> Why should the *status messages* go to the stderr (AFAIK the stderr > are for error messages)?
That's the UNIX way. The idea is that the output of one program is the input of another program, so only the 'main output' gets written to stdout, while various messages get sent to stderr, so that it can be separately re-directed. Now, we could argue about what the 'main output' of Plucker is, but I'm proposing that it's the document, not the status messages :-). In any case, I haven't removed any existing functionality, just taken a couple of error cases and given them useful meaning, and made it a bit easier for UNIX-heads to grok. Bill
