gv: pdf display not operational: error /undefined in copy_trailer_attrs

[veedeehjay@…]

`gv' does not display pdf-files but throws an error. the bug is principally known.

the bug is in file:

/opt/local/share/ghostscript/8.71/lib/pdf2dsc.ps.

here's the diff of the patch solving the problem:

@@ -116,7 +116,7 @@
    DSCfile PDFname write==only
    ( \(r\) file { DELAYSAFER { .setsafe } if } stopped pop\n) puts
    ( pdfopen begin\n) puts
-   ( copy_trailer_attrs\n) puts
+   ( process_trailer_attrs\n) puts
    (%%EndSetup\n) puts
 
    /.hasPageLabels false def % see "Page Labels" in the PDF Reference

[ryandesign (Ryan Carsten Schmidt)]

The file /opt/local/share/ghostscript/8.71/lib/pdf2dsc.ps is part of the port ghostscript, not part of the port gv.

What do you mean "the bug is principally known"? Known by whom? Is there an upstream bug report? If so where? Where did this patch come from? Is this an upstream fix? If you invented this patch yourself, have you already reported it to the developers of ghostscript?

[ryandesign (Ryan Carsten Schmidt)]

Is there any update to this issue? Any responses to my questions above?

[ryandesign (Ryan Carsten Schmidt)]

I found a response the reporter sent by email in August 2010:

The file /opt/local/share/ghostscript/8.71/lib/pdf2dsc.ps is part of the port ghostscript, not part of the port gv.

right. I did not not recall my old mails. of course this is actually a ghostscript issue. sorry for that.

What do you mean "the bug is principally known"? Known by whom? Is there an upstream bug report? If so where? Where did this patch come from? Is

at the time I found the solution in red hat's bugzilla:
​https://bugzilla.redhat.com/attachment.cgi?id=395186&action=edit

and here is another reference to the problem:
​http://savannah.gnu.org/bugs/?29023

here, too:
​http://groups.google.de/group/comp.lang.postscript/browse_thread/thread/24ade6df75ea896a/5f30f3545e657584?lnk=raot

if I recall correctly back in May I also found some remarks that it indeed had been reported upstream. right now I cannot locate these remarks again.

this an upstream fix? If you invented this patch yourself, have you already reported it to the developers of ghostscript?

no since I presumed that had already happened (it seemed that the problem had already been noted month ago in several linux distros). but maybe this is not correct...

[veedeehjay@…]

Replying to ryandesign@…:

Is there any update to this issue? Any responses to my questions above?

it seems that it has been fixed somewhere upstream and pdf display works at least with ghostscript 9.04 (and /opt/local/share/ghostscript/9.04/lib/pdf2dsc.ps implements the previously reported patch in line 119)

[ryandesign (Ryan Carsten Schmidt)]

Ok, thanks.