It's time for the 14th UBports Community Update! We didn't have Jan or Dalton this time, but that doesn't matter because Marius, Flo, and Wayne are great.
Thanks to our sponsors
Right now, as I type this post, our Patreon monthly contributions are sitting at $2,268. Thanks to every one of you.
We'd also like to thank Smoose, Private Internet Access, and Digital Ocean. Their support in monetary donations or services give us more time to develop.
Also, an extra-special thank you to Packet.net for allowing us to use a prerelease ARMv8 server free of charge as part of the Works on ARM project. It's been an incredible boon to our workflow, allowing us to build huge projects in minutes rather than hours. For example, Mir is building in ten minutes, where it took 3 hours on previous infrastructure.
We say it every time, but we really mean it. We wouldn't be here without you.
The Librem 5 was funded
The entire Linux community seems to have exploded at the idea of the Librem 5 phone from Purism. We are happy to see their successful funding, but still feeling cautious about the project. For example, we have serious concerns about the development roadmap.
Ubuntu Touch should run (with only a few modifications) on the 5, but we aren't really digging in to it until the world sees the hardware.
The very astute source sniffers around the community have been seeing one word quite a bit recently.
But how does it work? Why is it here? Why is it called Sudoku?
Project Sudoku is our latest plan to fast-track the development of an image based on Ubuntu 16.04. It involves taking a majority of the Unity 8 / Mir source that is running right now on 15.04 and rebuilding it on 16.04. Easy, right?
Actually, it seems like this will be much easier than continuing all of the conversion work from the final Unity 8 sources back to phones. At the end of Unity's life inside Canonical, the convergence vision was (mostly) dropped in the hope of bringing it to market faster. This included some work to convert Upstart scripts to SystemD units, rewriting slow QML modules in C++, and some other extremely thoughtful and useful additions. Unfortunately, none of this work was completed and we're left with something that's... half baked. Marius sent me a picture of two people walking up some stairs and then falling back down to illustrate the difficulty of making Xenial work.
By going back to 15.04 sources, we forego all of the improvements, but also all of the hardship of completing these conversions. This is definitely a huge compromise, but one that we hope can get ourselves on a more stable, updated, and secure base OS.
Sudoku refers to the idea that we need to take parts and ideas from everywhere we can to complete the puzzle that is Ubuntu Touch. It's also a prime example of why you don't let developers name things.
Right now, testing is being done on the Oneplus 3 and other experimental (Android 7.1 with CAF sources) devices. Mir and QtMir work. Unity 8 starts sometimes, if you don't breathe on it incorrectly.
If you're interested in the technical side of Halium, check out its website.
Flo has been working with the community to improve the Telegram app for Ubuntu Touch, and is pleased to announce that supergroup support is in the pipeline. Supergroups are appearing in the app's internals, but displaying and interacting with them isn't working so well yet.
Flo would like to thank all of the contributors that have been working with him on making Telegram for Ubuntu Touch an excellent chat experience.
It appears that Telegram is almost the only app that Canonical's notification service is used for at this time. We're expecting this service to be shut down at the end of the year.
Telegram is willing to change the endpoint for their notifications if we are willing to receive them and have a service with a compatible API to Canonical's. This is definitely not ideal as the current service has proven to be rather slow, unreliable, and doesn't seem to be very private. All messages are sent in the clear to the server, which relays them to the device. We don't want the responsibility of protecting the message data of every user. We also don't like the idea of requiring people to have explicit trust in us before using a third-party service.
OTA-3 Most wanted features
On September 2, Marius created a thread asking for the community's most wanted features in OTA 3. 142 posts later, we had a large set of possible (and some impossible) requests.
The most-requested improvement to Ubuntu Touch was Bluetooth stability. This was closely followed by being able to move the text cursor around in an easier way.
We're happy to report that the second request has been fulfilled. Users who are using the Devel channel will find that they can press and hold on the keyboard then swipe left or right to move the text cursor.
Bluetooth fixes will prove to be a bit more challenging. It's no secret that desktop Linux has trouble with Bluetooth. Since we just have a desktop Linux base (and a quite old one at that) under the hood, it's no wonder that we have major issues as well. We hope that a quick jump to 16.04 (possibly aided by Sudoku) will help us be able to focus more on backporting Bluetooth features from newer versions of Ubuntu.
Other notable requests were better battery life, an update to the browser, and running Android apps.
Mir, in the future
On October 12, Canonical announced the release of Mir 0.28. This brings a lot of new features to Mir and there is no way that I could do them any justice writing about them here. Check out the linked blog post to learn more.
There is one important feature, though. Mir 0.28 brings preliminary support for Wayland clients, meaning that any Wayland application should run on the Mir server. This is demonstrated in this Youtube video, also linked in the release announcement post.
This should also mean that Mir or Wayland client applications that are packaged as snaps should work on a 16.04 base with Mir 0.28.