Printers Printing some PDFs as Garbage Characters

[John Five]

> CUPS *does* support TBCP, but not on input - it'll just generate it.
> If you pass in data with TBCP encoding (which by definition is
> printer-specific), it needs to be printed in "raw" mode...

Should I enable TBCP for the "PS Binary Protocol" option in the CUPS web interface for each printer that is having problems? Will this solve my problem?

I'm assuming not, based on what you said above. What about enabling RAW printing, will that solve my problem?

If I enable raw printing, will non-postscript (e.g. PCL-only) printers choke on everything sent from Windows that goes through CUPS?

Also, if I enable raw printing by editing mime.convs and mime.types, should I also add the "server_args = -o document-format=application/octet-stream" line to my cups-lpd xinetd config file? I read somewhere in the docs that adding this forces CUPS to auto-detect the type of print job being sent via LPR.

Thanks for any insight.