Ubuntu Touch Q&A 80
Lomiri on Manjaro

 
 

Looking for the Audio-only version

of the Q&A?  You're just one big orange

button away -->

News and Update

This show was hosted by Marius, Florian, Dalton and Alfred but with them were invited guests from the Manjaro project, Philip and Furkan.
Furkan was attending using his PineBook Pro, on which he was running Manjaro ARM.

OTA-13

There has been more progress on OTA-13, with 42 PRs merged. Also Qtwebengine 5.14 has been merged. It offers about 25% more speed in loading and the Morph browser gets kicked from far fewer websites on the grounds of being old and outdated! There are now newer libraries for Qt glib and others. There are fixes to the address book app, including being able to record messages and the ability to manage the import of contacts from a Vcard file. Florian has added IPv6 support. It took him three years to fix it but here it is. PinePhone and Xperia X update settings now only show updates which will actually run on your device!

Lomiri

Lomiri packages are now arriving in Debian. Sunweaver has confirmed this week that Lomiri download manager works. URL dispatcher also works. Applaunch also functions now and finally Mir 1.8 has landed in Debian. These changes only apply to unstable, they are not yet in the testing channel. In the very long term, this will lead to Lomiri being in Universe, when Ubuntu rebases to Debian.

Lomiri API is in a UT app and of course we have already borrowed the Suru theme.

Teleports update for PinePhone

Teleports has a nasty bug with the keyboard on PinePhone which made it impossible to type. There is now a fix for that. If you switch to rc channel you can use the update.

Manjaro

The big thing for this Q&A was introduced by Marius, who showed Lomiri running Manjaro on the PinePhone.Philip explained some of the objectives for Manjaro and how they differ from something like Ubuntu. It has been called ‘Ubuntu for Arch’. Lomiri was able to run much faster on Manjaro than on other systems, which made it interesting. The ARM project started in 2016, initiated by Joshua. Marius confirmed that the package manager in Manjaro is a gem and very easy for developers to get into. Having up to date packages means that many bugs have been fixed, as well as providing a good increase in speed.

Manjaro uses a snapshot system rather than libraries, which speeds everything up. In effect, it builds from the gathered elements rather than from large assembled packages. Manjaro have their own system for handling Appimages and it might be possible at some point to move that into UT.

The PineBook Pro is an ARM device, just like the Raspberry Pi. The Manjaro project started on ARM with KDE, which is convenient because it is a rolling release, as Manjaro is. There will be three different flavour UIs for the PinePhone running Manjaro. Users will be asked to feed back what works best.

Marius approached the Manjaro task as a private project ‘for fun’ . Of course having Manjaro Lomiri will generate issues that will have to be resolved both on the Manjaro side and in UT. Having something practical to work with will take it up a gear, through practical feedback.

Marius reported back that Lomiri Manjaro went a lot smoother than expected. More than 80 packages are in his test build. His view is that the hard part is done especially getting to a systemd user session. It is actually already running on a systemd compositor.

Ubuntu Touch project and Ubuntu 20.04

Dalton explained that the UT project is very much about connecting with new upstream software and ultimately to 20.04, which is where all this fits in. This will mean more common ground between mobile and desktop platforms, enabling users across all of those to derive benefit. It means among other things that when we build software for UT, some of that will be available for use by people not on UT. It will be our way of giving back. It should also expand the developer base significantly, since their work will have wider use.

UT is mainly designed for the specific purpose of running on phones. As a result Lomiri has the reputation of being a ‘phone thing’ and that has limited interest in it. Using it within other systems will hopefully break down that (mostly wrong) perception.

In the test build created by Marius, he has already found that some UT apps work surprisingly well in that environment and of course this is just the early experimental phase.

Sponsors were thanked. 

Questions

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.

