Systemd For The Ubuntu Server Gets Discussed

[DrYak]

The need Mir now for Ubuntu phone primarily and if it starts to get traction, they'll keep working on Mir.
Jolla won't release their full source either.

The parts that Jolla isn't releasing aren't core functions, only some of their shiny polish (just like HTC aren't releasing their own specfic UI skin that they put above the base Android, but the core service, Flinger is open source)
and in fact some of that Jolla code (that they use to enable Wayland to function with Flinger drivers) ended up being used by Mir. Similarily XMir leverages some work done on XWayland.

As for platform needed for phone... well... let's be realist:
Wayland is actually shipping on device. Device that already exist, that you can go and buy today (Jolla, a few other embed platforms runing Tizen, Intel's IVI initiative).

Whereas there are no actual Ubuntu Phone device, beyond a few experimental installs on Google Nexus platforms. (Even WebOS has bigger shipped/install base than Ubuntu Phone), so Mir hasn't even shown proof in real world situation.

[chrisb]

The parts that Jolla isn't releasing aren't core functions, only some of their shiny polish (just like HTC aren't releasing their own specfic UI skin that they put above the base Android, but the core service, Flinger is open source)
and in fact some of that Jolla code (that they use to enable Wayland to function with Flinger drivers) ended up being used by Mir. Similarily XMir leverages some work done on XWayland.

As for platform needed for phone... well... let's be realist:
Wayland is actually shipping on device. Device that already exist, that you can go and buy today (Jolla, a few other embed platforms runing Tizen, Intel's IVI initiative).

Whereas there are no actual Ubuntu Phone device, beyond a few experimental installs on Google Nexus platforms. (Even WebOS has bigger shipped/install base than Ubuntu Phone), so Mir hasn't even shown proof in real world situation.

Let's be realist: All of these new phone platforms (Ubuntu, Firefox, Tizen, Sailfish) are going to have a massive fight on their hands to carve out a niche in the market. You know how many Android devices were sold today? Over a million. And yesterday? Over a million. And two days ago? Over a million... Consumers buy over a million Android devices every day. To be a success these new competitors will have to offer something that Android does not. Ubuntu has its "convergence" aim, Firefox has its $25 phone with 100%-web-apps aim, to be honest I'm not sure what the unique selling point of Tizen or Sailfish is.

Samsung's motivation appears to be that Tizen isn't controlled by Google. If they wanted to they could just do like Amazon and ship AOSP+Samsung changes, but contractually, this probably means abandoning "Google" Android altogether, and they aren't ready for that. But their main problem is that their customers now have an expectation that a Samsung Galaxy phone is an Android phone - customers expect that it will have Play apps and Gmail and Youtube. Which retailers are going to put their staff and customers through the confusion of selling the "Galaxy Z", and getting returns from customers when they find it is a completely different platform to the one they thought they were buying? A platform that doesn't run any of their favourite apps? Hmm.

Sailfish's unique selling point seems to be that they have a nice UI. Their problem is that at this point, consumers are unlikely to notice - people are generally happy with their Android phones. Their other problem is that, like Ubuntu, they want to "embrace and extend" Android by running Android apps, but they are doing this by running a proprietary compatibility runtime rather than using AOSP. I have no idea how compatible "Alien Dalvik" is with real Android apps (?) but I would guess that users won't just be able to sideload Android apps and have them work 100%. And unless it works really well, everyday users will get a bad impression, and just won't bother. The words "proprietary Android runtime" do not inspire confidence. It's also a bit suspicious that Jolla list less than 50 compatible Android apps. But this speculation could all be completely wrong - perhaps their Android compatibility is actually very good (?), they are certified after all.

Android is a good deal for manufacturers. They get a whole platform for free (including server-side stuff like Gmail, Youtube) that would previously have cost quite a bit of money to develop inhouse or license from a third party (eg. Apple spend tens of millions $ on platform development, licensing Windows Phone OS is $20-$30 per unit). There is no license fee for Android+Google apps. The only cost is mandatory third party compliance certification which works out at less than $1/device. It will be hard for the new platforms to compete, they fundamentally have to answer the questions, "Why should I buy this, instead of buying a Galaxy S5 or Moto G?" and "Will it run all the apps I want to run?" The astute might notice similarities between those questions and the questions asked of desktop Linux over the last 15 years.