June 04, 2010, 12:57 PM —
Here's what I used to think:
The wild success of Linux in embedded space would lead to application development on an unprecedented scale for the various mobile Internet devices, which would lead to a renaissance of application development on the Linux desktop, which in turn would lead to the Ultimate Success of Linux on the Desktop and a world where we will finally get to say "This is the year of the $Expletive! Linux desktop."
Now, my blueprint of Linux success goes something like this:
The wild success of Linux in embedded space will lead to application development on an unprecedented scale for the various mobile Internet devices. Period. And we still don't know if we will live in a world where we get to say "This is the year of the Linux desktop." $Expletive!
This is not some death curse on the Linux desktop environments, nor the "traditional" PC platform. The market is not done with this niche yet. Yeah, I said it: niche. Because very soon, I believe, most consumers are going to use computers they carry in their pockets and purses--and nothing else.
I can very easily see a market where there is only one computer per house, perhaps to take care of things like bills, music and photo storage, and the like, and all of the rest of the computing devices are mobile devices like smartphones and tablets. For families who can shift all of their computing needs into the cloud (think Mint.com instead of Gnucash), or can store big files on USB storage devices, having an actual computer may soon be unnecessary.
I don't think there will be wholesale abandonment of the PC platform. This week's announcement from AT&T that they were essentially eliminating the unlimited data plan was a clear sign to me that besides being greedy, AT&T is trying to prevent their networks from getting swamped with traffic that they can't handle. To be fair, I am not sure the other major US carriers, Sprint and Verizon, are in much better shape.
For now, then, I believe we will not see the total replacement of the PC/desktop, because cloud devices still for short of PC functionality and the cloud itself isn't strong enough to handle the bandwidth that a shift to all-cloud or mostly cloud would entail.
This puts us in an a situation where the PC and the MID platforms are going share some audience, but I do not believe the success of one will automatically equate to the success (of failure) of the other--because they are independent of each other.
Just today, for example, IBM VP Bob Sutor tweeted "Lest we forget: 'Android is based on Linux.' The rise of Android is part of the success of Linux." This is very true. But I would add that the success of Android doesn't effect the success of desktop Linux one whit. Apps built to run on Android don't run on my Linux desktop machine, nor can I imagine anyone porting them over any time soon.
The success of MeeGo or Linaro, however, could make a difference, since under the hood these distributions share much of the code from a "traditional" desktop distro. ChromeOS could be tossed into this example, too, except as far as I know, ChromeOS won't have a lot of native apps, favoring instead web-based applications. Any improvements on these platforms could theoretically have a straight-upstream effect on desktop distributions.
(Unfortunately, that success is a big "if," because Linux-based netbooks have yet to take off. The current market hype about tablets and smartphones may translate into genuine momentum towards these form factors and away from MIDs that are basically midget clones of laptops, which in turn are mobile copies of desktop PCs.)
Regardless of the success or failure of netbooks, I don't believe application development on netbook platforms will aid the "traditional" Linux desktop market. Right now, much of the development on netbooks seems to be centered around building more efficient/pretty interfaces and porting existing Linux desktop applications to netbook architectures. Other than application-based menus from Ubuntu's Unity project, I have yet to see anything significant come out netbook development that would benefit the broader desktop ecosystem.
It comes down to this: netbooks, similarities to PCs aside, have a different mission than traditional desktop devices. They're not meant to be heavy computing machines to do a lot of work, except in a pinch. They're lookiloo devices: look at your pictures, look at your movies, look at your documents, look at the Internet, and loo--er, listen to your music. Smartphones and tablets have exactly the same mission: let the user see anything they can, but there's few options to actually create content on anything but a small scale.
PCs let you create content on any scale you want. Yeah, I can write an e-mail on my Android phone, but it's going to be short. And I wouldn't dare try to write this blog on such a device. Or edit a video. Or build a spreadsheet. That's the mission of the PC.
MIDs are for consumers. PCs are for creators.
With such disparate missions, I question how much cross-platform application development is going to happen--even on MIDs with similar Linux tech as a desktop machine. Because even if there's a huge influx of apps built for Linaro, MeeGo, or Ubuntu Netbook, how many desktop users will find them useful on their desktops?
The success of Linux on the desktop will not depend on the success of Linux on embedded devices. There's too many differences between interfaces, form, and mission to provide a common link at the application layer. The success of the Linux desktop, as always, will rely on its own strengths and weaknesses.