Ok, the demand-dial part works fine.  [Thanks Paul!!]  Unfortunately, I
   appear to have some daemon that is keeping me from dropping my connx &
   will re-attach if I have dropped my connx.  (at least it's consistent)

   I've tried killing daemons that look possible, sendmail dies when pppd
   kills the connx.  Root is the only crontab entry and there's nothing in
   there to do this.  I've been trying to get tcpdump (and the like) to 
   track activity on the interface, lsof to see who has which devices open
   and I'm stumped.  Anyone see anything (below) that might be the problem
   with my config?

   Here's the process list:

gumby# ps ax
  PID TT  STAT      TIME COMMAND
    0 ??  DLs    0:00.10 (swapper)
    1 ??  IWs    0:00.37 /sbin/init
    2 ??  DL     0:04.47 (pagedaemon)
   44 ??  IWs    0:00.20 portmap
   52 ??  Is     0:02.27 syslogd
   55 ??  IWs    0:00.37 timed
   69 ??  Ss     0:04.67 update
   71 ??  Is     0:01.10 cron
   75 ??  Ss     0:00.56 routed -q
   81 ??  IWs    0:00.94 inetd
  271 ??  IWs    0:00.66 pppd
  280 ??  IWs    0:00.30 sendmail: accepting connections (sendmail)
  154 p0  Ss     0:05.25 -sh (csh)
  288 p0  R+     0:00.26 ps -ax
  155 p1  IWs    0:02.14 -sh (csh)
  273 p1  S+     0:07.70 telnet aab
  156 p2  IWs+   0:00.90 -sh (csh)
   96 e0  IWs    0:06.23 -csh (csh)
  153 e0  S+     0:59.36 /usr/local/bin/dt

gumby# netstat -rn
Routing tables

Internet:
Destination      Gateway            Flags     Refs     Use    Mtu
Interface
default          192.54.238.91      UG          2     2272      -  ppp0
127.0.0.1        127.0.0.1          UH          1        0      -  lo0
192.54.238.3     127.0.0.1          UGHS        1       24      -  lo0
192.54.238.91    192.54.238.3       UH          1        0      -  ppp0

   ... other than those two and perhaps the logfiles (which I'll send 
   pieces of to anyone interested) I am clueless as to where the problem
   might be.  Help?

   andrew.  (brennan@allegheny.edu)