Evince print fails with Postscript driver

Bug #151145 reported by Matt Vickers
38
This bug affects 1 person
Affects Status Importance Assigned to Milestone
cupsys (Ubuntu)
Invalid
Undecided
Pascal De Vuyst
Nominated for Hardy by Graham Clenaghan
gtk+2.0 (Ubuntu)
Fix Released
High
Ubuntu Desktop Bugs
Nominated for Hardy by Graham Clenaghan

Bug Description

Binary package hint: evince

Printer: Brother HL-1850

Driver: Foomatic/Postscript (Recommended)

Problem: Printing a pdf from evince results in data being sent to the printer (as evidenced by status light on printer), but the file is never printed. There is an eventual timeout where the job disappears from the queue and the printer data light stops. This occurs with numerous settings (duplex on/off etc) and running evince as root makes no difference.
However, with the same driver and settings, no such problem occurs printing a document from OOo Writer or Impress presentation.

Selecting a different driver from the list such as hpijs or HL1250 allows printing with evince, but not exactly as desired (duplex and resolution problems). The BRScript-3 (for HL1850N) driver displays the same problem as the Postscript driver.

Postscript is the recommended driver and has worked perfectly under previous releases.

I'm unable to tell whether the problem is with evince or the driver, but since it works from other programs I'm assuming evince.

Only one error in the logs, but this occurs even when a job does print so seems not relevant to the problem.

"Oct 10 12:17:15 gutsy-athlon kernel: [ 2750.424000] audit(1191982635.046:32): type=1503 operation="inode_permission" requested_mask="rw" denied_mask="rw" name="/dev/tty" pid=7512 profile="/usr/sbin/cupsd" "

Revision history for this message
Sebastien Bacher (seb128) wrote :

Thank you for your bug. Does printing from eog works correctly?

Changed in evince:
assignee: nobody → desktop-bugs
importance: Undecided → Low
status: New → Incomplete
Revision history for this message
Matt Vickers (mattyv) wrote :

Thanks for taking a look Sebastien.

I attempted to print with eog, but exactly the same problem was exhibited, the printer received data but never printed. Switching drivers again worked. It is beginning to look like OpenOffice is the odd one out in that it WILL print.

Obviously it's not only evince's problem, so thanks for helping get closer to the cause. I haven't had experience troubleshooting printing much before, so please let me know how to help and I will keep trying.

Revision history for this message
Sebastien Bacher (seb128) wrote :

Likely a GtkPrint or cups bug then

Changed in evince:
status: Incomplete → New
Revision history for this message
Thom Pischke (thom-pischke) wrote :

Seeing something very similar here. Have three ubuntu laptops, one running as a print server with feisty, and two clients running gutsy. After upgrade to gutsy, printing PDFs works very poorly. I've only manage to print one PDF since upgrading, and I managed that only by sending the job from the server itself running feisty.

When sending jobs from a laptop running gutsy, the job just hangs in the Queue. Sometimes the printer light blinks to show it's processing a document, sometimes not. Either way, nothing seems to ever come out.

I have managed to print non-pdf data.

I tried acrobat reader, more or less the same problem. Certainly the same result: Can't print PDFs.

I have an HP Laserjet 1320 and use the PostScript driver.

Revision history for this message
Dmtiry Shatrov (erdizz) wrote :

Evince doesn't print for me, too. I've got a fairly fresh Ubuntu 7.10 install and HP1018 printer.
The printer didn't work out of the box, so I had to follow instructions at http://foo2zjs.rkkda.com/
Now, gedit prints fine, kpdf prints fine, and evince doesn't print anything at all. The printing dialog appears as usual, nothing is grayed out, but after pressing the "print" button nothing happens.

Revision history for this message
Yumi (tempreg) wrote : Re: Evince print fails with Foomatict driver

Gutsy prints from Openoffice to my Laserjet1100. Printing a PDF file with Evince or Acrobad Reader and it hangs in the queue status "processed" seemingly forever.

Revision history for this message
alnicolli (lentakban) wrote :

Gutsy does'nt print PDF from evince with an HP1020, while others applications can print properly.

Revision history for this message
alnicolli (lentakban) wrote :

Reinstalled Evince, now it print.

Revision history for this message
Stuart Rossiter (stuartr) wrote :

I've had similar problems with Gutsy (and pretty sure the same problems on Feisty).

The last time this occurred was printing a PDF converted from a Word file via OpenOffice. Hung processing on the printer (and in the queue) until socket timed out, no matter what print client used (Evince, Acrobat). Eventually got it to print from OpenOffice but this had failed previously (perhaps OpenOffice always worked but I had clogged up the queue or something....?).

