I have an impression that the goal of all these new iOS/Android challengers is to get Linux on the mobile phone at all costs. Whereas the goal should be to create user experience that would be at least as pleasant and smooth as that of iOS, plus openness and transparency of the platform. At least.
That may imply Linux due to the lack of viable alternatives, but really, it should be built with users in mind, rather than perceived as a technical challenge of putting Touch on top of Linux.
(And please, stay away from Java in your next mobile OS. Among other things, you will ward off those of us, developers, who let's say are not neutral wrt. to certain programming languages. Edit: make Java optional if you wish.)
This has little to do with getting Linux on mobile. Maybe it's the motivation for some. But not the driving aspect.
Samsung wants Tizen to gain back full control of their platform. Canonical wants Ubuntu Touch (and Mozilla Firefox OS) because the PC is getting less important for consumers.
> (And please, stay away from Java in your next mobile OS. Among other things, you will ward off those of us, developers, who let's say are not neutral wrt. to certain programming languages. Edit: make Java optional if you wish.)
Neither of them is based on Java. Tizen and FirefoxOS are (sadly) JavaScript only. Ubuntu Touch supports JavaScript/HTML5 and Qt (JavaScript + C++).
That Android challengers are Linux-based doesn't imply Android is not. So logically it's still correct.
What Android competitors could do though would be to (1) forget Java, (2) make the platform completely OSS-license compatible, (3) make the UX understandable and loveable by children and kittens.
It's a shame the way that Canonical have isolated themselves from the rest of the Linux ecosystem. I wonder if the future for Ubuntu Touch would be brighter if they had more community appeal.
How so? Android and Firefox OS can run few if any GNU/Linux applications without major modifications. Ubuntu run every Linux app I can think of out of the box.
I was referring more to isolation in two related senses:
1. A technical sense: Canonical has a significant amount of distribution-specific software that has very little adoption outside of the Ubuntu ecosystem. Examples include:
* their own version control system, bzr (which I think has now been essentially abandoned in terms of development, and other than a few GNU projects never saw much uptake at all);
* their own desktop shell, Unity, which to the best of my knowledge has never been properly packaged for any non-Ubuntu distribution; their own display server, Mir, whose development is the subject of much controversy, and which has been rejected for well-founded technical reasons that have probably already been discussed at length on HN;
* their own init system, Upstart, which was initially adopted by other distributions, who then abandoned it in favour of systemd due to its fundamentally flawed design (for instance, there is no way to 100% reliably keep track of a process's children and the dependency resolution runs backwards, meaning that some configurations can cause hangs).
2. A social sense: Canonical's relationship with the rest of the Linux ecosystem has soured (beginning, in my opinion, around 2009-2010, but as a gradual thing, it's hard to pin down) to the point of outright hostility. In my opinion, most of the responsibility for this lies with Canonical's poor communication with others. There's a very detailed history of some events from 2009-2011 that you can read here [1] detailing their relationship with GNOME.
The culmination this year has been disgraceful personal attacks by Mark Shuttleworth on people who raise technical criticism of Canonical's software. I expect that more will follow whatever the fallout of Debian's systemd decision is.
The two are obviously related. The big problem is that there's (again, in my opinion) very little community appetite for contributing to Canonical's software. And why would there be, given that in order to submit patches for Ubuntu Touch, I need to give Canonical permission to relicence my code under any proprietary licence they choose? At that point, it's not a community contribution, it's just working for Mark Shuttleworth for free.
Does Canonical have the software engineering capability to maintain a bunch of stuff (an init system, a display server) that's really unrelated to their problem (making awesome products for users) and still be able to take on Android at the same time? I don't think so.
> Does Canonical have the software engineering capability to maintain a bunch of stuff (an init system, a display server) that's really unrelated to their problem (making awesome products for users)...
Thats an odd statement. It implies that an init system and a display server are not building blocks of an awesome user experience.
For upstart, you have to note that it was Ubuntus contender to write a modern and fast init system when there were far less mature init systems. Its main competitor (systemd) is about 4 years younger that upstart.
