> I just wonder if you're spending your effort in the wrong place.
> Yes, you have some legitimate issues with PPP, but those are
> implementation issues; they have nothing to do with the PPP protocol
> itself.

To an extent.  But on the other hand, "KISS".  PPP is way overweight
for what I really want, which is just a serial network link, manually
configured on each end, about as dumb as possible.  The state
corresponding to LCP and IP{,6}CP is completely uncalled-for here.

Teaching pppd to keep the interface up, and trying to debug either the
protocol or the implementation (whichever it turns out to be) so that
it will come back up reliably, either one of those (never mind both)
looks like more effort than hacking a few dozen extra lines into
if_sl.c.  IOW, the time and effort required to get bits moving is lower
with SLIP.

> I'd personally work on fixing the things in PPP that you don't like;
> you could even interoperate with other PPP implementations, which
> would be a bonus.

Potentially, yes, though all of the places I've used SLIP and wished
for v6 have been with both ends under my control, so it's not clear how
much of a practical matter that is.

> I can't really see anyone else in the known universe who would spend
> any effort on v6/SLIP; that may or may not be an issue, of course,
> but it's something to think about.

Well, who knows, maybe someone else is in my situation and has been
suffering with PPP links that don't always come back up and such.  And
if nothing else, it'll do what *I* want; I've got enough other code
that nobody else in the known universe is likely to be interestd in....

/~\ 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