KDE's Plasma Mobile Switches Off Directly Using The Ubuntu Touch Stack

[phoronix]

Phoronix: KDE's Plasma Mobile Switches Off Directly Using The Ubuntu Touch Stack

The Plasma Mobile stack being developed by KDE is almost one year old but they've now decided to shift their OS architecture a bit instead of relying directly upon Ubuntu Touch...

http://www.phoronix.com/scan.php?page=news_item&px=Plasma-Mobile-Stack-Shifts

[Julius]

Great news if this allows to run Plasma-mobile on more or less any Cyanogen phone/tablet in a container.

If they also implement a "Intel Big Screen Experience" like hybrid mode, it would bring the best of the two, i.e. a mostly standard stack of GNU/Linux and Android side by side, including convergence like functionality.

After following Ubuntu Touch for a while now (and using it on my Nexus4), I have become extremely dissatisfied with the direction they are taking. All that Snappy packaging (Sandboxing for commerical apps), the locking down of the main system and more and more diverting from base Linux just results is a bad copy of Android with few apps. The main point of convergence is lost that way, and then I rather stay with the original and run a proper Linux in a chroot

[bkor]

Great news! Kind of funny that they take something which runs Mir and put Wayland on it :-P

The test picture looks a bit ugly/distorted on the left. I wonder why that is. I was unable to leave a comment on the blog :-(

Related: Bionic ('glibc' alternative) needed some patches to make it work with Plasma. Some Devuan developer complained that systemd requiring things from a libc. Stuff should be fixed at the right place (the libc), not worked around in lots of places.

[Cerberus]

Great news if this allows to run Plasma-mobile on more or less any Cyanogen phone/tablet in chroot.

If they also implement a "Intel Big Screen Experience" like hybrid mode, it would bring the best of the two, i.e. a mostly standard stack of GNU/Linux and Android side by side, including convergence like functionality.

After following Ubuntu Touch for a while now (and using it on my Nexus4), I have become extremely dissatisfied with the direction they are taking. All that Snappy packaging (Sandboxing for commerical apps), the locking down of the main system and more and more diverting from base Linux just results is a bad copy of Android with few apps. The main point of convergence is lost that way, and then I rather stay with the original and run a proper Linux in a chroot

Care to elaborate how Ubuntu Touch is diverting more and more from base Linux? And I dont mean Unity 8 and Mir. You can do whatever you want on Ubuntu Touch, user has full control, and soon enough you will be able to run X applications on the phones, you can already do that but with some hacking. As for Plasma Mobile good luck to them, they will need it and lots of it as Plasma Mobile team does not have the same manpower that Canonical does, and Canonical is still polishing their OS and fixing bugs more than one year after it reached enthusiastic buyers, 2 years in development prior to that, mobile operating system is a massive undertaking, sure they can build upon Canonical's work and get some things done faster, but it will take a long time before Plasma Mobile is anything more than a toy for geeks. But I welcome any attempt an open source mobile OS as Sailfish OS and Firefox OS have already pretty much failed, Ubuntu is still going, more projects are good to increase the chances of a open source mobile OS ever reaching the mass market.

[shmerl]

Why aren't they using Mer like they used to (don't confuse it with Mir)? Also, what about native (not libhybis based) scenarios?

[Elv13]

Why aren't they using Mer like they used to (don't confuse it with Mir)? Also, what about native (not libhybis based) scenarios?

If I recall the talks at Akademy last year, this was due to some factors. First, the older Plasma Active / Plasma Mobile was Mer based. It unfortunately never really took off (as Mer itself, also unfortunately). Second, this is mostly developed by the Blue System (the "trust" backing Kubuntu) / Kubuntu / Neon team, so it made little sense to use anything else than their own base OS across form factors. This is mostly because they can share a continuous integration system. CIs are very hard problem to solve, especially for OSes. Having to duplicate effort across 2 base OS made very little financial/time/resources sense. It also made packaging workload and knowledge pool smaller and more maintainable. Finally, as Ubuntu Touch is also QML based, it can (or at least could) run their sandboxed apps.

I got a phone running Plasma Mobile. It is very early, but it does work. Of course, as with all other FLOSS "DE"/ large communities, this wont happen overnight and require tons of effort (including many full apps rewrite) to make the transition to (good) mobile apps. I got a KDE Modile port of mine (Ring-KDE) in alpha stage (not public yet). I did all the "heavy lifting" to be able to create QML versions. It took me 4 months, full time (sponsored, in 2012) to rewrite all the necessary components. It was one hell of a job. You have to really invest in separation of concerns between the logic layer and the GUI/UX and port everything to Qt models or QtQuick friendly APIs. Once this is done, creating other UI become trivial. The Qt based logic layer now back not only Qt based UIs, but also GTK and Cocoa (OS X) native ports (see www.ring.cx). It took this much abstraction to be able to have both QtWidgets and QML based "clients".

Edit: As for libhybis, it should not be an hard requirement. It just take less efforts to get ports running on Android devices. They can be converted to free drivers when the port mature.

[GreatEmerald]

If I recall the talks at Akademy last year, this was due to some factors. First, the older Plasma Active / Plasma Mobile was Mer based. It unfortunately never really took off (as Mer itself, also unfortunately).

That does not sound right. I used Plasma Active, at least, and also Sailfish OS is based on Mer.

[Julius]

Care to elaborate how Ubuntu Touch is diverting more and more from base Linux? And I dont mean Unity 8 and Mir. You can do whatever you want on Ubuntu Touch, user has full control, and soon enough you will be able to run X applications on the phones, you can already do that but with some hacking.

Did you actually use UT? Right now you can not do anything on UT except installing mostly bad apps from the limited selection in their app-store, which also requires you to register an Ubuntu One account before using it. Installing .debs is completely disabled, and when you enable it, it brakes the system, disables all OS updates and basically leaves you with a crippled non-functional system. Convergence is also still not really functional (any disabled on most devices by default), and all desktop applications need special packaging to run, so you get 3-4 apps right now (firefox, libreoffice etc.) and that's it. No "full desktop experience" at all... in fact worse than MS continuum right now.
And then they push their snappy sandboxing system, which is basically a stripped down version of Docker to run never up dated, closed-source apps conveniently without braking the system. And there are many other small issues that basically show that they are just trying to become a "better" Android, which will surely fail as they can't compete with Google by copying them.
All the good things about it, e.g. real convergence, full GNU/Linux/Debian compatibility, OS updates that do not depend on the device producer, heck even Scopes have been either abandoned by now, or are shadows of what they could have been and seemingly not the developers priority any more

Edit: yes early days, still under development... yadda yadda yadda. I am also really trying to give them the benefit of doubt and have used it on my Phone for a while, but it it not so much what is missing, but rather what they have already started to implement or have planned which has left me very disillusioned about UT.