands or that could
load a print command from scource code and this code to be changed to
fit new circumstances. Forth systems have
less need for outside compiled code and complex file structures. I do
see that the Linux
 system has a good level of security and a strong system logging and
error structure, but this was based on 30 years of development and
millions of programmer hours. The same amount of effort using good
computer science techniques on a forth system might have yielded much
greater results. One example I notice is when I use the Man pages -
there
is a good computer science listing of options for the command listed and
almost all these commands have a internal parser to deal with this
tremendous number of options.
 I compare this to using F-PC with browse command. When I want to
understand what a Forth command is doing and how it works I hit alt-b
and the scource lising comes up for
the command. The listing shows the stack diagram and sometimes a
comment. If I do not
understand a word in the source I hit the browse key down to that
description and then
return. The only bad part is that I have to know Forth code and intel
assembly code, but
if you are trying to fully understand Linux, you are trying to learn
different script
commands, c code, huge file structures and hundreds of system tools -
some that are there to deal with the complexity of the compiler, the
file structure, the security levels and the error logs. Anyway low fat
computer systems allow new ideas to be developed faster and with a
better understanding.
          Glad to to see Jeff's new website - keep the faith
                      Gary Lawrence