Practicalites of non-DSC-conformant PostScript

[John A. Murdie]

We've recently purchased two of the new small HP colour printer, the HP LaserJet 2605dn) - because it's the smallest HP colour printer we could find with automatic duplexing. We're printing to one of the printers from Windows XP (driver version PSCRIPT5.DLL version 5.02, apparently the lates) into CUPS and find that for certain documents the printout is mangled and the PostScript error message "ERROR: undefined OFFENDING COMMAND: CMAP-WinCharSetFFFF-H2" is printed after the mangled pages. Printing to file and printing the resultant PostScript with "lpr -o raw" results in a correctly-printed document. My guess is that pstops(1) is destroying non-DSC-conformant PostScript. (I can't post the example documents with a bug report, as they are confidential. Perhaps I could find a non-confidential document that is broken in this way.)

What can one do about this in general? We could set "raw" as default for that printer queue so that Windows could print there, but that would exclude its use by Linux with well-behaved PostScript. Perhaps we could have two queues for each of our printers, one set "raw" for use by Windows, the other a regular queue for our Linux systems, though this is a messy idea - Occam's Law and all that.

(Incidentally, what do other people who use this printer model think of it? We find it terribly slow on the network - perhaps adding more memory to the original 64Mb would help.)

John A. Murdie