[wyz9]

Mark Shuttleworth and Canonical are important to ‘desktop Linux’ mainly because they are attempting to kill it. Note that they do not even advertise Ubuntu as a GNU or Linux distribution.

It seems the current strategy of Canonical is to change enough of the system so that developing cross-distribution will become more and more of a hassle, and trust software developers to just target Ubuntu due to its market share in Linux-land.

Unity was a step in the direction. It’s not critical to other programs, but Unity itself appears to be very hard to port to other distributions (and to be honest it was the first thing that actually made Ubuntu distinct from other Debian derived distributions).

Mir takes it a step further, now wm and toolkit developers will have to target either just Wayland, and lose out on the vast Ubuntu userbase, or target just Mir.

Canonical does its best to bring closed-source commercial desktop applications to the operating system through the Ubuntu app store. With good reason: they know the developers of these commercial applications will only target Ubuntu, since unlike open source programs where the distribution’s packagers do the work of bringing your application to their OS for you, that can’t be done very well with just binary packages compiled against x version of y library. Thus forcing users who want to use one of these applications to switch to Ubuntu.

EDIT: and let’s not forget that Canonical ships what is basically spyware with Ubuntu. Local searches on your desktop should not be used to help Amazon advertise. Shuttleworth’s reaction to the complaints were extremely cynical as well.

[keithpeter]

"Mir takes it a step further, now wm and toolkit developers will have to target either just Wayland, and lose out on the vast Ubuntu userbase, or target just Mir."

Won't the toolkits simply support existing widget libraries? I mean GTK-Mir and GTK-Wayland or QT-Mir and QT-Wayland &c? Else there will be something of a dearth of applications! I'm genuinely asking as this is an area I don't know much about.

PS: There is a privacy settings manager in system settings probably as a result of the reaction to the Amazon search thing.

[slgeorge]

"Canonical does its best to bring closed-source commercial desktop applications to the operating system through the Ubuntu app store. With good reason: they know the developers of these commercial applications will only target Ubuntu, since unlike open source programs where the distribution’s packagers do the work of bringing your application to their OS for you, that can’t be done very well with just binary packages compiled against x version of y library. Thus forcing users who want to use one of these applications to switch to Ubuntu."

First, I drove the App Store both strategy and was responsible for development - in other words _I know_ why we did it. I am my own citation!

Strategically, what do you think the biggest reason is for why end-users don't move across platforms? Guess what it's application software, specifically lack of important application classes and brand names end-users are familiar with. It comes across in all the user-research. So, what you want to do is to encourage commercial software developers and prove out a market - which leads to a situation where familiar applications are available on an unfamiliar platform.

Guess what the other problem is in the Linux ecosystem compared to other 'alternative platform' ecosystems - the lack of a set of commercial software developers. Think back to Apple in the mid-90's a period where the media thought they "were dead". They could still attract thousands of developers to MacWorlds and there were lots of software companies developing great software for users of the platform. Desktop Linux doesn't have that because it never created a market. The result is not as much fully polished, long-term end-user grade software.

So finding a way to create a market place and create the ability for commercial software developers to target the platform is good for end-users. That's the strategic reasoning.

Next, consider the problems that developers face.

If you go to some conferences and ask developers why they don't target Linux - go ask at some Game developers conferences - they'll tell you two things a) Linux users don't pay for anything/market is too small b) developing for Linux is too hard. Now a) is just a hard problem and the only thing that changes is it is time. But, when you dig into b) what you discover they mean is that the development tools on Linux are hard and that the swathe of packaging options is confusing. Which is exactly why Ubuntu started cutting through the "tyranny of choice" to provide information on how to develop/port - see http://developer.ubuntu.com . It's also why we reduced the complexity of packaging, provided tools and a web portal which is what developers are used to on other platforms - now a developer just has to use the autopackaging tool and the system takes care of the rest of it.

Result is that there are more commercial applications available on a desktop Linux (Ubuntu) than every before. Result, some of the commercial tool developers are targeting a Linux desktop in a way never done before. I'm proud of that. And, since by definition it's easier to port from one distribution of Linux than it is to port from another platform to Linux - you should be proud of that as well.

Either way, that's the _facts_ on _why_ Ubuntu targets developers and full explanation of how it's not an evil genius plan.