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.
Rewrote the fix for blurry desktop wallpaper again to satisfy some fresh upstream reviews. And it’s more elegant. The cogl ABI addition is unchanged still.
Short term: After the above two fixes users can modify a config file to get working splash screens with the Nvidia driver.
Long term: A completely automated solution would require some nontrivial development work in future. Or maybe for us to use SimpleDRM in our kernel builds.
Polishing up and integrating of fixes from the Canonical OEM team to detect RST and Bitlocker and show a UI directing people to help pages for how to fix their system to be able to install in these cases
Reviews of xnox’s fixes to install the OEM kernel properly when needed
Sponsoring for Marcus of atk1.0 to fix a screenreader crash in the installer, see his status for details.
Release: queue reviews, discussions about archive opening, attending a virtual event to prep & ship the 20.04 release (ongoing, will be all this week) and open the GG-cycle
Fixing up proposed-migration tool to correctly not block packages when they’re out of date on riscv64, unblocking a few things which weren’t landing
Handling a Canonical-internal backport request for rabbitmq
snapd PR #5822 (user daemon support) received review approval from mvo, pedronis, and jdstrand this week. I’m hoping that this will finally be merged this week.
To support snapd PR #6258 (dbus activation support), I needed to make /etc/dbus-1/session.d writable on Ubuntu Core systems. That resulted in core18 PR #150 and core20 PR #38, which have both been merged. With those merged, the snapd PR is closer to passing CI. There are a few failures related to the snap-preseed tests, which seem to be due to the branch introducing some new directories to the snapd package.
Once both of the above branches are merged, there will probably be some further work to properly integrate both, and to get to a point where the feature flags can be turned on by default.
snapd portals support:
snapd PR #8356 has been merged. This adds a snap routine file-access command that reports whether a given path can be accessed within a particular snap’s sandbox.
I haven’t started on the xdg-desktop-portal side of this: upstream wants these features to be in a released version of snapd before merging changes, so there isn’t much point starting until 2.45 is out.
Snapcraft:
snapcraft PR #3031 has been merged, which makes SNAPCRAFT_BUILD_INFO work with the Multipass and LXD backends. When this makes it into a Snapcraft release, the snapcore/action-build Github action will be able to generate a snap/manifest.yaml file, which enables security vulnerability checking.
Snapcraft: Regarding the interface through which snapped apps can print via the CUPS snap continued investigation on how to implement in CUPS, found out the single point in the CUPS daemon code where the extra check for the client process ID and whether this is a snap has to happen. Will create a first patch for testing soon.
cups-filters: License change to Apache 2.0 plus (L)GPL2 exception: If the two remaining contributors do not answer or reject, their code contribution is not important (rastertops filter and PDF forms in banner pages) and could be removed if necessary.
cups-filters: Found that the filters are missing bi-level monochrome printing support. This is a dither-free mode which makes all pixels darket than a certain threshold black and lighter than this threshold white, needed for label printers for example to printer bar codes reliably scannable (See this issue).
PAPPL: The Printer Application Framework from Michael Sweet seems to appoach beta as Michael raised the version from 0.2 to 0.9.
gnome-shell: Found a workaround for the regression in keyboard mapping reported last week, the problem only shows when mouse pointer locating by pressing Ctrl is active.
Google Summer of Code 2020: Today we have to request the number of student slots from Google. All sub-orgs have done their student project selections and so we will request 15 slots for the Linux Foundation and therein 6 slots for OpenPrinting. Tomorrow Google will tell us how many slots we actually get, usually less than requested dpenednt on total budget of Google for GSoC and total number of selected students from all orgs.
Google Season of Docs 2020: Submitted the applications for the Linux Foundation to Google. Inside the Linux Foundation several sub-orgs have posted project ideas.
Gnome Shell and friends:
Housekeeping:
Plymouth:
firefox
thunderbird
chromium
other