The epson-inkjet-printer-escpr seems to be some problem of the test script design that the test sometimes fails on slow processors due to being to time-consuming, because it works on most architectures and repeating it can make it pass. So I simply re-triggered it on Xenial i386 and Bionic i386 to see whether it passes now. Note that I did not look into the actual test script of this package. At least the output of the test script looks like some standard script also used by CUPS and other printer drivers. This script for sure does not set MaxJobTime=0 in CUPS and so results should not change by this fix. I do not know about LibreOffice and OpenJDK tests.
-- You received this bug notification because you are a member of Desktop Packages, which is subscribed to cups in Ubuntu. https://bugs.launchpad.net/bugs/1804576 Title: MaxJobTime=0 results in jobs being cancelled immediately instead of never Status in CUPS: Fix Released Status in cups package in Ubuntu: Fix Released Status in cups source package in Xenial: Fix Committed Status in cups source package in Bionic: Fix Committed Status in cups source package in Cosmic: Fix Committed Status in cups source package in Disco: Fix Released Status in cups package in Debian: Fix Released Bug description: [Impact] Setting the cupsd option MaxJobTime to 0 should make the server wait indefinitely for the job to be ready for print. Instead, after updating job-cancel-after option with MaxJobTime=0 value it results in immediate cancelling. This leads to problems with using filters that take some time to process - the user needs to set MaxJobTime to a ridiculously high value to ensure the job is not going to get cancelled instead of just disabling the cancelling timeout. [Test Case] 1. Add MaxJobTime 0 option to /etc/cups/cupsd.conf. 2. Setup a filter that takes at least several seconds to process. (please find a sample imagetopdf wrapper introducing 5s delay) 3. Submit a print job matching the filter, e.g. lp -d my-printer someimage.jpg # jpg uses the imagetopdf wrapper Expected result: The job is printed after the 5s delay. Actual result: The job is cancelled. [Regression Potential] The scope of the change is limited to fixing the MaxJobTime handling in scheduler/job.c and scheduler/printers.c. There should be no difference in behavior except for the special value of MaxJobTime=0. [Other Info] Original bug description: When using CUPS filters, these filters can take a few seconds to complete. In this case no documents are allowed to be lost on printing failures, so we used to set "MaxJobTime 0" in cupsd.conf which worked on Ubuntu 14.04. With cups on 18.04, you get the following message in /var/log/cups/error_log whenever the filter takes a little longer: I [12/Nov/2018:14:43:26 +0100] [Job 18] Canceling stuck job after 0 seconds. Then, the job is deleted and lost. "MaxJobTime 0" is documented as "indefinite wait", but apparently cups treats is as "wait almost not at all". This issue appears to have also been filed upstream: https://github.com/apple/cups/issues/5438 Temporary workaround is to set the MaxJobTime to a very large value instead (e.g. 3 years) Trusty is not affected by this bug. To manage notifications about this bug go to: https://bugs.launchpad.net/cups/+bug/1804576/+subscriptions -- Mailing list: https://launchpad.net/~desktop-packages Post to : desktop-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~desktop-packages More help : https://help.launchpad.net/ListHelp