home .. forth .. machineforth mail list archive ..

[MachineForth] 25x Forth engine


At 09:56 AM 11/19/01 -0800, Jeff Fox wrote:
>Dear mail-list readers:
>
>Andy Valencia wrote in the [MachineForth] list:
>
>> I hope this list is still alive!
>
>It has been very quite for a long time.  But I figured
>that Chuck's press via slashdot would eventually result
>in more traffic in the MachineForth list.
>
>The SVFIG chapter's Forth day was held last Saturday ...

Many thanks Jeff for your detailed account of the current situation.

Quietness on the list does not necessarily mean lack of interest.  As I am
currently studying networking and looking at possibilities for building
simple, cheap and useful network server appliances, Chuck's essential
approach and activities seem as relevant as ever.

Paradoxically, the mass market movement in the direction of Ethernet at
Layer 2 and IP at Layer 3 --- driven by the dramatic lowering of costs
which connectionless packet-switching makes possible -- should also open
the door for low-cost server power at Layers 4 to 7, an area where Linux is
currently making large gains, but Chuck's visions should ultimately take over.

I don't know which routes others are planning to follow in this process,
which is a long one, but I am going to factor my existing Forth-based
software (which has been through a Novix phase) into GUI, business logic
(middleware), and replicated database on separate network machines at both
Layer 2 (Data Link) and Layer 3 (Network) levels.  Layer 2 for relatively
high bandwidth, Layer 3 for internetworking.  Then, I am going to replace
the GUI with any standard browser anywhere in the network, and the
replicated database with a mixture of my own code and cheap, open database
software or optionally any data base software that PHP will talk to.
Leaving the Forth middleware as the interesting service for me to develop
on ever-cheaper hardware.  And new middleware.

This transition may give a market opportunity for Chuck's chips in
multiplexing, coordinating and replicating a Storage Area Network.

There are currently some very cheap processor chips out there for the sort
of application I see first, which I will port my existing stuff to
hopefully in the new year, after first factoring it to run in 256 Kbytes on
old DOS boxes with packet drivers and my existing variant of Fig Forth.

The issue as I see it for bringing Chuck's designs to bear on this
development, is new applications that are very compute-intensive and very
parallel.  I believe that once we have in place both more network
infrastructure (bandwidth) and suitable software design for layers 4 to 7
that complements existing market/software/hardware developments in Layers 2
and 3, there will be a place for such new applications and that Chuck's
designs will then win as the cheapest platform on which to deploy them.

In the meantime, language development that supports parallel computing in a
connectionless, networked kind of way would allow us to move in the
direction of such new applications on whatever hardware we are currently
using (also a matter of cost).

This may all seem very vague, but the market is moving slowly (24 years now
since the first academic visions of specialized network nodes/servers) and
the thing is to get it right by experimentation and cooperation, I believe.
 Such experimentation is enabled by not standardizing the parts (Layers 4
to 7) that are not yet right, and that is a large area.  Instead it should
be aimed at the most dramatic cost reductions possible, enabled by the
simplest and most flexible architecture possible.  This will not be
achieved by standardizing, but by painstaking and time-consuming analysis
of particular key applications that are going to be valuable in the future,
removing all the irrelevant parts (GUI, database, replication...) until the
innermost loops can be redesigned to suit the (redesigned) hardware.  An
iterative process and not a short one.

Best wishes, David Walker


------------------------

To Unsubscribe from this list, send mail to Mdaemon@xxxxxxxxxxxxxxxxxx with:
unsubscribe MachineForth
as the first and only line within the message body
Problems   -   List-Admin@xxxxxxxxxxxxxxxxxx
Main Machine Forth site   -   http://www.