Comment 34 for bug 978120

What normally happens is that someone posts the Postscript that is sent to the printer from the Ghostscript workflow, preferably uncompressed by using the "psdebug=true" option. Ideally, in cases like this, also from the Postscript from the Poppler workflow.

We then go through a process where I cut the Ghostscript file down bit-by-bit, and a willing volunteer sends my edited files to their printer, and tells me whether it fails or not. That way we find the smallest file that will fail.

I may then have to "instrument" the Postscript to get further debug information back to narrow down *exactly* what causes the error, and hoepfully work out why - possibly by comparing it to the Poppler output. Although the Poppler output is often not directly comparable.

Obviously this can be a tiresome process for everyone involved, and can take some time. Equally obviously, the simpler the file we start with, the better!

Unfortunately, I can never remember how to tell CUPS to save the Postscript, instead of sending it to the printer..... I'm not a CUPS guy.