> In order for a device to work on our network, it needs to do the
following,
> which is kinda neat.  The protocol is pure ASCII, so it's very readable
to
> anybody watching the party line (RS-485).  The master assigns addresses
for
> communication, and auto-detects them.  By default, a slave device must be
> inactive when first powered up.
> 
> A slave will only accept commands after it has received a message
"COPYRIGHT
> XXXX" where XXX is the company name and copyright date.
> 
> A slave will not be spoken to unless it responds to a "COPYRIGHT" query
with a
> valid response, which just happens to be the company name and a valid
copyright
> date for that type device.
> 
> Now tell me, are you going to make a device for somebody when it has to
tell the
> world that its copyright belongs to your competitor in order for it to
work?

If the protocol allows for any garbage characters to precede a valid
response, this could be worked around by sending something like "THIS
PRODUCT IS NOT MADE BY" in front of the required company name.  In fact,
this has actually been done in a similar case... my understanding is that
IBM's official method of recognizing their original EGA graphics card was
to look for the string "IBM, Inc." (or something like that) at a certain
address in the card's ROM.  3rd party EGA cards simply put "Not a product
of " in front of it - satisfying the protocol, without committing a
copyright violation.
        Jason Harper