I'd love to use an Ubuntu phone, but the last couple of times I've tried, it was still pretty awful to use.
Currently, the only phone I could install it on is a Galaxy Nexus and it's not surprising that the OS would be slow there. However I've read that it's still stuttering on the Meizu MX4 (in this german review: http://www.golem.de/news/meizu-mx4-mit-ubuntu-im-test-knapp-... ).
Also, the lack of Ubuntu Convergence (the "killer feature"), is disappointing.
>However I've read that it's still stuttering on the Meizu MX4
I can confirm this. I do feel it got better with the last update, but it is still noticeably laggy in situations where it has little reason to be. It doesn't help the primary screen, or hub, or scope (whatever it might be known as) has a tendency to refresh itself frequently, and draws its various components in a very unpolished manner.
With that said, it isn't all bad - I've been really impressed with the podcatcher "Podbird". If every Ubuntu touch application was as well done as this, I could probably start recommending it to people I know. Here's a recent video of the desktop build, which is largely identical.
I've been running the Ubuntu Aquaris E5 since it launched. The main reason I want to stick with it through the difficult early time (browser crashes continuously), is that I want a proper Linux environment on which to develop applications for phones/tablets. I hate having to be forced into Java on Android, and IOS is too locked down and you can't use custom hardware. I hope they stick with it (the main phone guy recently left Canonical) but I can only agree that it is now becoming extremely urgent for Canonical to ship convergence, as without it, this OS is a poor cousin to the competition, even if the user interface is quite clean and fresh by comparison to the gaudy design of Android and IOS. It definitely has potential as much more of an open, blank slate for developers than the others, but the key word is "potential". It's not there yet.
I know a lot of geeks like to think Convergence is a killer feature as a device feature, but I think this is completely mistaken.
Convergence is about data, right? You want all your data to appear magically on whatever form factor of computing you are using.
The key mistake of the concept of convergence at the device level, is that it runs against Moore's Law.
Modern smartphones are pretty darn powerful, but the SoCs that drive them, are also pretty darn cheap. That means that whatever shell you want to dock your phone with, can just have the same SoC and be a computing device in its own right.
It seems odd to me that you'd want to have lots of shells (be they tablet shells, laptop shells, or docks with monitor/kb/mouse on a desktop) sitting around not able to do any computing, because their CPU is somewhere else in your pocket. For a few extra bucks per shell, you can make it a primary node in your computing life. Or for a few more extra bucks per shell, you can make it a powerful node in your computing life.
Convergence is much more convenient at the cloud level - all of your form factors have their own CPU/RAM/Storage (which can also then be varied, to allow devices with more power delivery to offer more computing power) and the data can flow magically between them. This is already a reality today. My watch, phone, tablet, laptop and desktop all have access to the same data and I can use all of them at once if I want to.
Canonical's speculative sales numbers may be further deducted by their Edge crowdfunding campaign that resulted 37K people participating forming only 40% (~13M $) of their target.
That's in the same ballpark as the 25k number the article comes up with.
For people who don't know, these numbers are very, very small. I wrote an app store dice puzzle game on Android no one here has heard of and it has 20x those numbers.
As a developer, you have to decide if the decreased competition on the platform is worth the smaller user base when you decide if you will write a version for that platform. So platform size numbers like these are important.
In my own case, the attraction of Ubuntu Touch is to actually ship an entire semi-custom device with custom software for vertical use cases, mainly in B2B. Different order of magnitude of revenue / profit margin, with much smaller unit sales. I don't think Ubuntu Touch stands a chance with consumers at least for the next couple of years.
This article makes me uncomfortable, it's pretty light on information and then reveals the author is a contributor and expects donations - shouldn't such a contributor have more information or at least more to say than "Canonical want to sell x, but x could be anything, so the results could be anything"?
Being a contributor to an open source project doesn't automatically give you any insight into the business goals of a company sponsoring the same project.
The worst thing to happen would be for early-adopters to get screwed over if/when Canonical decides to drop support for the mobile OS.
The truth is that just like Windows, Android will dominate market-share for a long time, but there are millions of linux users out there who will gladly purchase a pure-linux phone if a decently-polished one with LTS can be delivered.
It sucks that the market beyond Android and iOS consists of Windows-phone as a distant third and everyone else scrambling for 0.1%, but Linux can grow here too, cause Android is (just like Windows is/was) a malware breeding-ground (and worse, considering Google is the for-profit version of the NSA).
Tell me with Samsung Bada. Not only the user experience was bad. Samsung was promising to update my old Wave to Bada 2 for at least 6 months, and then drop the support. It was the last time that I buy a Samsung phone. Now I'm very happy with my BQ phone that simply works fine.
