Ubuntu Touch Q&A 45

OTA-8 and Beyond
16 marzo 2019

 
 

Looking for the Audio-only version of the Q&A?

You're just one big orange button away -->

Show Notes

News and Updates

Dalton appears alone!

Release of OTA-8 on 6 March.

It was a stability improvement release. He will talk a bit about Edge channel later.

Dark theme in tab headers of Morph browser can be added using UT tweaktool. Some apps have an over-ride, which allows you to use dark theme without needing any special tweaks to the system.

Some javascript changes and proper error pages for the browser.

UT when given to us by Canonical had built in tests. We turned those off in order to move faster but of course that created an accumulation of problems. Finally we have re-enabled tests. It has meant a huge change, especially as we now use a vastly different version of Qt.

OTA-8 has helped to sort out some bugs with address book.

There is some quite good porting documentation now available and even a video on porting to Ubuntu Touch. It is a slow and difficult process but you can do it with new devices if you set your mind to it.

What infrastructure should an open source project provide itself and what should it farm out to others?

This issue was highlighted recently by problems with our team notepad. This week it went down for a few hours and when it came back up, it didn’t retrieve the archive as it should. Of course we all want to do things ourselves but if it fails, you have to ask whether that is wise. This is a subject we would like feedback from you about.

We need to maintain a web site and a debian repository; maybe eventually services such as Nextcloud under a UBports umbrella? But we then have to manage those things, so there are resource issues.

Marius joined the conversation.

How big we are as an organization dictates the extent of what we can achieve ourselves.

Sponsors thanked.

Questions

Porting seems a hectic process and I am a noob. Is there an easier way to get Ubuntu Touch?

Yes - the answer is to get a supported phone! Joining an existing porting effort is also a good idea.

Are there devices running UT on ARM 64?

On a very few test devices there is a very experimental version running but it only works in outline and certainly is not in a usable state.

ARM 64 of course needs more space for the extra libraries. The obvious solution to that is increasing image size? That creates further issues …

Edge channel has to deal with that problem.

Is there an app for ‘porting’ Linux apps?

Well, you can use Libertine to run apps. For Android apps there is an experimental version of Anbox, which you can try on some devices. (The question may have been about taking a Linux desktop app and turning it into a native UT app? We will try to comment on that, on the forum)

When there is Wayland support in Mir, we won’t technically need Libertine to do that anymore. We may still want to use a modified Libertine though, because of the app confinement that it brings.

More information about Dekko2 development.

Dan is extremely busy with it. He is moving Oxide dependent elements towards support by qtwebengine. Not an easy task. If you want to help out, contact him at dekkoproject.org.

With complete removal of Oxide, will some webapps which depend on it  just stop running?

Oxide wasn't supposed to be removed in Edge -- we will bring it back soon. Before removing Oxide from mainstream UT we will warn users. That will not be an issue for a while yet though. Oxide doesn’t work with ARM64 incidentally, so that is an issue.

Will WebApp Creator development take account of this?

Actually, it already supports Morph so if there is a problem, you should see that as a bug, not an Oxide problem. Of course if someone can update that app it would be great

Mark says can you share a bit more about Mir and Unity 8 development in Edge channel?

New mir brings in Wayland support possibilities, so we can start using Wayland protocols. Pine has technical issues. QtWayland upstream will be a big improvement on using UT Wayland, where the burden of maintenance falls on us. That will save lots of effort.

There is a lot of information in the forum about Edge. You can also test it and report bugs, which will accelerate development.

Since Mir is almost done, what is the top priority after?

Well, Treble support; Pinephone support; Librem support; Halium improvements. All are contenders.

Our first priority after OTA-9 is stability, says Marius. But Dalton points to the often overlooked issue of maintaining our Debian packaging facility. This is really feeling the strain because we have grown. It is going to be completely redesigned.

At the moment we use Freight for Debian packaging. Scale is a problem with that. It recalculates everything for even a small change and takes c. 20 mins to rebuild a repository. We want dependency tree management etc. So a completely new solution, combined with further moves to Gitlab. Aptly is a possible tool. Debian’s options are too bloated for us.

Simon has been working on tooling for more complete Debian package management with Jenkins for Lubuntu. We may be able to take advantage of this.

We need to be able to load Unity 8 on desktops and compatibility keeps slipping between us and Ubuntu so we have to rebuild manually all the time. That would be corrected by adopting new arrangements.

Hopefully we can ultimately integrate Unity 8 fully into Debian.

In answer to a question, Intel processor phones are not supported by UT (eg. Asus Zenphone). There have been no successful ports of them and they are quite strange.