While reading debian-installer/installer/doc/TODO I spotted the following: * There needs to be a centralised logging method. Syslog will do, but we currently have a lot of nasty code that appends messages to /var/log/messages directly. All of that would be converted to pipe it to logger. However, this is a problem, because piping a command to logger loses the exit status of the command. Need some way to run a command, logging the ouput to syslog, w/o losing exit status.
Is this in fact still outstanding? Some time ago (for a very different piece of code -- our backup code in fact), I wrote a small set of functions which runs a command with the stdout (and/or stderr) being piped over where each line can be logged (to a file in my case) and then exits with the status of the child process (which is the advantage over using logger). I extracted the relevant code and made it standalone though it might need a bit of tidying up before I'd want to submit it. I thought I'd check if this might be useful before doing any work on cleaning it up, after all someone else might have already sorted this out years ago... $ ./wrapit -eop "ERR: " -- ls -al /tmp/nonexist ERR: ls: /tmp/nonexist: No such file or directory $ echo $? 1 is this the sort of thing you want (except with the result syslog'd)? -- Jon Jon Peatfield, Computer Officer, DAMTP, University of Cambridge Mail: [EMAIL PROTECTED] Web: http://www.damtp.cam.ac.uk/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