Ignoring the whole problem about wayland, the display server is a central part for providing the user interface. So how is writing a new display server "unrelated" to providing an awesome product to the user?
I think Canonical are determined to become financially viable. That means trying to generate a fairly high income stream somehow. I'm dubious about these deals with carriers/manufacturers but that is the direction they have decided to go in, rather than trying to do a Red Hat on server/virtualised support.
I would quite like a phone that becomes a desktop OS when I dock it. If they can actually pull that one off I think a lot of people might get interested.
I don't like various of their decisions, but at the same time wish them all the best. For a phone they're focused on doing their own thing. Really unfortunate for cross distro cooperation, but when I use a smart phone is important how great the experience is. If the phone is focused on the init system, then that's a phone I don't want to buy.
I don't think it can be said that they're isolated when the majority of Linux users and (non-system) developers run Ubuntu. And having run various Linux and BSD systems for close to two decades I can say that there is a good reason for that - it's by the far the best distribution out there.
Yes, but it's only in the AUR, so I didn't count it as "properly packaged", in the same way that I wouldn't count something as "packaged by Ubuntu" if it were available in someone's PPA.
It's a fair point, but I think that my assessment still holds.
I imagine the packager liked Unity and prefers Arch!
Unity as an interface is OK ish, apart from the bugs about accelerator keys in LibreOffice and other applications (e.g. Alt-F no longer brings up the file menu).
Sometimes an extra option is desired by companies like carriers just so they can pit all the options against each other and get better deals. Similar happens in the OEM space. An OEM may be happy with Qualcomm, but they are going to flirt with NVIDIA and use them occasionally just to get better deals from Qualcomm.
Regarding the first-edition Nexus 7 - Canonical will no longer produce Touch images for it, so the author of this article is going to be even more disappointed during 2014.
Samsung should release high-end Tizen (Bada3) smartphones in Europe. Most use Samung devices here with Android. If Samsung keeps its Touchwiz UI and ship an Android compatibility layer most people would buy it anyway. They buy it because of the Samsung brand name!
This was not unexpected; despite all the technical and political troubles iOS and Android have, mobiles OSes can be considered a commodity; there is little competitive advantage that any new contender could bring about at this point unless something truly radical emerges.
Ubuntu's unified OS on all platforms seems like an interesting idea at first glance, but Canonical has so far failed at providing the necessary vision for that to come through.
It's not so much that they're a commodity, but rather it's the fact that Android and iOS are so entrenched. Both ecosystems are networks. In fact iOS is a major competitive advantage for Apple, with all the apps that are iOS first.
Cyanogenmod is quite successful(10 million installs). It seems like a good model for creating alternatives in the os market(as long as google is happy and sharing it's apps and store). And they've inserted some interesting features to the os , for example strong encrypted text(TextSecure) as default.
Mobile OS seen is amazingly diverse; Symbian, WebOs, Blackberry OSes, Bada, ->Tizen, Ubuntu Touch, FirefoxOS, SailfishOS and then Android/s, Windows* and iOS. Anything else I omitted?
Various pre-smartphone Japanese phone OSes, e.g. I think au had one called "KCP", which was based on "BREW" from Qualcomm... BREW was apparently used in various forms by others too.
I think at this point the smartphone OS's that are new Firefox OS, Ubuntu, and Tizen are bascially an attempt to stay relevant in a world where the primary computer is a phone or tablet, not a laptop or desktop.
All three of them exist more because of the companies that built them, than because there is a big gaping hole in the market that they fill. Without a big problem they are solving I think the market as a whole will shrug at their arrival sort of like the Palm Pre.
Sailfish OS is built on all the same technologies that are upcoming in the desktop/server linux world, namely systemd, pulseaudio, wayland. They even use BTRFS for the root filesystem.
That may imply Linux due to the lack of viable alternatives, but really, it should be built with users in mind, rather than perceived as a technical challenge of putting Touch on top of Linux.
(And please, stay away from Java in your next mobile OS. Among other things, you will ward off those of us, developers, who let's say are not neutral wrt. to certain programming languages. Edit: make Java optional if you wish.)