John may be on to something (or maybe just "on" something ;-) ). My opinion is that you should always name the interpretor at the top of a script as he suggests. Heck it can't hurt to try. Jon Schewe wrote: >>>>>>"JF" == Johnny Fulcrum <johnnyfulcrum at mn.rr.com> writes: > > > JF> On Fri, 21 Nov 2003 10:13:40 -0600, Jon Schewe <jpschewe at mtu.net> wrote: > >> I've got some shell scripts that I've put on my desktop in KDE. Some of > >> them execute just fine, however when I click on others they just bring > >> up a > >> dialog box that says "subjective.sh not found" where subjective.sh is the > >> name of the shell script. > > JF> Humm.. do the ones that work have a #!/usr/bin/bash (or whatever > JF> shell/interpeter you're using) as the first line? > > Actually none of them have that. If I write a shell script that's like so: > echo `pwd` > /tmp/foo > That works. However if it's something like this: > echo `pwd` > /tmp/foo > echo $PATH >> /tmp/foo > java -classpath foo.jar foo > It doesn't work, however it does write to /tmp/foo and java is in the > path. If I open up a shell propmpt and type ./foo.sh (containing either > set of commands) it works just fine. > > -- Eric (Rick) Meyerhoff _______________________________________________ TCLUG Mailing List - Minneapolis/St. Paul, Minnesota http://www.mn-linux.org tclug-list at mn-linux.org https://mailman.real-time.com/mailman/listinfo/tclug-list