Bug 351849 - Margins often cut when printing
Summary: Margins often cut when printing
Status: RESOLVED WORKSFORME
Alias: None
Product: okular
Classification: Applications
Component: general (show other bugs)
Version: unspecified
Platform: Arch Linux Linux
: NOR major
Target Milestone: ---
Assignee: Okular developers
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-08-27 13:03 UTC by Dario Bertero
Modified: 2020-06-09 20:10 UTC (History)
6 users (show)

See Also:
Latest Commit:
Version Fixed In:
Sentry Crash Report:


Attachments
Example file that triggers the behavior (81.07 KB, application/pdf)
2015-08-27 13:04 UTC, Dario Bertero
Details
Particular of the printout (815.28 KB, image/jpeg)
2015-08-27 13:05 UTC, Dario Bertero
Details
HP Officejet 7110 ppd (34.59 KB, application/vnd.cups-ppd)
2018-03-24 14:31 UTC, Germano Massullo
Details
affected pdf (503.96 KB, application/x-xz)
2018-03-26 08:13 UTC, Germano Massullo
Details
cups debug log (122.00 KB, text/plain)
2018-03-27 20:53 UTC, Germano Massullo
Details
d01250-001 (29.77 KB, application/x-xz)
2018-04-05 17:12 UTC, Germano Massullo
Details
error_log (158.03 KB, text/plain)
2018-04-06 09:27 UTC, Germano Massullo
Details
Test PDF file generated while trying to reproduce the problem (352.13 KB, application/pdf)
2018-04-08 00:08 UTC, Michael Weghorn
Details
affected document 2 (181.94 KB, application/x-xz)
2018-10-29 16:34 UTC, Germano Massullo
Details
scanned Evince-20.04 and Okular-3.36.1 printings (257.31 KB, image/jpeg)
2020-06-04 15:41 UTC, kolAflash
Details
jobs from /var/spool/cups/ on Debian (Okular is 8, Evince is 9) (127.11 KB, application/x-xz)
2020-06-04 15:43 UTC, kolAflash
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Dario Bertero 2015-08-27 13:03:16 UTC
When printing, some content near the margins is often cut out. Playing with the various printing options does not help. In all these cases, the print preview is correct, but still some of the content is cut.

With another PDF viewing software the printout is correct.

Attached a sample pdf file that triggers the behavior (on left and right margins), and a picture showing a particular of what comes out after printing.


Reproducible: Always

Steps to Reproduce:
1. Print a document like the one attached
2.
3.

Actual Results:  
Some content near the margins in the printout is cut.

Expected Results:  
Correct print as in the print preview.

Printer: HP P4015n
This bug refers to version 0.23.0
Comment 1 Dario Bertero 2015-08-27 13:04:17 UTC
Created attachment 94244 [details]
Example file that triggers the behavior
Comment 2 Dario Bertero 2015-08-27 13:05:05 UTC
Created attachment 94245 [details]
Particular of the printout
Comment 3 James Funk 2016-02-18 17:33:06 UTC
I wonder if my problem is the same? Or is mine a similar problem?  When I try to print the PDF at http://www.ritetemp-thermostats.com/80XX/images/8082C_operation_guide.pdf  , the printout has only part of the page. It is as if the page image had been enlarged, and the part of it gets printed on my 8.5 inch by 11 inch paper.
Comment 4 James Funk 2016-02-18 17:38:20 UTC
(In reply to James Funk from comment #3)
> I wonder if my problem is the same? Or is mine a similar problem?  When I
> try to print the PDF at
> http://www.ritetemp-thermostats.com/80XX/images/8082C_operation_guide.pdf  ,
> the printout has only part of the page. It is as if the page image had been
> enlarged, and the part of it gets printed on my 8.5 inch by 11 inch paper.

