>> In answer to your questions: Yes, yes, and yes.
>
>
>Alright, is the particular failure mode useful/interesting/detectable?
>Presumably packets arrive at the remote end and fail CRC checks [let's
>see a packet trace]?

Ummm, it's been a while since I looked at this problem, but as I recall,
the packets passed the CRC check just fine (you could see the mangled packets
with tcpdump).  I believe the CRC was calculated after the VJ compression
code on the Netblazer mangled it.

>Perhaps a solution is to have pppd attempt to detect this particular
>sort of failure at startup time and proactively disable rfc1323, or at
>least loudly suggest the user do so?

I'm not sure there's a reliable way to detect these packets.  Maybe the
default should be for rfc1323 to be off?

--Ken