Bob Nielsen wrote in a message to Mike Bilow:
BN> I realize this borders on heresy, but a more-functional
BN> replacement for KISS is long overdue. At the time KISS was
BN> written by k3mc, ka9q, et al, it was only intended as an
BN> interim solution. Of course, many will say that the
BN> TNC2/Kantronics/AEA, etc., are obsolete as well, but there
BN> are thousands in use.
Not heresy, but a remarkably good idea. Something based on IEEE 802.2 would
get rid of the worst problems of KISS, such as the absence of a CRC, and would
be as extensible as the various flavors of LLC protocols now in use. Finally,
we might be able to reuse a great deal of the existing code already on hand for
processing LLC frames.
Another idea, which I recall being suggested many years ago, was to use PPP and
define a LCP for AX.25. In retrospect, this seems unwise, since AX.25 is
really mostly on interest to us as an encapsulation layer for IP, and this
would lead to a lot of reimplementation.
Anyway, these comments are just off the top of my head, and I don't want anyone
to mistakenly think that I have put enough thought into them to be willing to
defend them in public. On the other hand, I may be right. :)
-- Mike