I neglected to mention some things. I'm using a Brother HL-5140 laser printer, OpenSuSE 13.2
Comment 5 Rex Dieter 2016-02-22 19:21:51 UTC
What version of cups?  If 1.7.x, maybe a dup of bug #334653 and/or bug #330820
Comment 6 James Funk 2016-02-22 22:01:49 UTC
If I read software manager correctly, CUPS version is 1.5.4-21.9.1 (x86-64)
Comment 7 Michael Weghorn 2018-01-17 16:18:48 UTC
@Dario: Does this problem still exist with a current Okular version?
If so: Can you please attach the PPD file of the printer in use (which should be located at /etc/cups/ppd/<PRINTERNAME>.ppd)?
Comment 8 Germano Massullo 2018-03-24 14:31:30 UTC
Created attachment 111603 [details]
HP Officejet 7110 ppd

okular-17.12.1
I came to this bugreport while experiencing the problem on another PDF I was printing.
In ppd folder there is also a ppd.o file, do you need it too?
Comment 9 Germano Massullo 2018-03-24 14:32:46 UTC
(In reply to Germano Massullo from comment #8)
> Created attachment 111603 [details]
> HP Officejet 7110 ppd
> 
> okular-17.12.1
> I came to this bugreport while experiencing the problem on another PDF I was
> printing.
> In ppd folder there is also a ppd.o file, do you need it too?

I have forgotten to say that my distribution is Fedora 27
Comment 10 Michael Weghorn 2018-03-24 23:17:30 UTC
(In reply to Germano Massullo from comment #8)
> Created attachment 111603 [details]
> HP Officejet 7110 ppd
> 
> okular-17.12.1
> I came to this bugreport while experiencing the problem on another PDF I was
> printing.
> In ppd folder there is also a ppd.o file, do you need it too?

Thanks for attaching the PPD file.
Can you also attach the PDF file with which you are experiencing the problem?

Which page size is selected in the print dialog for the printer?
Comment 11 Germano Massullo 2018-03-26 08:13:04 UTC
Created attachment 111657 [details]
affected pdf

(In reply to Michael Weghorn from comment #10)
> Can you also attach the PDF file with which you are experiencing the problem?

I attached it in a compressed version it to avoid indexing from internet search engines. The cutted parts are:
- upper part of logo in upper margin of the page
- bottom text in bottom part of the page.
I experienced the problem also with other documents.

Evince instead prints the document correctly

> Which page size is selected in the print dialog for the printer?

In HP Device Manager: A4 Autoduplex 210x297mm
in Okular: A4 210x297mm
Comment 12 Michael Weghorn 2018-03-26 20:10:24 UTC
The error_log from CUPS (the printing system) might help to further analyse this. Can you set the 'LogLevel' directive in '/etc/cups/cupsd.conf' to 'LogLevel debug', restart CUPS ('sudo systemctl restart cups.service') and print again?

Depending on the CUPS configuration, the relevant log output will then either go to the file '/var/log/cups/error_log' or can be seen with the command 'sudo journalctl _SYSTEMD_UNIT=cups.service'.

Can you attach the error_log or the output of the above command here?

It would also be helpful if you could attach the file generated by Okular for printing. This one will be located at '/var/spool/cups/d<JOB_ID>_001' (where '<JOB_ID>' is replaced by the actual job ID).
Comment 13 Germano Massullo 2018-03-27 20:53:35 UTC
Created attachment 111689 [details]
cups debug log
Comment 14 Germano Massullo 2018-03-27 20:57:01 UTC
(In reply to Michael Weghorn from comment #12)
> It would also be helpful if you could attach the file generated by Okular
> for printing. This one will be located at '/var/spool/cups/d<JOB_ID>_001'
> (where '<JOB_ID>' is replaced by the actual job ID).

# ls /var/spool/cups/
returns a list of files of c<JOB_ID>
I took the file of lastest print job ID, and its size is 1,5 kB. Is it the file you asked for?

Best regards
Comment 15 Michael Weghorn 2018-03-28 02:44:09 UTC
(In reply to Germano Massullo from comment #14)
> # ls /var/spool/cups/
> returns a list of files of c<JOB_ID>
> I took the file of lastest print job ID, and its size is 1,5 kB. Is it the
> file you asked for?
> 
> Best regards

Thanks for the additional information.

Hm, no the c<JOB_ID> files are something different. You might have to set 'PreserveJobFiles Yes' in '/etc/cups/cupsd.conf' and restart CUPS in order for the "correct" files to be preserved there after printing. Can you try to print again after setting this? (You can also use 'PreserveJobFiles 86400', which will just keep the 'd<JOB_ID> file for one day or reset to 'No' afterwards, since keeping all documents forever might consume quite some space after a while if you print a lot...).

