Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!seismo!rutgers!ames!ucbcad!ucbvax!EDDIE.MIT.EDU!rich From: rich@EDDIE.MIT.EDU (Richard Caloggero) Newsgroups: comp.sys.apollo Subject: Re: SR9.5 bug (or am I mistaken)? Message-ID: <8708211541.AA22410@EDDIE.MIT.EDU> Date: Fri, 21-Aug-87 11:41:49 EDT Article-I.D.: EDDIE.8708211541.AA22410 Posted: Fri Aug 21 11:41:49 1987 Date-Received: Sun, 23-Aug-87 02:27:29 EDT Sender: daemon@ucbvax.BERKELEY.EDU Distribution: world Organization: The ARPA Internet Lines: 46 Hi, thanks for your response. I am, however still baffled... Here is my situation: We have 4 nodes in our net, 3 running sr9.2.5 and one running sr9.5 with bsd4.2 installed on it. I am the system admin, however, since I am blind, I must connect to the net via a dumb terminal over an SIO line. We thus dedicate one node's SIO port to my terminal. To facilitate process manipulation, I have written a DDM (dumb display manager) which functions in a simmilar way to the DM. It runs "underneath" my shells and other processes. I have set things up so that a ctl-C brings me out of the current process back to the DDM (I {rote this before we got UNIX; guess the csh will surfice, although I've gotten sorta attached to DDM). I thus need to have siologin start DDM when I log in. Following the manual, I put a script in `node_data/startup_sio.sh (sr9.2). However, this never got run so I explicitly ran it in my siomonit startup file. When I got your message, I tried putting it in `node_data/siologin/startup_sio.sh as you suggested; still didn't get run. As for my other problem, the "&" symbol works; I just didn't have the -b flag set in /com/sh. However, I should have been set for when I crp onto another node, I arranged things so that the crp command (my own shell script front end to the /com/crp command) runs /com/sh on the other node with the -start option thus executing a script which sets -b. This mass all works fine when crp'ing between sr9.2 nodes; however the -b flag never gets set when crp'ing from an sr9.2 node (to which my terminal is connected) to an sr9.5 node (so I can use unix). One more question: When I am crp'ed onto our unix node and start csh (either directly as /bin/csh, or via /bin/start_csh), it tells me that it can't access the tty. I assume that is because it knows its input is from an ipc connection (crp mailbox). It can't provide the standard job control facilities. Can I get arround this problem? Must I connect my terminal to the unix node directly? I would simply install is on all the nodes, however, with all the bug reports floating arround the newsgroup, I'd rather not commit ourselves to sr9.5 until I'm sure it will work well in our environment. *** Help! *** -- Rich Sorry for the verbosity and poor diction, but I am pressed for time. Thanx in advance!!!