Hi everyone, below you will find the updates from the Desktop team from the last week.
If you’re interested in discussing a topic please start a thread in the Desktop area of Discourse.
We also have our weekly meeting on IRC. We meet on Tuesday at 13:30 UTC in #ubuntu-desktop on Freenode. There will be an “Any Other Business” section at the end where you are welcome to raise topics. These topics might be discussed during the meeting, or afterwards depending on the time, depth of conversation, topic and so on.
followed @marcustomlinson work on the deb package and ran through the test plan whenever a new version needed testing
Build snap
continued debugging efforts to add mm-common and gtkmm to the build snap. After talking with an upstream gtkmm developer, I found the issue and now these two packages have been added to the build snap successfully and the build snap is currently in the snap store in candidate
triaging issues as @kenvandine found them (thank you for finding them!) with building packages using the gnome-3-34 extension.
PR #7456, adding a client library for the snap session agent has been merged. The final review was mainly focused on some test coverage gaps, and removal of some logging code that could become overly chatty if it was triggered.
I’ve been working on PR #5822 (user daemons support), bringing it up to date. I’ve updated it to work with the systemd API changes that occurred since the branch was started, and have it passing all the existing tests (unit and spread). At present I’m still working to integrate the control of user session daemons via the session agent, and will have to improve the spread tests to demonstrate that the services are started and stopped on package install/removal.
snapd portal support:
Answered a few review questions, but I’ve mainly been focusing on user daemons/dbus activation. The xdg-desktop-portal side implementation is also waiting on a lock down of how we’ll detect snaps on cgroupsv2 systems (such as current Fedora releases).
ghostscript: With the MIR for libopenjpeg2 being solved (9 years after being posted) I tried to sync from Debian, but unfortunately, this is not possible. Ubuntu’s default optimization is -O3 while Debian uses -O2 and ghostscript does not build with -O3 on ppc64el. Asked the Debian maintainer to add the exception in debian/rules to the Debian package but he rejects anything derivative-specific. So we need to live with this excption rule as smallest delta.
fonts-urw-base35: Posted MIR on the attempt to sync ghostscript. MIR got solved, posted bug reports on libwmf and python-reportlab to completely replace unmaintained gsfonts by fonts-urw-base35.
hplip: Investigated bug of HPLIP breaking Python updates and found out that the problem is people installing and uninstalling newest upstream HPLIP to support newest printers. Solutions: Snapping HPLIP completely, even better: driverless printing and scanning.
SANE: Tried to update to 1.0.29 to get the “escl” backend for AirScan (driverless scanning) to get rid all these nasty driver problems, including users messing with HPLIP.
system-config-printer: Removed IPP-over-USB support patch in udef-configure-printer, as we do not need to make exceptions for HP printers any more, thanks to driverless eSCL scanning.
cups-filters: Fixed several bugs in the filters and completed support for zero-page print jobs, with the great help of student candidates for GSoC.
IPP-over-USB: Further tests on the ippusbxd replacement ipp-usb, especially its debianization by the upstream author.
Google Summer of Code 2020: Continued to let student candidates work on assignments (like cups-filters bugs) to evaluate them for this year’s GSoC.