Unfortunately, the cups error log does not contain all of the information I expected. There should usually be more information on the print processing, e.g. some lines with 'argv' in them, e.g. 'argv[5]' would show the options passed to CUPS. Did the print job finish before you copied the file? May I ask you to do this once again to see whether the information is contained then?
Comment 16 Michael Weghorn 2018-04-04 08:35:22 UTC
The information requested in comment 15 is needed to further analyse this issue.
Comment 17 Germano Massullo 2018-04-05 17:12:20 UTC
Created attachment 111855 [details]
d01250-001

Solved with 'PreserveJobFiles Yes' in '/etc/cups/cupsd.conf'
Comment 18 Germano Massullo 2018-04-05 17:13:48 UTC
(In reply to Germano Massullo from comment #17)
> Created attachment 111855 [details]
> d01250-001
> 
> Solved with 'PreserveJobFiles Yes' in '/etc/cups/cupsd.conf'

I mean I managed to get the job file
Comment 19 Michael Weghorn 2018-04-05 20:40:53 UTC
(In reply to Germano Massullo from comment #18)
> (In reply to Germano Massullo from comment #17)
> > Created attachment 111855 [details]
> > d01250-001
> > 
> > Solved with 'PreserveJobFiles Yes' in '/etc/cups/cupsd.conf'
> 
> I mean I managed to get the job file

Thanks for attaching the job file. This is actually the PostScript file that is generated by Okular and passed to the printing system CUPS.
The file looks totally OK in my eyes; you can open it for viewing in Okular.

In my understanding, this suggests that the problem is only introduced during the processing inside CUPS (the printing system, e.g. by one of the CUPS filters or the printer itself). Unless Okular passes any "wrong" options, the problem probably lies outside of Okular's responsibility.

In order to have a closer look at this, it would be necessary to see which options are actually passed to CUPS.

Unfortunately those were not shown in the log output of comment 13. Was this the output of the  command 'sudo journalctl _SYSTEMD_UNIT=cups.service'? If so, could you please try to change the 'ErrorLog' directive in '/etc/cups/cups-files.conf' to 'ErrorLog /var/log/cups/error_log', restart cups ('sudo systemctl restart cups.service'), print again and attach the log file '/var/log/cups/error_log' that should have been created by then. (The mentioned directive makes the output written to this file instead of syslog/journal, in case the previous directive was 'ErrorLog syslog'.)

Some more thoughts:

How is the printer connected (e.g. a locally connected USB printer or a network printer or a printer shared from another host)? Can you please post the output of the command 'lpstat -v' (or 'lpstat -v | grep <YOURPRINTERNAME>')?

Is there any file '/etc/cups/lpoptions' or '$HOME/.cups/lpoptions' present?


What happens if you print the job file attached in comment 17 directly from the command line using the command 'lp -d <YOURPRINTERNAME> d01250-001' from a directory where this file is located? Is the output the same?
Comment 20 Dario Bertero 2018-04-06 02:54:45 UTC
Just to clarify my experience while waiting for his response. It happened to me some time ago with a certain configuration to a network printer managed outside my control, and a certain printer driver (I cannot remember anymore). Only on Okular, Evince and other softwares printed fine. After they have made some obscure updates to the network printing system and changed the authenthication (it is also possible I selected a different driver from CUPS list, there were many applicable to the Hp P4015n), I have not experienced the bug anymore.
Comment 21 Germano Massullo 2018-04-06 09:27:22 UTC
Created attachment 111870 [details]
error_log

I changed the 'ErrorLog' directive in '/etc/cups/cups-files.conf' to 'ErrorLog /var/log/cups/error_log'

# lpstat -v
dispositivo per Officejet_7110: hp:/net/Officejet_7110_series?ip=**removed by me**
Comment 22 Germano Massullo 2018-04-06 09:32:38 UTC
(In reply to Michael Weghorn from comment #19) 
> How is the printer connected (e.g. a locally connected USB printer or a
> network printer or a printer shared from another host)?

Network printer directly connected to the network

> 
> Is there any file '/etc/cups/lpoptions' or '$HOME/.cups/lpoptions' present?

