News and Update
Dalton, Florian and Alfred presented.
A big improvement has just been announced, with the release of Clickable 7.0. This is the software which UT uses to build apps and this version has new features across the board. There is a Clickable tool which will handle the format and library changes for those on the older version. It has way more options than any developer is likely to need and overall has more features than Xcode. It makes the whole process of creating apps for UT easy and seamless.
Work on Xenial has not stopped. Rotating lock screen now works. Thanks to Capsia for that change. Some ports now have background blur. A wallpaper behind the UI was a very simple arrangement and posed no risk to stability. Making things more fancy has the potential to introduce bugs and degrade performance but we have brought in a modification which works well. Porters have the final say but if they want, they can implement a translucency layer, so that when the app drawer is expanded from the left side, the wallpaper is visible in blurred form behind it. This feature will be in the next OTA release. Something similar was tried a long time ago and was tested on the Nexus 5. The performance was dreadful, so it is good to see that the problems have been solved.
Dialer has had an improvement. There is now a predictive feature which will offer a suggestion based on your first few numbers punched in.
WebGL in Morph is going to be in the next OTA release. Some of the very old devices such as BQ will unfortunately not be able to make use of it.
An improvement which you might not have noticed is that when swiping away tabs in Morph, the action is now more fluid.
Alfred has been continuing his refinement of the Pixel 3a. Landscape keyboard now works and there is an improvement in battery life. A really big one is the fact that you can at last fix in-call voice volume. That was a fundamental usability issue. “It will no longer blow Mark Shuttleworth’s ears – only his mind” :)
OTA 22 slightly delayed
OTA-22 was delayed by a bit. There was a synchronization fault which, unusually, affected the rc channel too. Apps started to break so we delayed and reverted. It will be sorted out soon and procedures changed around a bit to avoid future problems.
Dalton’s team has been continuing work on the 20.04 build. He showed a screencast of it working in qemu (it runs on the Pixel 3a but there is no convenient way to screencast it from there). A lot of parts are functional. Dialer is in there now. System Settings is in and ‘works’. Messaging has been installed but the are some size issues to sort out. Contacts has also been added and has functionality. Guido, Mike, Robert and Ratchanan have all done great work on it.
New apps in OpenStore
The first app in the roundup is Passes, which allows you to import PKASS files from emails or tickets and display them. That will work with things like flight and rail tickets, vouchers, sandwich passes etc. in a wallet and can be displayed.
SimplePrint is an experimental app which allows you to find open network printers and print to them in formats such as pdf and jpeg. It is possible to add drivers for particular printers.
Dalton will stay with us as chair of Membership Committee
As many of you will already be aware, Dalton has posted notice to everyone that he has resigned his employment with the Ubports Foundation. He didn’t wish to add anything to what he has already said, as his statement was carefully considered. There were really two strands. The first is the very obvious one that he has been doing four different jobs for the Foundation simultaneously and despite any appearances to the contrary he is not a superhero and it isn’t possible to sustain such a heavy burden over a long period of time without it taking a heavy personal toll. He has tried taking an extended break but that is not the answer. The way forward is not to find another superhero candidate and break them on the same wheel. We need four people to do four jobs. Dalton feels that only by leaving a vacuum can he bring that about. Time to stop pretending that a sticking plaster approach can work. There will be disruption in the short term but hopefully it will open the way to a more sustainable position in the medium and long term.
The second strand is related to the nature of where we are going. Dalton has led an enthusiastic, community-based project and felt very much at home with that. To advance, the project has to move to a more commercial phase, since business partners will be a necessary resource if we are to maintain a capable, attractive software/hardware combination. Dalton doesn’t oppose that path but he acknowledges that it is not so much fun for him. The community part has always been the most important aspect for him.
Right now, Dalton needs a proper rest and time to reflect. His personal views on how to go forward from here will only crystallize when he is thoroughly refreshed. Please respect that and give him space.
Dalton will retain his role as chair of the Membership Committee.
The News section of our Forum is the best place to post questions for the Q&A. YouTube live chat, Telegram and Matrix are other places to post a question.
If you didn't know, the Forum questions get priority.
Ubuntu Touch encryption
Tera asked what would be needed to enable encryption of user data on UT devices? Alfred commented that there are arguments for full disk encryption or just file level encryption, with pros and cons for each. It would be possible in theory to build either functionality into the Ubports installer.
In a conventional Linux system, the userdata and System are in the same space and there is no way to determine whether either has been tampered with. Encryption is necessary for safety in that kind of arrangement. UT uses a read-only file system, so the System is a specific known thing and checks can be run to ensure that it is uncorrupted and complete. Android and iOS take the same approach so it is not actually a necessity for them to have encrypted userdata. The encryption is tied to two elements. One is derived from something such as the serial number of the device, and an attacker will never be able to learn that. The second layer of encryption is derived from your pin. We could have a system which would result in those ‘system insecure’ messages on boot disappearing. This is a complicated subject and both Dalton and Alfred have written blog posts on it, so the best thing to do is head over and read them from the links.
Of course there is a potential collision between encryption and syncing/backup so there are some technicalities to solve and probably some user choice.
Clarification about Android references
Hugehead asked about our constant references to ‘Android 9 devices’ or ‘Android 7 devices’. If UT is an independent system, why does it have anything to do with these Android versions or any Android version? The reason is simply that we make use of Android kernels and Android drivers. The exception of course is the PinePhone, which makes the point that UT is something which can operate completely independent of Android.
With hardware built for Android, it was by definition not built for us so we have to find ways for UT to communicate with all the hardware elements. We use an abstraction layer to translate between our system and the Android drivers. In Android 9, Google introduced project Treble which brought manufacturers together in some common protocols, which we could then share across a whole range of device builds.
Florian put up a schematic showing the architecture of Ubuntu Touch and how things fit together. We are limited by the fact that the driver mechanisms are closed source. The only practical approach is to use them rather than somehow create our own substitutes. Some devices run on kernel 3.4. That is a bit crazy but we have to compromise with it and our ability to patch is very limited.
The Android architecture was specifically designed to create a barrier between the Linux kernel and the hardware of the device. It is a deliberate block, so it is not surprising that there is no easy way to sidestep it and have a build which simply implements the upstream Mainline kernel.
Florian emphasized again that we plan to move to a much smaller range of devices which are ‘supported’ in terms of having the backing of the core team. With the resources we have, choices must be made and our efforts have to focus on a manageable number of devices which are used by a significant number of people and perform really nicely. If you just bought a device which is only used for UT by ten people on the planet, we can’t justify a focus on it. Having phones with excited maintainers who have a good dialogue with their users is very good for the community. We want every device to have that.
KVM virtual machine
PinePhone Pro development
At present there are no decisions on what future if any there is for the Q&As in their current format. It has been a great run.
See you next time :-)