In some cases, prints off 1 or 2 pages before hanging (in CUPS, can see it processing 1 page, 2 pages, etc. *very* slowly before the timeout).

My suspicion would be that it's some specific syntax within the PDF file that then translates into dodgy PostScript or similar. Having said that, had the same issues using a PostScript driver and a Gutenprint one (to an HP Laserjet 2300). Still appears to apply to only certain PDF files (have printed others OK, including ones converted to PDF via OpenOffice).

If it helps, attach PDF formats of file that caused problems (c. 300KB) and will attach Word version in separate post...

Cheers,
Stuart

Revision history for this message
Stuart Rossiter (stuartr) wrote :

And the Word original....

Cheers,
Stuart

Revision history for this message
Braice (commercial-braice) wrote :

I have the same issue since I updated evince from 0.8-0.9 to 2.20

I can print well with acrobat reader

I've got a Brother HL-5250DN with BR-script 3 drivers

The printer receive the data but prints nothing

The is no error in cups logs

I never succed to print something with the new evince version

I've got the same issue with ubuntu and gentoo

Revision history for this message
sja821 (sja821) wrote :

Printing pdfs from evince fails for me as well. All of my printers are networked with native postscript. Evince hangs at 100% cpu after printing until I exit evince. I can print to file (postscript format) and then lpr the ps file, so it is doing the hard parts right. When running evince from the command line, I get a message that the "file is empty" when I try to print to the printer. The same printers worked properly in edgy (I skipped feisty).

Revision history for this message
huckfinn_ca (huckfinn-ca) wrote :

I'll go along with most of what has been written here above, and I never had problems before upgrading to 7.10:

I'm running Ubuntu 7.10, Evince is 2.20.1, and have a networked Brother 7820N using a BR-script 3 driver. Evince creates a hang trying to print "postcript format" stuff to the printer after taking an long time on the print queue (>1min) for just 1 page. After finishing in the queue the Brother 7820N does print an error on the paper, and nothing from the queue:
ERROR NAME;
   ioerror
COMMAND;
   fill
OPERAND STACK;

However if I print the file using the lpr command. I'm successful. Yet it still takes a long time on the print queue(>1min).

Revision history for this message
Christopher Hübner (huebnerc) wrote :

I have the same problems using the pdfs from http://www.mvv-muenchen.de.
I am using an old LaserJet 4 over Network (JetDirect). The timetables are no problem, but the footpath (graphic-pdf) won't print. With kpdf printing works fine.

Revision history for this message
Mark Van den Borre (markvdb) wrote :

* printers: laserjet 4000N, 4050N, 4100N, 16 Mb ram, all PostScript with JetDirect module
* OS: clean 7.10 install, three different machines
* similar symptoms: evince and other gtkprint applications fail to print all but very simple and short documents

According to my interpretation of https://wiki.ubuntu.com/Bugs/Importance, this bug should be reclassified critical, or at least high priority:
* users with ethernet postscript printers constitute a large portion of all ubuntu users
* this bug has a severe impact on our ability to use ubuntu

Revision history for this message
Alex Mandel (wildintellect) wrote :

Agreed this should get bumped up. I noticed issues before gusty, anything in evince or kghostview with graphics results in an Error Operand stack message being printed. Just tried Kpdf on a Gnome install and it took a while but worked with the BR3 script driver and the foomatic postscript ppd.

Definitely an issue in the gnome/cups print stack.

Brother 5150D

Revision history for this message
Stephen D Kamm (s-kamm) wrote :

evince 2.20.1

Samsung CLP-300N

Won't print:

Revision history for this message
Stephen D Kamm (s-kamm) wrote :

evince 2.20.1

Samsung CLP-300N

Will print

Revision history for this message
Stephen D Kamm (s-kamm) wrote :

(Sorry about multiple posts - It is not clear to me how to include multiple attachments in single comment or edit previous comments.)

In the "Won't print:" case, evince creates blank "output.ps" and blank preview, as well as not printing.

Given these are rebate forms, I'd agree this is critical.

Revision history for this message
Dave (dave-fsinfo) wrote :

I can confirm the error when printing from evince (on a freshly installed ubuntu 7.10).