Both yes

> 
> What happens if you print the job file attached in comment 17 directly from
> the command line using the command 'lp -d <YOURPRINTERNAME> d01250-001' from
> a directory where this file is located? Is the output the same?

lp: No such file or directory
Comment 23 Michael Weghorn 2018-04-08 00:08:26 UTC
Created attachment 111896 [details]
Test PDF file generated while trying to reproduce the problem
Comment 24 Michael Weghorn 2018-04-08 00:19:28 UTC
@Dario: Thanks for the update. Good to hear you're no longer having the problem. This also confirms my assumption that the problem is/was somewhere during the print processing.

@Germano: Thanks for the additional info. That new 'error_log' does actually contain the required information and I think I can somewhat reproduce the issue in a Fedora 27 VM now. (not using a real printer and just running part of the CUPS filters and using an external tool to view the generated file, but the result looks like it should be what you described).

Can you confirm that attachment 111896 [details] resembles the output you get on your real printer?


(In reply to Germano Massullo from comment #21)
> Created attachment 111870 [details]
> error_log
> 
> I changed the 'ErrorLog' directive in '/etc/cups/cups-files.conf' to 'ErrorLog/var/log /cups/error_log'

Thanks, this helps!


(In reply to Germano Massullo from comment #22)
> > 
> > Is there any file '/etc/cups/lpoptions' or '$HOME/.cups/lpoptions' present?
> 
> Both yes

Can you please try to delete or temporarily rename those, print again and see whether this resolves the issue?

In case the thing I can reproduce is actually representative for your problem as well, the problem is related to the print options that are passed to CUPS in your setup. I cannot reproduce with the options that are passed when I print the document to my dummy printer in my Fedora VM.


> 
> > 
> > What happens if you print the job file attached in comment 17 directly from
> > the command line using the command 'lp -d <YOURPRINTERNAME> d01250-001' from
> > a directory where this file is located? Is the output the same?
> 
> lp: No such file or directory

Sorry if my instructions were not clear enough. This command would have to be run from a directory where attachment 111855 [details] was downloaded and extracted (I guess you put the original file into a tar.gz archive and it was not like that in the '/var/spool/cups/' directory).
Anyway, in case removing the two 'lpoptions' files resolves your problem, that test is no longer needed in my eyes.
Comment 25 Germano Massullo 2018-04-14 09:11:45 UTC
(In reply to Michael Weghorn from comment #24)
> @Germano: Thanks for the additional info. That new 'error_log' does actually
> contain the required information and I think I can somewhat reproduce the
> issue in a Fedora 27 VM now. (not using a real printer and just running part
> of the CUPS filters and using an external tool to view the generated file,
> but the result looks like it should be what you described).
> 
> Can you confirm that attachment 111896 [details] resembles the output you
> get on your real printer?

Yes it has exactly the same cutted parts (upper and bottom side of the document)


> (In reply to Germano Massullo from comment #22)
> > > 
> > > Is there any file '/etc/cups/lpoptions' or '$HOME/.cups/lpoptions' present?
> > 
> > Both yes
> 
> Can you please try to delete or temporarily rename those, print again and
> see whether this resolves the issue?

# mv /etc/cups/lpoptions /etc/cups/lpoptions_OLD
$ mv .cups/lpoptions .cups/lpoptions_OLD
# systemctl restart cups
then I printed the document and the margins were perfect.

To double check I runned
# mv /etc/cups/lpoptions_OLD /etc/cups/lpoptions
$ mv .cups/lpoptions_OLD .cups/lpoptions
# systemctl restart cups
and I printed again, **BUT** this time the margins were perfect too. I think this is very confusing. How could you explain the perfect margins even resuming old files?

> > > What happens if you print the job file attached in comment 17 directly from
> > > the command line using the command 'lp -d <YOURPRINTERNAME> d01250-001' from
> > > a directory where this file is located? Is the output the same?
> > 
> > lp: No such file or directory
> 
> Sorry if my instructions were not clear enough. This command would have to
> be run from a directory where attachment 111855 [details] was downloaded and
> extracted (I guess you put the original file into a tar.gz archive and it
> was not like that in the '/var/spool/cups/' directory).
> Anyway, in case removing the two 'lpoptions' files resolves your problem,
> that test is no longer needed in my eyes.

