Submitted by At the recent Linuxcon Conference in Portland, Oregon, there were hints of new Moblin-powered hardware being announced at the upcoming Intel Developer Forum. Normally, this would be moderately interesting news, but some of the ambitious comments made by Linux luminaries at Linuxcon merit further examination. People from the Linux Foundation, Intel, and IBM spoke at the conference, and it’s evident that they see the netbook market as the epicenter of the movement to raise Linux’s profile in the consumer space, and whittle away marketshare from Windows. Update: Intel has also announced Moblin 2.1 for phones.There are a couple of interesting quotes in the PC World coverage of the conference:
“We need to stop pretending that it will be a drop-in replacement [for Windows] and make it something better,” said Bob Sutor, vice president for open source and Linux in IBM’s software group.
Linux failed to retain its early lead in netbooks because it was trying to emulate Windows. “We were trying to win at their game. We in the Linux community are trying to be successful by mimicking what someone else is doing successfully. To me, that is a losing strategy,” he said.
Linux got an early lead in the netbook market because of its favorable economics. The profit margins were so thin that the manufacturers were drawn to Linux because of its lack of licensing fees. But Linux-based netbooks lost momentum early on due to two factors: Microsoft’s extremely aggressive price-cutting of Windows XP OEM licenses from the bargain bin, and, perhaps more importantly, the netbook vendors’ lackadaisical attitude about software development. To a large extent, after putting a fair amount of creativity into designing small and cheap laptops, these vendors failed to put a complementary effort into adapting the available Linux distributions to the hardware and smaller format. More importantly, they failed to support netbook users with a halo of web-based services that would have made using, troubleshooting, customizing and upgrading their new Linux netbooks fun and easy.
So when typical computer users bough these new mini-laptops, they found themselves not only in an unfamiliar environment, but unable to use these machines the way they would normally use them – that is, by installing Windows software on them. In comes Microsoft with cheap XP licenses, and the rest is history.
As much as we can all scoff at Apple’s claims that the iPhone is their answer to the netbook, there’s a lesson to be learned there. New iPhone users are faced with a radically unfamiliar computing environment, user interface, form factor, and software ecosystem, even more so than a Linux netbook buyer. So obviously the iPhone completely failed in the marketplace. Oh yeah, it didn’t. And not only because of the reality-distortion field. The iPhone overcame the unfamiliarity hurtle because its radical user interface was well-suited to its form factor, and because Apple built a scaffolding around the iPhone experience that (eventually) made routine personalization of the device easy.
The netbook’s strength and weakness is that it is a rather familiar form factor, so people expect it to act like a laptop PC. I believe that people will always be more accepting of alternative operating systems when they come in unfamiliar packages, such as a mobile device, DVR, NAS box, or alarm clock. But failing to target a very popular mini-laptop market just because there’s a built-in expectation of Windows-like functionality would be a pretty timid strategy, and it doesn’t look like Intel, IBM, and the various Linux distros are ready to throw in the towel.
When I first learned that Intel was putting money behind Moblin, I’ll admit I wasn’t enthusiastic. Intel is one level abstracted from the OS market, since they don’t make devices themselves, and I doubted that such a big company would be able to sustain interest in such a seemingly non-core project. But I’m coming around, because I now see that it’s really in Intel’s best interests. Intel had an easy time of it all through the eighties and nineties, because every computer user in the world was perpetually unsatisfied with the speed of their computer. But early in this decade, Moore’s Law turned around and bit Intel on the ass, when suddenly everybody’s three-year-old computer was still faster than they would ever need for their routine computing tasks. Intel has spent the past decade desperately trying to think of things for regular people to do with computers that would require vastly faster processors (video editing? virtual reality? weather simulations?) but have so far come up short. And where the real action is these days, in mobile and novel computing devices, Intel is working at a disadvantage, since it has robust competition in the low-power space, such as Freescale and ARM. But for devices that span the gap between low-power mobile devices and high-power PCs, Intel dominates with the Atom, and if they can figure out how to get mid-power computing devices to catch on big-time, then they’ll have a strong opportunity in that market.
As the presenters at Linuxcon pointed out, Microsoft is about to give the Linux netbook market a big gift: for strategic reasons, Microsoft can’t continue to sell Windows XP on any device. Everything has to move to Windows 7. But there’s no way that OEMs will be gettgng Windows 7 licenses for next to nothing like they were for Windows XP.
“What I hear when I talk to netbook vendors is Microsoft does not want to repeat the extremely aggressive pricing with XP Home. They want to significantly increase the price for Windows 7 netbooks,” said Dirk Hohndel, chief technology officer of Intel’s Open Source Technology Center.
So here’s Linux’s big second chance. Microsoft will likely let them lead on price again, and with Intel beind Moblin, there’s a good chance that the next generation of Linux netbooks will have not only a robust underpinning, but a pleasant and different user experience. Many different players (including Intel and Novell) are already discussing setting up “App Store” type sites for Moblin devices that are likely to appeal to a broad spectrum of users. Adding to this perfect storm is the rise of cloud computing and the usefulness of apps like Google Apps, which chips away at Microsoft’s desktop stranglehold.
Microsoft, of course, will play to its strengths, and the fact that Windows 7 is a very good desktop operating system and should run quite ably on varous kinds of lighter-weight devices and enable users to tap into the vast library of familiar apps will ensure that it’s a major player in this market, even if it will be a struggle for them to balance profit margin and marketshare in this space.