With a Foomatic-based printer driver, I get ghostscript errors in my CUPS error_log (if logging is set to debug)...
------------
D [25/Feb/2008:18:15:49 +0100] [Job 9] Error: /rangecheck in --xyshow--
D [25/Feb/2008:18:15:49 +0100] [Job 9] Operand stack:
D [25/Feb/2008:18:15:49 +0100] [Job 9] (\034\023\035\024\036\037\007\033\036\022-\021\027\022) --nostringval--
D [25/Feb/2008:18:15:49 +0100] [Job 9] Execution stack:
D [25/Feb/2008:18:15:49 +0100] [Job 9] %interp_exit .runexec2 --nostringval-- --nostringval-- --nostringval-- 2 %stopped_push --n
ostringval-- --nostringval-- --nostringval-- false 1 %stopped_push 1889 1 3 %oparray_pop 1888 1 3 %oparray_pop 1872
   1 3 %oparray_pop 1755 1 3 %oparray_pop --nostringval-- %errorexec_pop .runexec2 --nostringval-- --nostringval-- --no
stringval-- 2 %stopped_push --nostringval--
D [25/Feb/2008:18:15:49 +0100] [Job 9] Dictionary stack:
D [25/Feb/2008:18:15:49 +0100] [Job 9] --dict:1161/1684(ro)(G)-- --dict:1/20(G)-- --dict:113/200(L)--
D [25/Feb/2008:18:15:49 +0100] [Job 9] Current allocation mode is local
D [25/Feb/2008:18:15:49 +0100] [Job 9] Last OS error: 2
D [25/Feb/2008:18:15:49 +0100] [Job 9] GPL Ghostscript SVN PRE-RELEASE 8.61: Unrecoverable error, exit code 1
------------
...or, with the CUPS+Gutenprint driver, the printer (a HP LaserJet 4M via JetDirect) prints a postscript error:
  ERROR: rangecheck
  OFFENDING COMMAND: xyshow

Looks like somewhere in the print stack, the xyshow postscript command is generated improperly. My guess would be that evince (or a library used by evince) generates the corrupt postscript code before submitting it to CUPS.
After a bit of googling, it seems to be the same/similar bug as described here: http://lists.cairographics.org/archives/cairo/2007-November/012147.html

KPDF prints the same document just fine.

My estimation is that this bug has a moderate impact on a large number of Ubuntu users, given that it effectively prohibits printing of some PDF documents via the default PDF viewer. Therefore, it should have at least "high" importance. At least to me, it does have ;-)

Revision history for this message
TerryG (tgalati4) wrote :

Marking as Confirmed to preserve comments.

Changed in cupsys:
status: New → Confirmed
Bryce Harrington (bryce)
Changed in cupsys:
importance: Undecided → Critical
Revision history for this message
Pascal De Vuyst (pascal-devuyst) wrote :

Matt Vickers,
Please provide the information described here: https://wiki.ubuntu.com/DebuggingPrintingProblems.

Others,
These kind of bugs are very hardware specific and more than likely the problem reported by Matt Vickers will be unrelated to your problem. If you don't have Brother HL-1850 printer please open a new bug report and provide the information from https://wiki.ubuntu.com/DebuggingPrintingProblems.

Changed in cupsys:
assignee: nobody → pascal-devuyst
status: Confirmed → Incomplete
Revision history for this message
Richard Oxland (roxland-deactivatedaccount) wrote :

I've reported a similar bug for an HP LaserJet 1320n here:

https://bugs.launchpad.net/ubuntu/+source/cupsys/+bug/201033

Revision history for this message
Steffen Krumbholz (skrumbholz) wrote :

Evince fails printing this official german tax form (attached). XPDF and HPLIP-GUI were printing perfectly.

Changed in cupsys:
importance: Critical → Undecided
Changed in gtk+2.0:
importance: Low → Critical
milestone: none → ubuntu-8.04
status: New → Confirmed
Changed in cupsys:
status: Incomplete → Invalid
Revision history for this message
Till Kamppeter (till-kamppeter) wrote :

Corrected status and package assignments: The problem described here is that Ghostscript gets fed with broken PostScript and this leads Ghostscript to exit with an error. The broken PostScript comes from evince or whatever library generates PostScript for evince. The library is supplied by the gtk+2.0 package and therefore I raise the priority for this package. CUPS has no influence on this, therefore I am closing the CUPS task.

Revision history for this message
Sebastien Bacher (seb128) wrote :

did anybody try on hardy? does printing to a ps give you something you can open using ghostscript? I've tried several example on the bug and that works correctly

Changed in gtk+2.0:
status: Confirmed → Incomplete
Revision history for this message
Matt Vickers (mattyv) wrote :

I haven't been able to test for a while (stayed with Feisty on main PC) but will grab the Hardy beta and give it a shot now to see if it has been resolved on the reported equipment (since it's mine).

Sorry for not being active on this bug report recently.

