News & Updates
We now have a presence on Gitlab. The calculator app is the first to be going over. One of the benefits of the platform is that it will assist with translations. We are of course not going to move everything tomorrow. We are starting with some simple apps, to try it out. Thanks to Brian Douglass for the calculator build.
Morph and Qtwebengine improvements have been put in place. Chris Clime and others have done a lot of work. It is a big jump from Oxide. For now we have disabled GPU Acceleration in QtWebengine, since it does not work yet. We need later integration with Mir before re-implementing it. That will take a while longer.
On system software development there is a big update to docs. It will now be a dedicated header "System Development" on the Docs website. A lot of the new knowledge set out in the documentation came out of the development of the OS over the past year.
By the middle of next week, we should have most of the new documentation in place on those web pages.
Marius wants to talk about Nexus 5 and camera. Pretty deep problems there. Wait a bit or concentrate some energy on a fix?
The possibility of the project Treble as a solution.
That could use a uniform API, via a binder, instead of going through bionic and libhybris as we do now. A Linux hardware interface definition layer uses C++ to create an interface. A service starts a device side server and connects to a vendor server. This is in a way similar to the way desktop systems operate.
So it is similar to a DBus call. It all happens in kernel.
We can continue with these experiments. Will it be used in UT? That is not clear yet. What we can see is that it might provide an entirely different way to access the camera.
Jonatan joins the Q&A.
He does articles, blogposts, documentation, system bug fixing etc. A true generalist!
A German group started up in March. It gathered enough pace that it was soon possible to get a social and networking meeting organized.
There was some discussion of Raspberry Pi and some work done around Mycroft. That was the first working meeting. The Raspberry pi group which was set up was to scope what might be possible but with so much else going on, not much has happened yet. There are lots of ways in which a build on the Raspberry Pi could be interesting and productive.
Florian has been busy with lots of different work, so the appearance of a group coming to help him was great. Telegram has produced a new working lib which has made it possible for a lot of the latest functionality to be implemented much more easily. Hence the Vienna meeting. Mycroft was not urgent but Telegram really needed attention.
TelegramPlus lacks management actions such as adding members and groups but the messaging functions work well already.
Jonatan enjoyed the event very much but felt a bit left behind because of his lack of Qt knowledge.
This and That
There will be another hackathon in the new year. Details will be made known later.
Jonatan came to UBports because he does not like the way that Google etc extract data. He had used Ubuntu desktop for a long time previously, so he wanted to try it with a FP2 that he repaired. It was the obvious alternative to Android and Apple.
60 live viewers watching the Q&A this time. That is a record! Those watching live are usually around 20.
Will TelegramPlus be published soon?
First, we need a proper, permanent name that refers but is not the exact Telegram name. The logo can look similar but must not be identical. Why not send us some ideas? Like... Florigram! By the beginning of December there should be a beta in the Open Store. The clicks are available as nightly builds until then.
Sponsors acknowledged. On Patreon, especially Darko Balke, Morgan McMillian, and Thoralf Sann.
Thanks to our corporate sponsors Scaleway, Packet.net, Digital Ocean, Smoose and Private Internet Access.
What are the midterm and long-term goals of UBports. How about a year from now?
We will have the Foundation, for sure. We will then be a legal body, able to act officially. Better marketing, bounties to developers. We will have a "Head" of each function.
We will by then be looking towards a new LTS, probably 18.04, with experiments around the end of 2019. We can experiment with 18.04 without affecting 16.04. There is no timeline for those experiments and will not be for immediate practical use. We don't know how many developers we will have as we go through that so time frame is difficult to predict.
Next year there will be regular South German meetups doing practical stuff.
More devices will be able to run Ubuntu Touch, either natively or as Halium phone ports.
Unity8 will be done for desktop.
We have 1.5 developers now. How many would we need for sustainable maintenance of UT?
Android and iOS use thousands. We can take a completely different approach. We would cooperate with developer teams from businesses with which we have a close working relationship. That would be in addition to our own volunteers. The organizing principle would remain the community, rather than a monolithic company. There would be lots of strands to the community, commercial as well as enthusiast.
10 full time paid developers working directly for the Foundation would be a great level to have. Around that could be a community of commercial partners, for example phone manufacturers.
So much is changing in the tech world that it is impossible to plan really long term. So much new and unexpected could happen.
Can we translate blog posts on Weblate?
Florian said better to do it as a whole texts otherwise it can fall out of sync. Merge it that way but not via translation server
Is there a recent bug affecting webapps that needs to be fixed? What are good steps to handle Oxide deprecation?
If standard containers are used with webapps there should be no problem at all but if things are more intricate and Oxide is used directly that will need tweaking. Look at the QtWebEngine documentation for now but a lot more documentation on all of that stuff will arrive soon, before Oxide is retired. The structure of UT itself still has some Oxide dependencies so we will need to patch those before it can go.
Any news on Unity 8?
We need to get the latest Unity8 onto the phones. At the moment there are different versions on desktop and mobile. The older version on mobile has to be replaced, so that we can work with two platforms yet one version. Then we can move ahead with both. The necessary upgrade also involves Debian, Gitlab, versioning etc
Rodney is doing a lot of work on Unity8 currently.
Upstream Mir is needed for X11 etc.
At any rate, Unity8 is usable for desktop at the moment.
Github is getting a cleanup. There should be a shout-out re old Github requests. Please close them yourself if they are no longer relevant. That will help maintainers a lot.
The more we move stuff to GitLab, the more we should tidy Github first.
Widevine is an Android DRM. Will it come to Morph browser?
Netflix needs Widevine to function. It is a Google technology and they have made it very difficult to use on ARM. On mobiles, it is embedded deep in its closed architecture. Finding some way to get it to work on ARM would be very challenging and certainly a huge amount of work. A nice to have perhaps, but really extremely difficult to achieve.
There are other DRM systems out there, so perhaps the Google system is not the only one that could be considered.
In OTA6 will the Morph browser become the equal of Oxide, since that is a major obstacle to some users upgrading?
The answer is that in devel channel, Morph is in many ways either close or better than Oxide already. So basically, yes.
Are there any new ports?
Sony Xperia X is moving quickly. Halium has a lot more coming through though. If Treble experiments are successful, very likely a surge of new devices.
Any news about the Foundation?
Preparations are continuing. No promise of a time when it will be agreed.
Have we received the Librem5 test boards?
No, we don't have them and we don't know if we will get one or more than one.
What about Samsung's DEX?
That probably will not affect us much. It is an Android phone which has a mini Linux installed for when it is plugged into a desktop.
Maybe Unity8 will work on it? In any case there are all the usual privacy issues associated with a Samsung Android device. By raising interest in the subject though it will do us a lot of good in marketing Ubuntu Touch.
Bug fixing is not discussed in Q&As so please do not submit bug questions.
A reminder that www.ubuntuusers.de is a good link for those who speak German. The site has a good forum but the articles there about Ubuntu Touch are very out of date. Volunteers to refresh them are very welcome!