I'm not trolling, but genuinely curious. Android runs on a Linux kernel, as does Tizen and probably several other phone OSes. What makes the Canonical phone more Linux than these for Linux fans? Presumably it's more like a desktop linux distro in some ways, but if so what ways?
Android uses the Linux kernel. That's where the similarities between Android and GNU/Linux end.
Ubuntu Touch is GNU/Linux. It is largely[1] made up of "standard" GNU/Linux components (e.g. glibc, systemd [soon], BASH, gcc, Pulseaudio, Gstreamer, NetworkManager, Telepathy, oFono, BlueZ, deb, apt).
This has several advantages:
1. People familiar with GNU/Linux can re-use their existing knowledge
2. The GNU/Linux ecosystem is developed by a lot of different parties. If you feel you can do better you can write your own components and swap them in (e.g. as Canonical did with Mir)
3. A community (or business) can develop their own "spins" (distros), which could extend the lifetime of your mobile (security fixes!)
4. It's a real pleasure to develop software on your Linux desktop, knowing there'll be less friction getting it to work on Linux on the mobile.
[1] The exception probably being using Mir instead of Wayland
Linux on the mobile has made some great progress in the last couple of years. A large part of this progress is thanks to the efforts of Jolla (SailfishOS), Canonical (Ubuntu Touch) and Samsung/Intel (Tizen).
Linux on the mobile can today make use of Android drivers thanks to libhybris (https://en.wikipedia.org/wiki/Hybris_%28software%29). The overwhelming majority of software that makes up "Linux on the mobile" is made up of mature FOSS components (e.g. Qt, systemd, Pulseaudio, Gstreamer, ffmpeg/libav, NetworkManager/Connman, Telepathy, oFono, BlueZ, rpm/deb, apt/zypp -- Wayland for SailishOS [and as of recently Tizen]).
I think at this point Linux on the mobile has a mature base. We now need a mature, polished, FOSS set of core applications. At that point (thanks to libhybris, and being able to leverage Android drivers) -- I think a self-sustaining community will be able to take over. I think we're about 2 years away from this goal.
I currently use Jolla (SailfishOS) as my daily driver, and I'm extremely happy with it.
One more than there should be by rights! Mine fell in a river at night a week ago, I thought it was gone but then I saw a light flashing from the bottom.. It was the screen flickering, coming and going for about a minute like it was calling to me. By the time I got in it had gone out but I found it by diving and feeling around on the bottom.
Packed it in rice and 4 days later it came back to life :) I'm very happy, great phone and surprisingly hardy!
I'd love to see the Ubuntu phone try to do things that the iPhone just can't do - things like give developers the tools they need to write apps: onboard, i.e. use the thing itself to build software for it.
Anyone know if thats possible, or do you still have to have an external cross-compiler/IDE to write proper apps for it?
> ...write apps: onboard, i.e. use the thing itself to build software for it.
I do this on iOS using Pythonista. It includes a basic on-device GUI design tool, hooks into various iOS built-in libraries and some useful third party Python libraries. There are even a few (admittedly trivial) apps published to the App store written in Pythonista. There's also Codea if you prefer Lua to Python. I tend to do most work in an iPad as a phone is a bit cramped for working on code.
I think they've done the right thing in catering to their existing base of enthusiastic developers. Trying to mass market it right now would just get them a bunch of bad reviews from consumer electronics sites that don't understand it and don't have the time to bother (to be fair, their readers are just more interested in the what color the next iPhone comes in). I think they've also done the smart thing in bringing it to countries where Android and iPhone have not already obliterated the whole market. That's a bummer for me though as I'm a US Ubuntu user and I'd love to buy one of their phones. (I know I can now, but 2G or wifi is not worth paying my monthly bill for). Last I heard they're working with a manufacturer to bring a different option to the US market. I'll be first in line to buy it and I'm excited about building whatever applications I find that I need on it that may not already be there. I think they could do well by really marketing to what they do and the others can't: no questionable agreements to sign, ability to choose what to do with your device that you bought and own, and nobody using your information. Sadly that's a small market, but it's probably large enough for them to keep the lights on.
I'd love to see a "laptop shell" where once i slip in my ubuntu phone, it goes from the phone os, to the regular ubuntu desktop experience. There's a lot of advantages of only having to own one device. For instance, as long as i'm updating my phone every 2 years... why not get a free laptop upgrade as well!. I take the train home from work, every day I struggle with the tether from my phone to my laptop. If it was just one device life would be a lot simpler.
27 comments
[ 2.6 ms ] story [ 59.8 ms ] threadAlso, the lack of Ubuntu Convergence (the "killer feature"), is disappointing.
