Chrome OS will presumably include Google Native Client support so that native code can run (with some limitations) (long but good video on Google Native Client: http://www.youtube.com/watch?v=zgng4C18nNk).
I'd much rather have an OS with the capabilities to sandbox applications, but that can still run an arbitrary binary. For example, I cannot imagine hardware device manufacturers writing drivers that are "cloud apps".
Chrome OS doesn't support native apps but the drivers are regular Linux drivers. Of course, you probably can't install drivers, so whatever drivers come with the computer are what you get. I don't see this as a problem since netbooks don't have slots and most USB devices use class drivers.
If you can't run Skype and other instant messangers, interenet experience is going to be incomplete. I wonder of they are going to do something about it.
Where do you draw the line? I'd love text editors and RescueTime as well. My hope is that Google will succeed in driving innovation on the Web and these types of apps will become first-class citizens.
Right now all netbooks sold out there have a camera and it is mostly used for video calls with Skype. I don't think netbook manufacturers would give such a major feature up.
I see Google has videochat in GMail with a browser add-on, perhaps it will come pre-installed with Chrome OS but it would still have little installed base.
I can't seem them leaving the code required for voice/video on Google Talk off of this thing. This is interoperable with any xmpp client which happens to support libjingle.
No, but openness has a lot of advantages. It means that it's free, it's guaranteed to be supported by someone even if the founders all go away, you can trust it more since everyone can look at the source, and so on.
You can't see the code inside Google's network so even though the client app is open you have no more proof that all your calls aren't being streamed to Langley or Beijing than you do with skype.
An you can't have pay for service like calls to 'real' phones because if the billing was done in the client you could simply remove it, forcing all the billing to be done at their servers - which makes it much more expensive.
XMPP is an open protocol. There are encryption extensions. Gtalk is used for routing, so it's route-around-able if you have concerns about snooping, still letting you communicate with grandma's gtalk account who doesn't have concerns about the FBI. This is much harder to do with software like Skype (especially considering the obfuscation that has gone into hiding the operation of the Skype code).
An you can't have pay for service like calls to 'real' phones because if the billing was done in the client you could simply remove it, forcing all the billing to be done at their servers - which makes it much more expensive.
You mean like how land lines and cell phones do their billing on the provider's servers and don't do the billing in the phone itself because you could simply remove it?
Search engines are completely different from downloaded software - you can't avoid reliance on the provider unless you feel like writing your own web crawler and waiting for it to gather web content (even though it will never catch up with the volume of new stuff constantly being put on the 'net).
With HTML 5 browsers should be able to access CPU processing power and other parts of the computer previously available to only native applications. So this is possible.
[Google: We believe the web platform is a much simpler way, where the machines are essentially stateless, more cache-like, but can still be performant. And yet much easier and simpler for individuals to use.]
My experience is, that for certain tasks, native client application is still much more convenient. Evernote is a classic example - although online version of this note-taking software is available, I still enter a note quicker in native app, which then synchronize the data on the cloud.
It's the same with photo editor, word processor, etc. So, what is really needed are good native applications, which are able to synchronize their data with the cloud. But maybe Google will convince me that web-only platform is better, we will see...
I haven't tried out a netbook, but the reason I don't turn off my Macbook is because I can simply close the lid, and I'll immediately get a login screen when I open the lid. Boot time is the least of my worries.
I feel the same way; but I suspect people in different parts of the world might think differently.
E.g., in my experience, some people have been very vocally opposed to the idea of leaving computers on overnight owing to the energy wastage etc. They were all from Europe, and from precisely the same types of environments where we would usually leave our machines on permanently (eg postgrad university labs).
Not to generalize, but I wonder if (at least at that time) there was not more sensitivity to such environmental issues in Europe.
Unfortunately (at least on PCs) the suspend-resume functionality just has never seemed to work properly ...
I carry a n800 everywhere. I got it two years ago - I had various PalmOS products ("instant on") before that. What blew my mind when I got it is that the n800 is always on, and there is no reason to turn it off.