Ok so I did not run it
Comment 26 Michael Weghorn 2018-04-16 19:42:32 UTC
(In reply to Germano Massullo from comment #25)
> Yes it has exactly the same cutted parts (upper and bottom side of the
> document)

Thanks for the confirmation.

> > (In reply to Germano Massullo from comment #22)
> # mv /etc/cups/lpoptions /etc/cups/lpoptions_OLD
> $ mv .cups/lpoptions .cups/lpoptions_OLD
> # systemctl restart cups
> then I printed the document and the margins were perfect.

So the problem actually is some print options that are set "outside the control of Okular".


> 
> To double check I runned
> # mv /etc/cups/lpoptions_OLD /etc/cups/lpoptions
> $ mv .cups/lpoptions_OLD .cups/lpoptions
> # systemctl restart cups
> and I printed again, **BUT** this time the margins were perfect too. I think
> this is very confusing. How could you explain the perfect margins even
> resuming old files?

That's really surprising. Does this mean that you can no longer reproduce the original problem at all?
In the end, the print options that are/were effectively set did cause the problem, so you might want to have a look at the CUPS error_log again and compare the value of "argv[5]" (i.e. the options passed to CUPS for the print job) to those that were set previously (where the problem occured) -- and then look at what the two 'lpoptions' files contain and what options are no longer applied.


In any case: Do you think we can close this bug report as "WORKSFORME" or "UPSTREAM", since you don't seem to have the problem any longer and it seems to have been caused by options set outside of Okular?

(We can still discuss on why the options are no longer set after renaming the lpoptions files twice, regardless of whether the bug is open or not.)
Comment 27 Germano Massullo 2018-04-18 21:05:59 UTC
Let me try with some other computers
Comment 28 Christoph Feck 2018-05-09 23:54:28 UTC
Germano, any success with testing on different computers?
Comment 29 Christoph Feck 2018-06-01 01:11:53 UTC
No response, changing status. If you have new information, please add a comment.
Comment 30 Germano Massullo 2018-08-09 09:20:00 UTC
I tried to print using other computer and I had no problems

No problem with (In reply to Michael Weghorn from comment #26)
> (We can still discuss on why the options are no longer set after renaming
> the lpoptions files twice, regardless of whether the bug is open or not.)

Yes I would be interested in discussing this
Comment 31 Michael Weghorn 2018-08-09 09:29:50 UTC
(In reply to Germano Massullo from comment #30)
> I tried to print using other computer and I had no problems

I'm glad to hear this.

> Yes I would be interested in discussing this

OK. Do you have a way to reproduce this in some way? (Otherwise it's impossible to further investigate the issue.)
Comment 32 Germano Massullo 2018-08-09 09:44:58 UTC
(In reply to Michael Weghorn from comment #31)
> (In reply to Germano Massullo from comment #30)
> > Yes I would be interested in discussing this
> 
> OK. Do you have a way to reproduce this in some way? (Otherwise it's
> impossible to further investigate the issue.)

Unfortunately I am no longer able to reproduce the issue :-(
Comment 33 Michael Weghorn 2018-08-09 10:49:12 UTC
(In reply to Germano Massullo from comment #32)
> (In reply to Michael Weghorn from comment #31)
> > (In reply to Germano Massullo from comment #30)
> > > Yes I would be interested in discussing this
> > 
> > OK. Do you have a way to reproduce this in some way? (Otherwise it's
> > impossible to further investigate the issue.)
> 
> Unfortunately I am no longer able to reproduce the issue :-(

Then I'd suggest to just be happy things are working as expected now. Should the problem still reappear at any point in time, there's still the option to investigate it again...
Comment 34 Germano Massullo 2018-10-29 16:34:42 UTC
Created attachment 115965 [details]
affected document 2

Experiencing again this problem on a different computer + different printer.
I attach the PDF, I will provide the other needed infos + tests as soon as possible
Comment 35 Germano Massullo 2018-10-29 16:52:38 UTC
Can you please tell me where I can add the missing PreserveJobFiles entry in my cupsd.conf file?
https://paste.fedoraproject.org/paste/eViZES2mtZ7TaK71JxFa8g
Comment 36 Michael Weghorn 2018-11-06 11:57:11 UTC
(In reply to Germano Massullo from comment #35)
> Can you please tell me where I can add the missing PreserveJobFiles entry in
> my cupsd.conf file?
> https://paste.fedoraproject.org/paste/eViZES2mtZ7TaK71JxFa8g

Due to my late reply, the above URL is no longer valid.
However, you can just add the line "PreserveJobFiles Yes" into your cupsd.conf pretty much anywhere, e.g. at the very top and restart CUPS so that the setting is actually applied.