Revision history for this message
Graham Clenaghan (origin191) wrote :

I had this problem in Gutsy with evince-print (installed xpdf for my pdf printing needs, works fine there), and Hardy did nothing to fix this.

Revision history for this message
Sebastien Bacher (seb128) wrote :

that seemed to be a libcairo issue and should be fixed in hardy, does anybody has the issue using the current versions?

Changed in gtk+2.0:
importance: Critical → High
Revision history for this message
Graham Clenaghan (origin191) wrote :

As I said previously, I have a computer running Hardy beta and I still have this issue with evince and the hpijs driver -- I can test other applications if you want, I can tell you that firefox, openoffice, and xpdf work, I dont print much else.

Revision history for this message
Matt Vickers (mattyv) wrote :

It now seems fixed/almost fixed. The first PDF I tried (see below) took a very long time between data being sent and a page coming out. Each subsequent page also took some time and the last three didn't print at all (duplex on). However I tried another PDF of similar size and it printed quite quickly with no missing pages - also with duplex. Perhaps the slow printing, for whatever reason it occurred, caused a timeout and thus left the last pages off. Nothing in the cups log to indicate why it failed.

After that, I printed an image with EOG, it was fast and successful.

Evince 2.22.0
Brother HL-1850 with Postscript driver
PDF full text article available for free: http://jssm.org/vol6/n4/19/v6n4-19abst.php

Any other applicable info I should provide?

Revision history for this message
Steve Langasek (vorlon) wrote :

According to Sebastien, the original issue of incorrect postscript generation has been resolved for hardy per upstream. If you are having other printing issues, please open separate bug reports for those. In particular, problems when using the hpijs driver seem to be unrelated here.

Changed in gtk+2.0:
status: Incomplete → Fix Released
Revision history for this message
Mike McNally (m101) wrote :

Hate to be a bug necromancer, but this bug describes exactly a problem I've had and continue to have in Karmic. Printing PDF files from "evince" and even Adobe's own Acrobat often results in exactly the same failure: an error printout with an "Operand Error" or "Configuration Error". The error message itself is of no diagnostic help to me.

The exact same PDF file prints fine from a Windows 7 machine via Acrobat Reader.

(Brother MFC-7820N)

Revision history for this message
Andy (pwndy) wrote :

Issue still exists in Ubuntu 10.04.
Trying to print pdf from Evince or Okular is displayed in the print queue, and the printer displays "processing...".
But nothing is printed.
I have tried both Foomatic and BR-Script3 driver for my Brother HL-1870N printer.

Revision history for this message
beggi dot com (beggijons) wrote : Best appz place

HAL kr LOWEEN SA pua LES!
EVERYTHING YOU NEED for a great and exciting day you can buy HERE!
Only TODAY ANY GOODS you can get with a next code 4437 which gives you a 40% dis jh count.
Be prepared for a scary day and make a show!

Get the great di dal scou be nts on popular so hn ft lv wa etj re today at www.greatrivercityone.com.ua
All s oq of xwx wa za re is instantly available to do vuc wnl pf oad - No Need Wait!
ALL OUR SO xkx FTW nkm ARE bs S ON ALL EUROPEAN LANGUAGES -
USA, English, France, Italy, Spanish, German and more!!!

SO kw FTW wf ARE:Windows 7 Ultimate 32 bit99.95Windows 7 Ultimate 64 bit99.95Windows XP Professional with Service Pack 369.95Office Professional Plus 2010 32-bit89.95Office Professional Plus 2010 64-bit89.95Adobe Photoshop CS5.1 Extended99.95Office Professional 200769.95Adobe Acrobat 9 Pro Extended59.95Office Home and Student 200749.95
Also we have so mu lx ch s pk of qe t for MA pg CIN cp TO phv SH!!!Adobe Creative Suite 5.5 Master Collection for MAC269.95Adobe Creative Suite 5.5 Design Premium for MAC219.95Microsoft Office 2008 Standart Edition for MAC119.95Aperture 3 for MAC79.95Adobe Photoshop CS5.1 Extended for MAC89.95
To re gbd vi qj ew full list of the offers, v cvi is qvi it www.greatrivercityone.com.ua

Revision history for this message
caspar_wrede (caspar-wrede) wrote :

Evince fails to print certain PDFs under the following conditions:
* using an HP LaserJet P2015 Series (connected via Wifi)
* Ubuntu 12.04
An empty sheet is printed. Other PDFs print successfully.

Revision history for this message
Till Kamppeter (till-kamppeter) wrote :

caspar_wrede, your problem is bug 994477. Please test the fixed package proposed there and report your result there.

Please do not hijack old bugs.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.