* When I walk in the door at home, it automatically connects to my wireless access point.
* I can ssh into it or scp files to or from it using my laptop (after unsuspending the laptop and waiting for it to get back into X :-O) regardless of where it is physically.
* I use rsnapshot nightly to back it up.
* I tap on the screen and it is "instant on" because it was already on, just the backlight was off.
In short, I never turn it on, I never plug it in (other than to recharge :-()
That is the way Life Should Be (other than needing recharging).
Most people don't realize What Could Be because their computer usage paradigm (imagination) is limited by what Windows + Intel is able to do.
There are a class of users that really only do a bit of email, a bit of web surfing, and maybe a bit of document editing. (Like my parents eg). From my experience such users tend to remain longer on older hardware, and its turned off most of the time.
I daresay such users would naturally gravitate to whichever environment gets into a usable state the fastest.
I treat my laptop like I treat my phone -- always on but it spends most of its time asleep. Only reboot in case of OS upgrade or system failure. I don't really get why people don't do it this way... just close the lid.
Well it presupposes you have an environment where it actually works consistently.
I cannot remember the exact number & types of machines I've used or had provided to me at customer premises over the last few years, but I'm pretty sure its > 10. Its only ever been my MBP that wakes up consistently with everything still intact. <= 10% is a pretty poor success rate...
even though Chrome OS is meant for Netbooks, the number of people that own netbooks and use them to communicate via Messenger, Skype etc is enormous. And people who do use Skype on netbooks will not opt for the Chrome OS because you cannot use Skype in an online application like MSN or Yahoo for example.
Personally I use skype a lot and was looking forward to Chrome OS as a substitute for Windows/Linux on my netbook, and was even considering buying a netbook for my girlfriend just for this purpose, however now I will not be doing that.
I think those criticizing Chrome OS for missing obvious features are missing the point. Google is making a long term play here -- and not even that long term. In five years, the distinction between the web and desktop will have completely eroded for non-professional users.
Chrome OS supporting only web apps is not like the iPhone only supporting web apps; it's like how the Palm Pre only supports web apps.
I found this OS by opensuse very similar (not sure what to call it, as it is mentioned that it is an OS around the chrome browser and not affliated to Google)
There are some issues with the data portability IMHO. For example, if you edit doc in Office Live and then wanna send out with GMail, how would you do that? Or how will you check out the photos in the camera and see them instantly? The different security model for web app made those tasks harder.
Seems like you could have some type of remote directory that would allow normal file operations like uploads to work while still storing the data in the cloud. Something along the lines of a samba share.
Finally..! Anyone else hear the war drums? Microsoft should definitely be worried.. Very worried. In fact msft should have been on the ball with this years ago but they were too concerned about losing revenue from licensing fees and their dominance over the desktop. Just goes to show those afraid of change will be destroyed by it.
It's also interesting to note that this was the computing model used 'back in the day.' Everything was dumb terminals logging into a mainframe. Looks like we had it right the first time.
There's always been this kind of pull between putting more smarts at the end points vs having everyone login into one behemoth multivac (mainframe back then, the Internet now). It's obvious that Google is putting its stock in the latter approach. I would go out on a limb and say that computing has been heading in this direction for a long time and companies like msft have slowed it down as long as they can but the flood gates are about to burst open..
This should definitely raise some eyebrows over at W3C.. This is a huge development.
It's a tectonic shift in computing. We're going back to the dumb terminal/mainframe model. Apple will weather it just fine I'm sure b/c they control a shitload of devices. Microsoft will not. If they were dead before, they're worse than dead now.
Google will control the other half of devices but hopefully in a 'non-evil' way. I sincerely hope owning a device with Chrome OS doesn't force a user into using Google services although I'm sure there will be some tie-in.
What google are proposing is to sell a web browser only appliance - for a limited number of users who only use their windows laptop to run ie this is great, they no longer have to worry about viruses or steal ms-office to write a simple letter.
