>> It would be pretty much a no-brainer to extend SLIP to handle IPv6;
> You're aware that SLIP has _no way_ to distinguish multiple L3
> protocols?

Right.  That's what I intend to fix.  And I believe that, as I said,
>> it could even be done compatibly
in that provided you send only v4 packets, it is interoperable with a
stock implementation, and that both v4 and v6 can share the same wire.

> Worse, as it has also no way to distinguish Van Jacobson - compressed
> TCP from normal IP, so the first byte is used to do this decoding -
> with the result that I doubt that the IP version half-byte is
> transmitted transparently.

Yes, if you turn on header compression that first byte is one of the
things that gets compressed out, since it's approximately constant.

That doesn't interfere.  In fact, the mechanism I have in mind could
easily be used to distinguish compressed-header v4 from uncompressed
v4, though of course stock implementations wouldn't then recognize the
resulting compressed-header packets.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B