Today I had this message mailed to me by a failing cron job (on Unix):

/usr/bin/env: No such file or directory

But the machine on which the cron job is running, of course, does have /usr/bin/env. This error is being reported by /usr/bin/env due to failing to find something else, rather than being reported by something else due to failing to find /usr/bin/env. I'm increasingly starting to feel that the VMS error reporting syntax isn't as needlessly idiosyncratic and verbose as I always used to think; if the message had been of the order of

%ENV-F-NOSUCHFILE, No such file or directory

then, although I would still be required to work around the programmers' laziness in not bothering to state what file or directory it had failed to find, I at least wouldn't have to stop and wonder which actual program had generated the wretched message!