I'm running a 1.0 code kernel on a 486DLC @ 20Mhz
The 0.9 kernel I used to have was keeping slip up for several days
at a time.  Currently slip is dying every 4 or 5 hours.  This is
at 115200 with a 28.8 modem and a hayes ESP board.  Hardware has not
changed at all from 0.9 to 1.0

I have a cron job which cycles once a minute, pinging our connect host,
and resetting the line and redialing if we get no response.  Sometimes
I find core dumps from csh or grep which are used in this process.

Anyone having or not having problems keeping their slip lines up?

-john