Can you also attach the /etc/cups/lpoptions and ~/.cups/lpoptions files, if present, as well as the corresponding lines from the CUPS log (as previously, in particular those containing the 'argv[...]' lines)?
Comment 37 Michael Weghorn 2018-11-13 19:36:27 UTC
Setting status to WAITINGFORINFO (s. comment 36).
Comment 38 Bug Janitor Service 2018-11-28 04:46:09 UTC
Dear Bug Submitter,

This bug has been in NEEDSINFO status with no change for at least
15 days. Please provide the requested information as soon as
possible and set the bug status as REPORTED. Due to regular bug
tracker maintenance, if the bug is still in NEEDSINFO status with
no change in 30 days the bug will be closed as RESOLVED > WORKSFORME
due to lack of needed information.

For more information about our bug triaging procedures please read the
wiki located here:
https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging

If you have already provided the requested information, please
mark the bug as REPORTED so that the KDE team knows that the bug is
ready to be confirmed.

Thank you for helping us make KDE software even better for everyone!
Comment 39 Bug Janitor Service 2018-12-13 03:44:30 UTC
This bug has been in NEEDSINFO status with no change for at least
30 days. The bug is now closed as RESOLVED > WORKSFORME
due to lack of needed information.

For more information about our bug triaging procedures please read the
wiki located here:
https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging

Thank you for helping us make KDE software even better for everyone!
Comment 40 kolAflash 2020-06-04 15:41:12 UTC
Created attachment 129046 [details]
scanned Evince-20.04 and Okular-3.36.1 printings

I can reproduce the problem with the original attachment:
https://bugs.kde.org/attachment.cgi?id=94244

I attached a photo. (printed with Debian, see below)
Left on the photo is Evince, right is Okular.
You can clearly see, that Okular cuts the first letter of each line.
Additionally Okular shrinks the document too much.
Evince can print the document without (so much) shrinking.

I have other documents where Okular also cuts the lower end.


I've tested this with Debian and openSUSE and Okular always gets it wrong, and Evince always gets it correct.

Debian-Testing (11 a.k.a Bulleye 2020-06-03)
Okular-20.04 (a.k.a 1.10.0)
Evince-3.36.1
Printer: HP Deskjet 1000 j110 Series, hpcups 3.20.5

openSUSE-15.1
Okular 18.12.3 (1.6.3)
Evince 3.26.0
Printer: HP Deskjet 1000 j110 Series, hpcups 3.12.11
Comment 41 kolAflash 2020-06-04 15:43:34 UTC
Created attachment 129047 [details]
jobs from /var/spool/cups/ on Debian (Okular is 8, Evince is 9)
Comment 42 kolAflash 2020-06-04 16:23:41 UTC
P.S.
The Debian-Testing (Bullseye) system is freshly set up and I changed any Okular or printing dialog settings. So it's all still at default (margins, etc., ...).
Comment 43 Michael Weghorn 2020-06-09 10:07:46 UTC
(In reply to kolAflash from comment #40)
> Created attachment 129046 [details]
> scanned Evince-20.04 and Okular-3.36.1 printings
> 
> I can reproduce the problem with the original attachment:
> https://bugs.kde.org/attachment.cgi?id=94244
> [...]

Please create a new bug report and leave a link here.

(And also attach CUPS's error log with LogLevel set to "debug" in /etc/cups/cupsd.conf, and your printer's PPD, located at /etc/cups/ppd/<PRINTERNAME>.ppd)
Comment 44 kolAflash 2020-06-09 20:10:39 UTC
(In reply to Michael Weghorn from comment #43)
> [...]
> Please create a new bug report and leave a link here.
> 
> (And also attach CUPS's error log with LogLevel set to "debug" in
> /etc/cups/cupsd.conf, and your printer's PPD, located at
> /etc/cups/ppd/<PRINTERNAME>.ppd)

Done:
https://bugs.kde.org/show_bug.cgi?id=422690