News and Update
Marius, AppLee and Alfred presented.
Marius explained that he has been working on a road map and that the top priority now is to work on the issues which impact most on users.
The lack of agps has been a major problem for a lot of people because it interferes with having route guidance ‘out of the box’.
Bluetooth audio is another weak point – especially in call handling. Even USB-C audio doesn’t work reliably with some devices and in the absence of a microphone jack that is a big problem.
Another annoyance is an occasional failure to switch between cell towers when travelling.
Some people don’t want emergency alerts by SMS but many users do and UT doesn’t deliver those reliably.
There have not been many regressions in the move to Focal but there have been some and we need to tidy those up.
In more of a ‘nice to have’ direction, 5G support and backup/restore are worth working on but the priority for those is not as high.
Some support for Snaps would be good. At the moment you can do a bit of that but only by making the phone RW, which is flaky and unsupported.
Someone asked about Snaps in Libertine. That is highly unlikely because it would amount to a nest within a nest. Way too slow and complicated. Reworking so that Snaps can run with RW is a big task but will probably come in the long run. Getting rid of Clicks is unlikely because they work so well.
Q&A every two weeks
Also on the roadmap are tasks which we need to complete before the 24.04 update. These are necessary but they are very small scale compared with the transition we just made. The next transition should take very little time. It will happen independently of the issues described above. It is a parallel process and one does not block the other.
QtWebEngine upgrade coming
Alfred on Ubuntu summit
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.
Alan G asked about the state of Lomiri desktop. Marius uses it every day but that is because he wants to drive out the bugs in it. That doesn’t mean it is daily driver ready. It will crash from time to time. All the important functions are there but it isn’t stable. Alfred tried it with 23.04 on his Macbook and found that it was extremely fast and fluid. Marius added that it doesn’t cache any values, so there are serious gaps still.
Caedem Asked whether it will really break out into desktop territory or is it just going to remain as a core UT architecture? Alfred explained that apps and system establish themselves at the same level. They intertwine in quite a rigid way. You can have some apps which sit on top of that ecosystem but they are kind of in boxes and self sufficient so that they don’t disrupt the main event.
If you give freedom to app developers to pick any version of Python they feel like, then you run into compatibility issues. UT can be used as a hacking toy but that is not what we are about. We want a robust system. You can exercise freedom by adding components to a Click package, rather than messing about with the system. You can add whatever you like, even duplicating what is in the system already. Not very efficient but it is very safe.
Morph on Xenial
Domubpkm asked if we would consider just keeping Morph browser up to date under Xenial? Sorry but no. Doing that would be more complicated than it sounds. To be clear – Xenial is abandoned. We just don’t have enough resources to spend any time at all on it. Focal is about going forward and we are not going to make any concessions to allow backward compatibility. The bottom line is get a compatible device and join us on Focal. That is where we are now.
Kuwailexs asked about progress with VoLTE. It is a horror. For a start, everything is written in Java. It is all proprietary and in Alfred’s words it is all about taking SIP and covering it with dirt, so it can be hidden in the woods. Although VoLTE is allegedly a standard, in fact everyone ignores the standard and does their own thing. It becomes a walled garden for countries and way for network providers to push their own phones onto users. In the USA, you will find that AT&T provided phones are the ones that work best with the AT&T network. Communicating with the modem and radio drivers in the phone needs adaptations specific to the phone. It really is a mess.
Alagurlalagurl asked whether UT runs on any device? No, not exactly. Every device has to be ported by someone and some are so challenging that it isn’t worth the effort. We have a lot of documentation about how to do that. A porting group on telegram provides a lot of support and encouragement. Patches will be needed on the Android side and the UT side, so it is not something for people without technical capabilities but it is not at a very high level.
Porting takes quite a long time so it must be a hobby that you enjoy.
Halium 12 RH asked about the stability of Halium 12 and whether there are any plans for Halium 13? Well Halium 12 is used on the Volla phones so that is definitely pretty solid. Halium 13 is used on the Volla 23x, so also pretty complete. There has been some testing with Halium 13 on Fairphone 4. That doesn’t mean it was a ‘project’. It was just a convenient test device and a lot worked well on it. There are lots of shiny, exciting devices which could use it and those will be our next crop of cutting edge UT devices. Binarisation made things easier in a way but now we have shifted from HAIDL to AIDL. So we have to rewrite everything. Halium development has become much faster though and much more rules based.
Fairose asked about the current state of FP4 with pre-installed UT. Is there an expected date for sales to start? Actually we may skip a model and go for launch with FP5. One of the things we are particularly keen on is to have 5G support. Marius has got that working in beta. It should be in Devel by the time of the next Q&A.
In the EU we can get by without VoLTE but in many countries it is essential and we would want FP5 to have that at launch.
Rocket2infinity asked about SIP support, especially with push notifications. These have been a lifesaver as a workaround for the lack of VoLTE in North America. It is of course not just a workaround but a great alternative approach. One thing little known is that Dialer has some unused coding which could be used in handling SIP. It would need someone to take on that work. Push notifications are server bound so that is another workstream. Is there anyone out there with the time and skills to work on those? Certainly it is doable.
Kristjan asked whether UT apps can be developed using Rust framework? For anything you use it must be able to interface with QML and it must be able to negotiate a path through ContentHub. It will need to play nice with other apps. Technically anything can be made to work but the question is whether the effort is worthwhile. In a simple app you would need QML/Rust bindings. Taking it to desktop via Lomiri opens up another box of problems.
Overall, we want UT to have a coherent look and a coherent behaviour. People who want to import ‘every app from everywhere’ will therefore be disappointed. Apple and Android have the same approach. Standard toolkits are there to prevent chaos.
Marius answered a niche question about Norwegian bank ID. The SIM toolkit approach has been junked and the new approach is app based. Now it is OS linked. Marius finds that WayDroid works for him. The apps are integrated with hardware and there are many different ones so this is an area where UT will probably not be able to follow. We are moving into an era where individuals are locked into the two main OS.
Alfred showed his development environment on tablet. His TIDE. It is a mechanism for programming and building in C++. The kit also includes a debugger. Hreadin, jamu and JASN are all implemented. It even works on a phone. It is convergent. Sponsorship to Alfred is welcome because he is not funded by UBports. Details for donation are in the show-notes.
See you next time :-)