GNOME Bugzilla – Bug 153012
Printers automatically configured as Postscript shouldn't print with -oraw
Last modified: 2006-06-20 19:43:23 UTC
From Debian bug report http://bugs.debian.org/272118: I just printed an image with the GIMP, and was rather surprised to see PostScript code come out of my printer. I looked around the printing dialog a bit, and discovered that it had defaulted to printing raw postscript and telling the print system not to touch it; moreover, there was a stern warning telling me not to touch the -oraw or Bad Things would happen. This is, of course, completely bogus: I already have CUPS configured to translate PostScript to my printer's language, and I would appreciate it if other programs wouldn't second-guess the print system (besides which, the GIMP doesn't even have a driver for my printer). Unsurprisingly, the image printed perfectly once I removed the -oraw. ----- When the Print dialog retrieves a list of printers from CUPS (or wherever, through lpstat), it configures the printer with Postscript Level 2 and the -oraw option. Additionally, the tooltip on the Command text entry box says not to remove the -oraw option or else it will probably fail. AFAICT this is absolutely incorrect. Any modern UNIX print system is supposed to be able to handle PostScript as input and convert it to whatever native format the printer expects it in, or just passes it right through to the printer if it understands PostScript or the print system doesn't know what to do with the PostScript input. According to the Mandrake printing manual, "-oraw overrides the driver which is assigned to the queue by CUPS" which is certainly not the correct behavior when printing in PostScript most of the time.
Well, the functionality of the print plug-in is maintained by the gimp-print people. We will need their advice so this should perhaps better be filed at their bug-tracker.
https://sourceforge.net/tracker/index.php?func=detail&aid=1030681&group_id=1537&atid=101537
From the gimp-print bug-tracker: This issue, which was previously reported as bug 700792 has been fixed in the 5.0 line (currently at beta 2) by completely reworking the printer command mechanism. It is not going to be fixed in the 4.2 line, in part because there's no clean and compatible way to do this in that environment and in part because there are no planned future 4.2 releases. This makes it dependant on bug #127957 then.
Closing as WONTFIX. See bug #127957 for details.