Jim Zemlin, executive director of the Linux Foundation says that within a year, “no one will buy hardware or software” and that the market will look a lot like the mobile phone market, with wireless carriers giving devices away for free with service contracts. “Microsoft’s economics don’t fit into that at all,” he said. I’m not sure I want to live in that world, but if that world is coming, then he’s probably right.
For your usual netbook I find it lacking. Taskswitching was way too slow and I don’t need all the social stuff.
For a touch based MIDs Moblin might be even better than Android, but very similar to Maemo.
I think you cannot forget Google in all this. They wanted to bring out a OS for a long time, way longer than Intel and I think they decided once XP is off the market for good will be the right time to do it.
Chrome OS probably has a much stronger technological basis than Moblin, although it is all vaporware at this point.
I think one Chrome OS sanctioned by Google will have a stronger marketing pitch than a lot of Moblins from Canonical and Novell and Xandros and ..
What makes you think that Google won’t be using moblin as the basis for their ChromeOS? After all, Intel seems to be doing a pretty good job in this space.
Well it definately is. Both Moblin and Maemo are part of the Gnome Mobile initiative and work and use all of the same technology. In the future they will be quite difference since Maemo is moving to QT.
The deal is is that Android and ChromeOS are going to be a much more radical departure from the traditional Linux/Unix model.
This means that Moblin, being based on Gnome and GNU/Linux (as in not Android) is a hell of a lot mature then anything Google could hope to produce.
There is the ideal “new from the ground up” and wart-free software that people want.. but then there is the years and years or real-world usage and development that you simply can’t pay for or replicate for any price.
Using a phone is now “computing”? A proud day for computing and computer science.
Maybe I’ve missed something, but practically all netbooks (with maybe some minor exceptions) are PCs. Nothing have changed except power consumption and price (both more than healthy), but under the hood all netbooks are basically PCs, typically equipped with cheap Intel chipsets.
Dear Mr. Zemlin. You have made quite a few venturesome but delusional comments in the past. This goes to the end of the list.
But if Mr. Zemlin’s views are right in some distant future, perhaps the most fundamental change in computing will be the separation of consumer-users (locked bulk products, phones, gadgets) and the rest (everything else). Which is kind of sad.
Edited 2009-09-24 03:25 UTC
It sounds like you’ve never used an iPhone or any other advanced smartphone. You know, these things can do a lot more than just make phone calls. I’ve been using personal computers since 1985 and I didn’t have a computer as powerful as my iPhone is until 2001. Aside from the user interface limitations, there’s nothing to prevent you from using the iPhone or a similar handheld computer for “real computing.” What do you consider computing anyway?
Intel is poised to steal a major portion of market share if they can continue to offer their OS in the face of Windows 7.
I hope that the system’s development community will get SSE3 optimized compilations of this OS to increase the speed and efficiency of netbooks.
My ASUS 1005HA netbook is fast, but Windows XP Home is slowing it down. http://bit.ly/44CHFm
As always when Mono flamefest is heating up, I took a quick peek at Boycott Novell (who love this topic).
I bumped into this:
http://boycottnovell.com/2009/09/24/moonlight-maybe-dead-now/
It seems they have/are doing a Silverlight Linux port (er, “Moblin port”) based on original Silverlight code, NOT Mono and Moonlight.
I used to work at Intel (and would like to again, with all this cool Moblin stuff) and there was one interesting point that I picked up whilst I was there that helps you understand the way they think. My boss told us this during a meeting and it rather helped me make sense of things.
They’re not a CPU design company. ARM is a CPU design company, they license their designs to manufacturers based on their technical strength. Intel *does* do CPU design – and do well at it – but unlike ARM they don’t make their living by selling intellectual property. Intel is primarily a manufacturing company; they really want their factories to make lots of stuff and for the stuff to get sold. That’s the way they operate.
From this perspective, Intel’s investments in Open Source make somewhat more sense. They’re not being altruistic – Intel have presumably identified an opportunity to increase their sales if certain software exists and works well on their hardware. It makes business sense. No reason *not* to Open Source the code, since they don’t (generally) sell intellectual property. One assumes that they’ve simply decided they ought to “help along” the state of the art a bit, so that there will be better software available, sooner.
Nb. I’m not informed by any insider policy knowledge on this one, beyond knowing their mindset is purely “sell lots of devices”.
You might note that they do sell *some* intellectual property, for instance they sell compilers and have been involved in IP licensing deals. So they do sometimes work in IP, just that IP sales aren’t the core motivator.
Thanks for including that insight. I think there’s a similar story over at IBM, which over the years has become more-and-more a professional services company. To some extent, for them, even selling IBM hardware is secondary. They’ll sell other people’s hardware and software as long as they’re the ones to get to run the multi million dollar decade long implementation. And that explains why IBM has turned into a Linux booster too.
Amuses me how things have come full circle in a way – back in the day they used to sell you a mainframe and software to do what you want was part of what you were paying for AFAIK.
A big period in the middle where the software companies (MS in particular) were vastly more powerful than the hardware companies. And now it seems like the hardware companies have – through Linux and Open Source – found a way to bring it back to the original state of affairs where they can be complete system vendors (and, as you mentioned, in IBM’s case that “system” extends waaaaaay beyond the devices and into all kinds of other services).