Chromium in Ubuntu – deb to snap transition
by Alan Pope on 10 October 2019
We have recently announced that we are transitioning the Chromium deb package to the snap in Ubuntu 19.10. Such a transition is not trivial, and there have been many constructive discussions around it, so here we are summarising why we are doing this, how, and the timeline.
Chromium is a very popular web browser, the fully open source counterpart to Google Chrome. On Ubuntu, Chromium is not the default browser, and the package resides in the ‘universe’ section of the archive. Universe contains community-maintained software packages. Despite that, the Ubuntu Desktop Team is committed to packaging and maintaining Chromium because a significant number of users rely on it.
Maintaining a single release of Chromium is a significant time investment for the Ubuntu Desktop Team working with the Ubuntu Security team to deliver updates to each stable release. As the teams support numerous stable releases of Ubuntu, the amount of work is compounded.
Comparing this workload to other Linux distributions which have a single supported rolling release misses the nuance of supporting multiple Long Term Support (LTS) and non-LTS releases.
Google releases a new major version of Chromium every six weeks, with typically several minor versions to address security vulnerabilities in between. Every new stable version has to be built for each supported Ubuntu release − 16.04, 18.04, 19.04 and the upcoming 19.10 − and for all supported architectures (amd64, i386, armhf, arm64).
Additionally, ensuring Chromium even builds (let alone runs) on older releases such as 16.04 can be challenging, as the upstream project often uses new compiler features that are not available on older releases.
In contrast, a snap needs to be built only once per architecture, and will run on all systems that support snapd. This covers all supported Ubuntu releases including 14.04 with Extended Security Maintenance (ESM), as well as other distributions like Debian, Fedora, Mint, and Manjaro.
While this change in packaging for Chromium can allow us to focus developer resources elsewhere, there are additional benefits that packaging as a snap can deliver. Channels in the Snap Store enable publishing multiple versions of Chromium easily under one name. Users can switch between channels to test different versions of the browser. The Snap Store delivers snaps automatically in the background, so users can be confident they’re running up to date software without having to manually manage their updates. We can also publish specific fixes quickly via branches in the Snap Store enabling a fast user & developer turnaround of bug reports. Finally the Chromium snap is strictly confined, which provides additional security assurances for users.
In summary: there are several factors that make Chromium a good candidate to be transitioned to a snap:
- It’s not the default browser in Ubuntu so has lower impact by virtue of having a smaller user-base
- Snaps are explicitly designed to support a high frequency of stable updates
- The upstream project has three release channels (stable, beta, dev) that map nicely to snapd’s default channels (stable, beta, edge). This enables users to easily switch release of Chromium, or indeed have multiple versions installed in parallel
- Having the application strictly confined is an added security layer on top of the browser’s already-robust sand-boxing mechanism
The first release of the Chromium snap happened two years ago, and we’ve come a long way since then. The snap currently has more than 200k users across Ubuntu and more than 30 other Linux distributions. The current version has a few minor issues that we’re working hard to address, but we felt it’s solid and mature enough for a transition. We feel confident that it is time to start transitioning users of the development release (19.10) of Ubuntu to it. We are eager to collect feedback on what works and what doesn’t ahead of the next Long Term Support release of Ubuntu, 20.04.
In 19.10, the chromium-browser deb package (and related packages) have been made a transitional package that contains only wrapper scripts and a desktop file for backwards compatibility. When upgrading or installing the deb package on 19.10, the snap will be downloaded from the Snap Store and installed.
Special care has been taken to not break existing workflows and to make the transition as seamless as possible:
- When running the snap for the first time, an existing Chromium user profile in
$HOME/.config/chromiumwill be imported (provided there is enough disk space)
/usr/bin/are wrappers that call into the respective snap executables
chromedriverhas been patched so that existing selenium scripts should keep working without modifications
- If the user has set Chromium as the default browser, the
chromium-browserwrapper will take care of updating it to the Chromium snap
- Similarly, existing pinned entries in desktop launchers will be updated to point to the snap version (implemented for GNOME Shell and Unity only for now, contributions welcome for other desktop environments)
apporthook has been updated to include relevant information about the snap package and its dependencies
If you’re experimenting with Ubuntu 19.10 then you can try Chromium as a snap and test the transition from the deb package right now. However, you don’t need to wait until the release on the 17th of October to start using the snap and sharing your feedback. Simply run the following commands to be up and running:
snap install chromium
snap run chromium
Once 19.10 is released, we will carefully consider extending the transition to other stable releases, starting with 19.04. This won’t happen until all the important known issues are addressed, of course.
Now is the perfect time to put the snap to the test and report issues and regressions you encounter.
We appreciate all the feedback and commentary we’ve been sent over the last few months as we announced this project. We honestly believe delivering applications as snaps provides significant advantages both to developers and users. We know there may be some rough edges as we work towards the future and will continue to listen to our users as we chart this new journey.