G. Pascal Zachary has written a very interesting article about the importance of Google’s Chrome OS announcement, and what the move to the cloud (internet!) could mean for Microsoft and the operating system in general. I do think he missed an important part of the possible consequences of the move towards the cloud: the hardware.
If you look at the history of personal computing, you’ll notice how focus has been moving up the stack. It started with the hardware – the hardware was key, the hardware was where the differences were made. In the ’80s, numerous different hardware platforms formed the basis for home computers, like the Mac and the Amiga.
Over time, focus has shifted away from the hardware platform to the operating system where the differences are made. The various different hardware platforms in home computing all lost the battle with Intel’s x86, and with Apple ditching PowerPC, this shift was complete. It’s now the operating system that differentiates one computer from the other; Windows, Linux, Mac OS X.
And now we’re moving even further up the stack. A lot of people seem to think that the operating system as it exists today is becoming ever more irrelevant; it will become nothing but a simple gateway towards applications on the internet in the cloud. Applications no longer care about what operating system they’re running on, much in the same way that operating systems currently do not have to worry about what hardware they’re running on.
The hardware part of personal computing has been standardised. It’s x86 all the way now. And here is where it gets interesting: because applications no longer care about what operating system they’re running on, the hardware platform has suddenly become less relevant and defining too. In other words: the x86 monopoly is now more vulnerable than it has been in a long while.
Because the operating system is becoming irrelevant, people will no longer care about using Windows or Linux, and with that, new opportunities are once again rising for other hardware platforms, such as ARM. Regular Windows won’t run on ARM, and with applications becoming ever less operating system dependent, people will be just as happy accepting an ARM machine running Linux as they were with an x86 running Windows.
And this is where the Linux world shines. Thanks to its open nature, it already runs just about everywhere, and a possible ARM netbook revolution will leave Windows in the dust while providing a massive opportunity for Linux to finally gain a serious foothold in home computing.
1. Look at popular “new” technologies
2. Expound virtues of one or more technologies
3. Claim it’ll kill off the old
4. ???
5. PROFIT
Rinse, repeat every so often.
*yawns*
Edited 2009-08-10 22:03 UTC
Everybody expects these things to happen over night. The replacement of X86 Windows will take a long time, it won’t be as fast as the replacement of Windows Mobile because Windows is much much more entrenched.
But it will happen and X86 will decline with it .. just don’t expect it to happen tomorrow.
yeah, exactly.
when did windows mobile get replaced? when did it ever own the market for that matter?
But look at the practice – XP-based netbooks have been battling with Linux for some time now, and the result is clear: XP won hands down. The appcompat and familiarity aspects were just too strong.
It’s possible that this will change at some point. However, since netbooks represent the “computing in the cloud” and “low cost” visions and even there Intel and Microsoft still dominate, this transition if it occurs is far from imminent.
Which is why most of the ARM netbook manufacturers aren’t positioning their devices as netbooks with really long battery life, but rather smartphones with a clamshell formfactor, a full keyboard, and no actual phone calling features, but phone OSes.
Not that it’ll work, but…
I agree with you. But remember, that the Linux installations on the Netbooks weren’t exactly great. Obscure distros, update breakage, non-working drivers, etc.
The OEMs shipped Linux, because they couldn’t ship anything else at that moment.
I doubt that had much to do with it. It had more to do with the cost of producing Linux netbooks. Switching to Linux meant that they had to select hardware that actually worked with Linux, not just what was the cheapest this week. They also had to educate, train and keep people with Linux support skills. Having just one support team with windows skills, was much cheaper, espesially as Microsoft sold them XP Home at a very resonable price
Most venders also made the “mistake” to sell two versions of their netbook, one with too little memory, too little disk running Linux, and one with decent hardware spec running windows. Guess what, people preferred windows versions. Especially as the Linux versions they provided forthese systems often was a bit substandard.
Even people that wanted to run Linux on their Netbook would buy the Windows version. If not for other reason that they got windows for free, in case they should be forced to use it for some reason in the future.
The appcompat and familiarity aspects were just too strong.
Ohh. I’ll be very happy if not a single one of “popular” pc-world bloatware programs will ever be spotted on my ARM smartbook (which I’m about to buy).
XP lost really. Cost has been cut to bone or hardware markers would just ship Linux.
Linux is now sold by almost all OEM at least as a token.
MS cannot keep running loss making departments for ever.
Basically all MS did was slow Linux take up. So slowing it development in the desktop area. It has not stopped.
Remember some of that development is wine the means for Linux to run windows application. Numbers of testers and developers speed up open source development.
Is progress now faster on desktop development of Linux than before netbooks. Answer is yes. Linux is on the way to winning just slowly.
I don’t think anyone expects x86 or Windows to remain as dominant or grab the kind of profits they once commanded.
However, articles like this seem to suppose the dominant players remain static.
Intel is positioning itself against ARM with a focus on low-power processors. Customer relationships are already there…
Microsoft might have been late to the web-game, but they’re fully in it now.
Now the question is about pricing and yes… support.
I really don’t expect many netbook manufacturers to support their own version of linux. I’m sure they will try, but they WILL fail. The costs are too high and are much better distributed.
People underestimate the amount of support needed. Both on the technical side and on the consumer end. On the MS side, MS does a lot of the driving to make sure everything works together, the user experience is good… On the Linux side, the distros will do it… and they won’t do it for free.