I can confirm this. I do feel it got better with the last update, but it is still noticeably laggy in situations where it has little reason to be. It doesn't help the primary screen, or hub, or scope (whatever it might be known as) has a tendency to refresh itself frequently, and draws its various components in a very unpolished manner.
With that said, it isn't all bad - I've been really impressed with the podcatcher "Podbird". If every Ubuntu touch application was as well done as this, I could probably start recommending it to people I know. Here's a recent video of the desktop build, which is largely identical.
https://www.youtube.com/watch?v=SLSqz7skYvw
Convergence is about data, right? You want all your data to appear magically on whatever form factor of computing you are using.
The key mistake of the concept of convergence at the device level, is that it runs against Moore's Law.
Modern smartphones are pretty darn powerful, but the SoCs that drive them, are also pretty darn cheap. That means that whatever shell you want to dock your phone with, can just have the same SoC and be a computing device in its own right.
It seems odd to me that you'd want to have lots of shells (be they tablet shells, laptop shells, or docks with monitor/kb/mouse on a desktop) sitting around not able to do any computing, because their CPU is somewhere else in your pocket. For a few extra bucks per shell, you can make it a primary node in your computing life. Or for a few more extra bucks per shell, you can make it a powerful node in your computing life.
Convergence is much more convenient at the cloud level - all of your form factors have their own CPU/RAM/Storage (which can also then be varied, to allow devices with more power delivery to offer more computing power) and the data can flow magically between them. This is already a reality today. My watch, phone, tablet, laptop and desktop all have access to the same data and I can use all of them at once if I want to.
For people who don't know, these numbers are very, very small. I wrote an app store dice puzzle game on Android no one here has heard of and it has 20x those numbers.
As a developer, you have to decide if the decreased competition on the platform is worth the smaller user base when you decide if you will write a version for that platform. So platform size numbers like these are important.
http://www.phoronix.com/scan.php?page=news_item&px=Ubuntu-Ph...
Spoiler alert: Ubuntu phone can be found among the users, but it's pretty much behind everything else in terms of market-share.
The truth is that just like Windows, Android will dominate market-share for a long time, but there are millions of linux users out there who will gladly purchase a pure-linux phone if a decently-polished one with LTS can be delivered.
It sucks that the market beyond Android and iOS consists of Windows-phone as a distant third and everyone else scrambling for 0.1%, but Linux can grow here too, cause Android is (just like Windows is/was) a malware breeding-ground (and worse, considering Google is the for-profit version of the NSA).
EDIT Also Android functionality is gradually moving into the propriety Google libraries.
Ubuntu Touch is GNU/Linux. It is largely[1] made up of "standard" GNU/Linux components (e.g. glibc, systemd [soon], BASH, gcc, Pulseaudio, Gstreamer, NetworkManager, Telepathy, oFono, BlueZ, deb, apt).
This has several advantages:
1. People familiar with GNU/Linux can re-use their existing knowledge
2. The GNU/Linux ecosystem is developed by a lot of different parties. If you feel you can do better you can write your own components and swap them in (e.g. as Canonical did with Mir)
3. A community (or business) can develop their own "spins" (distros), which could extend the lifetime of your mobile (security fixes!)
4. It's a real pleasure to develop software on your Linux desktop, knowing there'll be less friction getting it to work on Linux on the mobile.
[1] The exception probably being using Mir instead of Wayland
Linux on the mobile can today make use of Android drivers thanks to libhybris (https://en.wikipedia.org/wiki/Hybris_%28software%29). The overwhelming majority of software that makes up "Linux on the mobile" is made up of mature FOSS components (e.g. Qt, systemd, Pulseaudio, Gstreamer, ffmpeg/libav, NetworkManager/Connman, Telepathy, oFono, BlueZ, rpm/deb, apt/zypp -- Wayland for SailishOS [and as of recently Tizen]).
I think at this point Linux on the mobile has a mature base. We now need a mature, polished, FOSS set of core applications. At that point (thanks to libhybris, and being able to leverage Android drivers) -- I think a self-sustaining community will be able to take over. I think we're about 2 years away from this goal.
I currently use Jolla (SailfishOS) as my daily driver, and I'm extremely happy with it.
Packed it in rice and 4 days later it came back to life :) I'm very happy, great phone and surprisingly hardy!
Anyone know if thats possible, or do you still have to have an external cross-compiler/IDE to write proper apps for it?
I do this on iOS using Pythonista. It includes a basic on-device GUI design tool, hooks into various iOS built-in libraries and some useful third party Python libraries. There are even a few (admittedly trivial) apps published to the App store written in Pythonista. There's also Codea if you prefer Lua to Python. I tend to do most work in an iPad as a phone is a bit cramped for working on code.