Red Hat Bugzilla – Bug 475583
openoffice.org prints PostScript source
Last modified: 2009-03-05 00:54:46 EST
Description of problem:
When printing a document from openoffice.org (Writer in this case) the PostScript source is printed, rather than the document as it appears on screen.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Start OpenOffice.org Writer.
2. Load or write a document.
3. Print document.
PostScript source is printed in plain text, i.e.:
%%Creator: (OpenOffice.org 3.0)
PostScript should be interpreted and printed.
Printing to PDF and printing the resulting file from evince or Adobe Reader works normally. Printing from gedit works normally. CUPS test pages print normally.
This system was upgraded from Fedora 9, in which OOo printed normally.
This probably needs more info but I don't know what info you may need.
Sounds truly remarkable. Not heard that one before. What about printing to a file as .ps rather than a .pdf, does that cause the same effect ? And what is the type of printer shown under system->administration->printing->right click->properties->make and model
I guess it be mismatched mimetype. Can you
1. attach /etc/cups/mime.types here
2. edit /etc/cups/cupsd.conf: set LogLevel to debug
3. restart CUPS (or run 'service cups reload' as root in terminal)
4. try again the printing
5. attach /var/log/cups/error-<todays date> here
You can also try to install cups-pdf and print to PDF printer: if the problem is between OpenOffice.org and a CUPS filter the resulting PDF file should contain PostScript source too.
might help here to get the exact type of printer in use, i.e. what it says in
system->administration->printing->right_click->properties->make and model
caolanm->twaugh: Heard of anything similar to this happening in F-10 that might give us a clue as how to reproduce it ?
Nope. Getting the error_log, as suggested in comment #2, would be best for diagnosing it.
Incidentally, it might be easier to run the printing troubleshooter (System->Administration->Printing, then Help->Troubleshoot) to collect diagnostic information for this, rather than following the steps in comment #2. Just select the printer you're using, then when it asks you to print a test page do so using OpenOffice.org and mark the job. Then we'll get to see all the job attributes as well.
printing works ok for me in F-10, feel free to re-open if you have additional information that could help reproduce this
Created attachment 333974 [details]
Troubelshooter debug output
I have also run into this problem. I have a Brother HL4040CN and am using the Brother supplied CUPS driver (not foomatic). OpenOffice is the only product that exhibits this problem. Printing with other applications works fine.
Here is the troubleshooter debug output from an OpenOffice print job.
A few observations.
Printing from OpenOffice worked just fine three days ago. I don't know what has changed in the interim.
Printing works for root but not for a non-root user.
When printing from OpenOffice, I get this line in /var/log/messages:
localhost kernel: brcupsconfcl1: segfault at 626f6a20 ip 00000000626f6a20 sp 00000000ffb59e30 error 14
The segfault occurs for both root and non-root.
The line appeared in the logs for the first time today.
It does not appear when printing from other applications.
'D [03/Mar/2009:21:20:49 -0800] [Job 61] /usr/local/Brother/Printer/hl4040cn/cupswrapper/brlpdwrapper_hl4040cn: line 35: PPD=/etc/cups/ppd/HL4040CN.ppd: No such file or directory',
I think this could be the problem. Try to copy PPD file for the printer to /etc/cups/ppd/HL4040CN.ppd .
There is quite a lot of similar problems posted at http://www.cups.org/newsgroups.php?s21+gcups.general+T+Qpostscript; in most cases printing is OK from one application, but wrong (i.e. PS source is printed) from another.
I don't know why the 'No such file or directory' error occurred since the file is there.
The strange thing is that now, without changing anything, printing from OpenOffice is working again. The only difference between now and yesterday is the time that has passed. I hate "magical" fixes.