Google can have many more apps for this than apple, because every website is n 'app', developers can write apps simply using javascript/html - there is no need to buy a mac, learn objective C and join apples dev program and learn what secret signs you need to know to get your app approved.
Google don't need to approve the app in the same way that they don't need to approve your website.
To Microsoft, who make their money from business customers running SQLServer and Windows server licenses, this is about as relevent as Nokia shipping a new edition of symbian.
Problem with your argument is that everything that Chrome OS runs (web apps), also runs on Windows, Mac, Linux. The reverse is not true. Win32/64/PEF/Mach-O binaries do not execute on Chrome OS.
And this was tried in businesses with the client/server model in the 90's as well. That failed, as much as I think what Google is aiming for here will fail except for a small niche of computer users (unless there is a light-years advancement in webapps replacing desktop).
In any event, I actually have no worries for Microsoft. They have a huge share of the web space already, and are already targeting webapps of their products (even if they have to play catch up, they'll get there).
Apple will have problems because their market consists of people that like their products just to be different and to appear in vogue. When that can be accomplished without buying overpriced hardware, see Apple's share of the PC market drop.
The code signing is a bit worrisome, but I suppose this isn't really a hacker OS. And it along with the sandboxing is really the only "complete" solution to malware.
Flash on linux is currently less than stellar, and I see that as something that would be used pretty commonly. Any ideas on how they plan to deal with that?
Why would we expect Google to fix Flash Player? Fundamentally it's Adobe's problem.
It should be easier for Adobe to make Flash work on Chrome OS than on Linux, though, because Chrome OS doesn't have the diversity of duplicate APIs in normal Linux.
This whole endeavor still doesn't make a whole lot of sense to me. If you want to make an idiot-proof OS for Grandma, fine, that's a great idea. But why shoehorn the whole cloud-computing philosophy into it? Most people that a very simple OS would appeal to are not going to know or care what cloud computing even is. And furthermore I don't really see much value-add in that anyway, seeing as how the device is MOBILE and you can already take your applications and information with you...
Did you believe it when you heard it back then? I didn't.
Apple offered "web apps" to placate would-be developers and buy time to refine the real SDK and work on the App Store model of delivery. (I know that jailbreakers love to give themselves credit for the SDK, but I think that fails the Principle of Parsimony in the case of Apple, who tends not to disclose things until they're ready, and who holds user experience - which web apps do not provide - above all else.) A lot of people didn't believe an SDK was coming until it was announced, but I don't think any of them really thought it through.
I do believe Google's plan is to make the browser the centerpiece, on the other hand. Their model is to ensure a good web experience, to drive ad revenue, isn't it? Why would they add a native app API?
I'm tired of this meme. I bought an iPhone because it's a browser with a permanent 3G connection. I use gmail, google reader, wikipedia, iPlayer and many other web apps that are simply saved bookmarks. Despite the initial app goldrush the few free apps that I've installed seem to be ad supported and are generally simple enough that they would probably work fine as Html5 apps.
Obviously some apps, notably games, need native access but the initial bitching came from existing Mac developers with Cocoa experience. The big problem For everyone else was lack of access to iPhone hardware features via JavaScript.
I think the Trojan horse here is that Chrome itself will force you to use Google as your "base cloud" (for preferences, files, login, docs, mail, etc.), though then you're also free to use other web services beyond that.
If you want your own base cloud, it looks like you'll have to build your own system and ship it.
Making a development environment for ChromeOS that works on ChromeOS would be a first step towards answering PG's clarion call for a woman with a hammer, no?
When I submitted this, the story was half-told. Apparently, Gears will still be allowed to store some things locally. But all data will be synched to the cloud automatically at all times.
Current operating systems already support multiple computing paradigms (including the ability to run apps and store data in the "cloud"). Forcing people into a specific paradigm has been historically proven to be disastrous (Network Computers, Javastation, etc).
I'd much rather have an OS with the capabilities to sandbox applications, but that can still run an arbitrary binary. For example, I cannot imagine hardware device manufacturers writing drivers that are "cloud apps".