Network forth

I found this in the archive of

!19960405  Prior art for Internet "applet" patents - Eolas, Prodigy, etc.

    There has been much discussion lately about software patents for Internet
"applets" and related technology, the Eolas patent application for one, and
Prodigy's attempt to cash in on Java.
    One of my readers recently forwarded me the following story and example
of prior art that will impact any of the Internet "applet" patents.  It shows
you how tricky this prior art search stuff can be.

    In the very early days of amateur packet radio, a group at the
    Linkoping University in Sweden built an experimental packet radio
    network they called Softnet. The distinguishing feature of Softnet was
    that every network packet was a program, written in Forth, that was
    interpreted by the receiving node. For example, a packet could forward
    itself by prepending a Forth routine to do so.

    The idea was cute and powerful, but the obvious problems of security
    and stability kept it from being accepted by the ham community.

    Here are some relevant quotes from papers presented at the ARRL
    Computer Networking Conferences:

    "In Softnet each node acts as an interpreter of packets containing FORTH
    statements which are immediately executed. The statements are typically
    'treat the rest of the packet as data and forward it to node B', but they
    can also define new functions as 'forward all my packets to node C'."

    (From Jens Zander, SM5HEV, "SOFTNET - PACKET RADIO IN SWEDEN", First
    ARRL Amateur Radio Computer Networking Conference, October 16-17,
    1981. Appears in Pioneer Papers on Packet Radio 1981-1985, ARRL,
    ISBN 0-87259-022-4.)

    "The main concept behind SOFTNET is that all packets are considered
    to be programs of a *network language*. These programs are interpreted
    in the nodes as soon as they arrive. Nodes can be programmed by any
    number of users simultaneously without unwanted interaction..."

    (From Jens Zander and Robert Forchheimer, "SOFTNET - AN APPROACH TO HIGH
    LEVEL PACKET COMMUNICATION", Second ARRL Amateur Radio Computer
    Networking Conference, March 19, 1983. Same volume as above).

    Copies of the ARRL proceedings are available from

    It would indeed be ironic, by the way, if the main societal
    contribution of amateur (ham) radio's early work in packet radio were
    to be the shooting down of the many bogus patents filed by commercial
    companies over a decade later when they finally (re)discovered what
    we hobbyists had been quietly doing all along.

In fact, there are a fair number of issued and pending patents for which
these papers are relevant prior art.  The obscurity of the reference (as
I hadn't thought of searching the amateur radio world for software prior
art), illustrates the complexity of finding, acquiring, organizing, and
distributing software prior art, a complexity missing from current sanctioned
efforts to deal with software prior art.  And as my preliminary 1996 software
patent statistics (out next week) will show, the problem continues to get
worse and worse, deliberately so.