Date:        Tue, 03 Feb 2004 08:46:02 -0500
    From:        William Allen Simpson <wsimpson@greendragon.com>
    Message-ID:  <401FA609.5600B9E7@greendragon.com>

  | PPP should "just work".

Actually Bill, there's one PPP question you can probably answer, which
relates to the way I used to use SLIP, which I have never worked out how
to do with PPP.

That is, using SLIP, I could shut down my system, reboot (new system, just
to unglue something wedged, whatever) config the slip interface again,
and just keep on running, without the other end being aware of anything
having happened (of course, connections into my system all died - it
rebooted, but stuff passing through just stalled for a few minutes).

With PPP, I haven't figured out how to get it to start, without going
through the PPP config process (LCP negotiations, authentication, and
all that).   None of that was relevant for me, as I had (note past tense)
a hard wired line (copper from home to work with a modem on each end,
with my end configured to ignore DTR) - nothing ever changed, all I
wanted was a silent IP pipe (like an ethernet would be).   SLIP
provided that.  When I switched to PPP, I could never make it work again.

Maybe it was just the implementation, but to get the LCP stuff to
negotiate, meant I needed to cause the other end to notice that my
end had gone away - which generally meant, causing my modem to drop
carrier to the remote modem for a while, so the remote end would
see the hangup, then go through the modem re-sync, and go back to
expecting PPP startup, which was the only thing I could make my end
do (no data frames till that was done).

There must be a better way, mustn't there?   What is it?

kre