Flutter protocols

Povoq asked what thoughts we have about Canonical’s decision to port Google’s Flutter protocols? If it runs on Wayland this could provide a great facility for UT ports? Dalton commented that the native Matrix client for UT known as Fluffychat has now spawned a Flutter version, making it cross-platform. At the moment the focus is on using snap to package them but that is not compulsory. We could look at other ways. Just to clear up something – the developer used Mir to run it during the development phase but that does not mean that Mir will be used for the publicly available version.

Click packages architecture

Povoq also asked whether it would be theoretically possible to divorce the whole click packages architecture from core Ubuntu eventually? Basically, yes. The new build means that someone would be able to build an app for Manjaro but have that packaged as a click, to run in UT. That crossover has already been achieved, to some extent. Another consequence is that some day – not yet – the door will be opened to packages such as Flatpak and snaps.

The point of click packages is that they will run on read-only file systems. That is what they were designed for. It went along with confinement and made safety independent of trust in developers. It is therefore not possible to ‘just’ run clicks in Manjaro. It would be necessary to build applications for UT which are constructed with the intention that they should also be able to run on Manjaro. In practical terms that would mean building for a slimmed down version of UT, using a snap or Flatpak package.

Lomiri desktop

Alan has heard some chat about Lomiri desktop recently. Does that mean there has been a corresponding increase in interest from developers? Well of course having it spread across more platforms will increase the number of eyes on it and therefore speed refinement.

Aury88 asked what parts of Lomiri are most in need of love, to provide a passable experience on desktop? Marius was surprised how much did work, especially the compatibility with X11, once patches were put in place. Workspace and multi-display are other major issues, obviously. There are also lots of gaps with mouse handling. Furkan said that a priority was getting apps working. The lack of graphics drivers also needs to be resolved. That is a longer term issue because it has to go hand in hand with kernel development and that is a slow process.

Furkan acknowledged that there will be a need for equivalents for apps in daily use on Android (including Whatsapp) across linux platforms, once the PinePhone becomes established as a general daily driver option.

PinePhone feedbacks

Furkan asked about feedback from PinePhone users on the UT side. All agreed that although some have taken the plunge to use it as a daily driver, it is not yet ready for that. The special thing about it is that everything in it is open source software, so even in its current form it is an amazing achievement. The modem and wifi are a ‘black box’ but at least they are isolated.  The same is true of the USB-C controller, although at the moment that is not running with UT.

In principle, UT is the most difficult to adapt for the PinePhone because it is so different. If we can do it, more mainstream operating systems should have it easier. Marius made a lot of hacks for his Lomiri build so the first task is to go back and do some tidying. After that, keyboard support is a priority, as well as adding apps. The modem then needs to be hooked up for wireless functions. Telephony handler is running and already detected a SIM card. Ofono is already running, so lots of the necessary elements are there.

There is no timetable because as usual this is something that has never been done before, so there is no way to measure.

Furkan suggested that Manjaro community members can pull in all the packages from Manjaro unstable and feed their experiences back. Everything needed is there. Lead developers randomly test a few things when building but there are huge amounts they don’t see and a large community of testers is vital to taking a project forward.

Philip recommended the Manjaro forum, which is very friendly. Even if a question has been asked there many times before, the community are very positive and help out.  Of course it is also a great place to tell what is working and what is not, in projects like the one discussed today. Support between communities is also vital. If a problem has been fixed by one community, tell the others so that they can adopt the solution, if appropriate. We are in some sense ‘competitors’ but in reality we all work from the same Ubuntu base and we advance by mutual help. It is great that this is working particularly well at the moment.

Furkan made the point that it is only by that close working and solidarity that the linux based mobile systems as a whole can present a challenge to Android and iOS. We are all of us about providing to a mass of ordinary users. Our ambitions are not just about pleasing a band of developer ‘tinkerers’. Communities of users can develop some tribal instincts but that is not the way developers feel. They know the debt they owe to each other.

See you next time :-)

Ubuntu Touch Q&A 79
Volla phone supported by UBports installer