As long as Microsoft’s pricing is reasonable relative to the costs of a supported Linux distro, it will still be quite dominant. Just think of every other area in life where people are willing to pay a premium for a brand name (TVs…). It means a lot.
This is not to say Intel and Microsoft will remain dominant. It’s just not as guaranteed as many people think.
Hardware vendors will keep fighting in their own arena over price and some acceleration technology and users… won’t care. Is there any technology out there that offers 100x better X or Y over x86 ? That’d could be a game changer… I don’t think there is one today. It’s all small improvements here and there and x86 catches up (in its own Frankenstein way). The world doesn’t care about purity alone.
x86 will be replace, for sure.. where it makes sense. I got the idea the article suggests x86 will die. Not in my grandchildren lifetime.
Speaking of demographics, I think the base will shorten… leaving more options at every layer.
Google announces an OS, thus killing Microsoft and the whole x86 ecosystem. That was quick.
Sarcasm aside, x86 is way too entrenched, not only on the desktop, but in the server market, too. This means lots of money for Intel (and a little bit for AMD), which of course reinvests that money in x86, making the chips even better.
There isn’t anything from anybody that is in a position to compete. PPC is dead on the desktop. ARM isn’t here, either. (A handful of netbooks doesn’t count). Just because something different exists doesn’t mean people will flock to it for the sake of being different. After all, I can still by UltraSPARC laptops. Whoop-de-do.
No, x86 isn’t going anywhere for a long, long time. For that matter, neither is Windows.
I think one of the biggest limiting factors is flash. The web might be platform agnostic, but flash isn’t. Still, I’d be rather shocked if google wasn’t able to get adobe to port flash to whatever platform they want.
IMHO the cloud is just another buble, And I don’t think Google’s next os will be the hit everybody thinks it will be.
Half of the article was just repeating how “operating system is irrelevant”, not even bothering to use different words to explain it, and that made me lose the point, but finally the last sentence said it: “This is the year of Linux on desktop.”
As if.
Edited 2009-08-11 05:32 UTC
I have been making the point that Linux will weaken the x86 position years ago. It’s not happening, x86 only got stronger and currently no other architecture has processors that are competitive to x86. There currently is no business case for buying a laptop, desktop or server with anything else.
Yes, netbooks, but even there the case for buying ARM rather than x86 is dubious. Maybe I would buy one, but I really like the ARM architecture, but for an “average consumer” this motivation doesn’t hold.
I’d say it’s exactly the average consumer where this can hold. I mean developers are more knowledgeable in what processing power and architecture they need, and in many cases ARM doesn’t cut it, while for average netbook use it would do nicely. I’d say the netbook arena could be the first place where cpu choice shouldn’t matter, because of the generally low requirements.
You people (e.g. article writer above) just seem to forget some things: i) apps have to run somewhere, ii) not all apps are twitter-related, iii) you can’t always eliminate local storage, local hw, and local computing power necessity. And in these cases the OS and the HW will matter for a long time.
No I don’t think local apps will fade away, but I still want a ARM netbook. I’ll still be running local apps, it’s just they will be compiled for ARM. There is a great deal of Linux software that compiles for ARM. In the GNU/Linux world, processor architecture isn’t as a big deal. Everything is installed from repositories anyway, a normal user might never know their processor architecture.
I reaaally would like to see more MIPS in the market. But I think the author of this article is not thinking about the everyday user. That user that like fancy 3D games. That uses an “alternative” version of Photoshop just to change the face of his sister. The user that wants to use his camera to make a video of his little child and then edit it at his computer. etc, etc.
Yeah, most of this stuff can be made in a “cloud”… it’s possible to edit a picture and (i think) even a movie. But how fast will it be? Will it have all the features of the original off-line one?
Users like to have “the best” computer they can get. They dont mind to pay some extra just to have more Ghz and mor RAM. Even if they will not use it, they like the felling that they have a superior machine. They will look at you with your 1Ghz computer and will say “haha mine is 7 time faster than yours”. That’s why we almost dont see low-specs netbooks (like the original eee) anymore.
Like I said, I like MIPS, I like Linux, but I dont see a real market change in a near future.
This is another case of the following assumption being made in the IT market, that because the details of something become less relevant, then this will favor the alternatives to the dominant implementation of that thing. The most frequent example : ” The OS becomes more and more irrelevant, therefore Microsoft will die and Linux will triumph “.
The problem is that this silent assumption happens to be wrong in most of the cases, and also that its opposite can be made as well, and actually is true indeed more often than not : the details of something become more and more irrelevant, therefore the _alternatives_ to the dominant implementation become irrelevant along with their supposed differentiation and value proposition.
Contrary to what many people think, there’s nothing in the x86 architecture that keeps implementations to be whatever you want them to be. Multi core ? check. Fast, large-die ? check. Small, power efficient ? check. And in the case of OSes, it’s even more obvious since the OS actually hardly does anything of the actual work that you’re using your computer for, apart from switching tasks, driving your printer and webcam and configuring your network. And that brand new webcam that your mother bought, works as advertised when you plug it on a windows box, and it’s a brick when connected to your linux box, and that’s a fact.
As the OS, the metal, the network protocol, the storage media, etc, disappears under the next layer of abstraction and becomes “irrelevant”, there is less value in bothering to switch to alternatives, not more.