I recently tried taking care of some rebates that I needed to mail. In the process, I had to print several things. Many of the print outs (especially of Newegg.com order invoices) had missing text. Downgrading to app-text/ghostscript-gpl-8.71-r6 from app-text/ghostscript-gpl-9.00-r1 fixed the issue. I use Gentoo Linux's testing tree on an amd64 system. I use a HP LaserJet 1018. I have attached a sanitized Newegg order invoice for use as a test case. The downstream bug is: http://bugs.gentoo.org/show_bug.cgi?id=341207
Created attachment 7264 [details] A scan of a printout of the testcase Here is a scan of a printout.
Created attachment 7265 [details] Sanitized Order Invoice, HTML + GIF only. It seems that I failed to attach the test case when I opened this bug. Hopefully it will attach itself with this comment.
Please attach PS or PDF file that has problems. Your last attachment has HTML + GIF only. Ghostscript cannot process this directly. The bug may be in the program that generates PS or PDF.
Created attachment 7266 [details] Sanitized Order Invoice Postscript Testcase (In reply to comment #3) > Please attach PS or PDF file that has problems. Your last attachment has HTML + > GIF only. Ghostscript cannot process this directly. The bug may be in the > program that generates PS or PDF. I had Google Chromium output a PostScript file and then printed it via Okular. The text is still missing.
From the discussion in the referenced Gentoo bug, this is a duplicate of bug 691961 (which is marked as WORKSFORME even when it still doesn't work).
(In reply to comment #5) > From the discussion in the referenced Gentoo bug, this is a duplicate of bug > 691961 (which is marked as WORKSFORME even when it still doesn't work). That's because nobody has followed up with any newer information. From the thread: "But, without knowing how CUPS is invoking Ghostscript, there isn't really a lot I can offer in the way of advice. I'm going to close this as 'WORKSFORME', feel free to reopen it if you can find a Ghostscript command line which shows the problem." Running Ghostscript on the file supplied here *also* works correctly (by the way, despite the name and MIME type this is in fact a PDF file). I do not doubt that the printed output is incorrect, but we have no idea how this print out was created. (I'm assuming the file was printed through CUPS, just like the report in #691961) In fact we don't even know for sure that Ghostscript was involved in the print path at all, but the fact that reverting to an earlier build works suggests that it is. Neither this report nor #691961 supply an example command line for Ghostscript which can be used to reproduce the problem. In the absence of this I cannot reproduce the issue (I *have* tried) and if I can't reproduce it its well-nigh impossible to fix it. If someone would please like to use the script supplied by Oldrich (apologies, I can't seem to reproduce the r macron in your name) in comment #14 of bug #691971 then it should be possible to retrieve the command line being sent to Ghostscript by CUPS. That should allow us to reproduce the problem. Any other means of retrieving the Ghostscript command being used is of course also good. In the absence of a Ghostscript command line we aren't going to be able to resolve this. We may need more information, but we can't even tell that at the moment. The only thing which springs to mind is that the operation of SAFER was altered in recent versions of Ghostscript (at the instigation of Linux distributors concerned about security issues) and this *might* be the source of the problem. This is pure speculation of course.
mark worksforme as indicated in comment #6.
I tried printing again today. This time, I used my HP PhotoSmart C3180 printer to print some photos. The colors were incorrect. Futher inspection showed that Gentoo's package manager had upgraded ghostscript-gpl to version 9.0.1. Downgrading to version 8.71 fixed this. This issue also occurred with the HP logo when printing out the HPLIP test page on my HP PhotoSmart C3180. Is it possible for the ghostscript developers to use the HPLIP test page as a test case for this issue? By the way, I have no idea how to intercept the files that are being passed to ghostscript. If anyone has an idea on how to do this via CUPS, please post it.
(In reply to comment #8) > By the way, I have no idea how to intercept the files that are being passed to > ghostscript. If anyone has an idea on how to do this via CUPS, please post it. As mentioned in comment #6 (albeit with a typo in place) comment #14 of this report: http://bugs.ghostscript.com/show_bug.cgi?id=691961 shows how to capture the GS output and command being sent to Ghostscript, this article: https://wiki.edubuntu.org/DebuggingPrintingProblems#Capturing%20print%20job%20data describes how to capture a CUPS input job.