Path: utzoo!utgpu!watserv1!watmath!att!att!linac!uwm.edu!cs.utexas.edu!mayoff From: mayoff@cs.utexas.edu (Robert Mayoff) Newsgroups: comp.windows.x Subject: Re: Problem when killing of processes. Message-ID: <997@langtry.cs.utexas.edu> Date: 28 Nov 90 01:32:58 GMT References: <5170@rex.cs.tulane.edu> Organization: Dept of Computer Sciences, UTexas, Austin Lines: 20 In article <5170@rex.cs.tulane.edu> keating@rex.cs.tulane.edu (John W. Keating) writes: > Occasionally, someone will kill of all their processes while running > windows. This causes the console (the terminal curently being used) > to start misbehaving. It logs the user off, gives a login prompt, > and then stops receiving input except for a carriage return and a > new login prompt after each key is hit. One wonders what system in particular you are using. I see this problem quite frequently (to my eternal sorrow) around here on our SPARCstations 1+ and our Suns-2. The problem here is that people kill -9 their X server, or crash it in some way, and fail to restore the state of the keyboard afterward (possibly by also killing the shell script which started the server). The solution in this case in to log in remotely and run the kbd_mode program thus: "kbd_mode -a". This tells the keyboard to report keypresses as ASCII-encoded characters, which is what a Sun expects in console mode. You don't have to be root to do this, by the way. If you're not on a Sun, then this probably won't help. -- /_ rob /_ Fun things to do with UNIX (#12 in a series): / cd /dev; cat mouse # Try this on a Sun. Really!