Google has just unveiled its Chrome OS operating system during a press event at the company’s headquarters, and it’s pretty much exactly what we expected it to be: a streamlined Linux kernel booting straight into the Chrome web browser. The code is available starting today.
Web web web web web web web
The Chrome OS replaces the traditional software stack on computers with a more lightweight one, all made possible because it can only run web applications. Instead of going from BIOS to kernel, services, start-up applications, and eventually the browser, Chrome OS will cut a lot of stuff out, ending up with just the BIOS, a streamlined Linux kernel, and the Chrome web browser.
The end result is that Chrome OS boots in 7 seconds, with every application on the operating system being a web application; there are currently no plans for native applications. You can pretty much mimic the interface of Google’s new operating system by maximising your Chrome window, but there are also a few new things in Chrome OS, UI-wise. For instance, there are panels, which are lightweight windows which pop-up from the bottom of the screen.
Under the hood, Chrome OS does a few interesting things. The entire root file system partition is read-only, and the user file system is encrypted. You shouldn’t see the user file system as a place to store files – it’s actually a cache. All user data is stored in the cloud on the internet, and continuously synced. As you probably already anticipated, the operating system updates itself automatically.
Another interesting concept is that of the verified boot. The boot process verifies the bytes on your drive using cryptographic keys, and as soon as a mismatch is detected (because of malware? Data corruption?), the system will reboot itself and re-download the operating system image. The operating system basically re-images itself.
Security is a focal point for Chrome OS. It uses the browser security model, meaning that the operating system does not trust any application. All web applications run in a sandbox, and can’t touch local files. This is fundamentally different from traditional operating systems, where every application has the same privileges as the user – in other words, they can potentially destroy your data.
Google is taking a rather different approach to distributing Chrome OS. You won’t be able to just download it and install it on any machine – Google will be very strict about what hardware Chrome OS supports, and provide reference hardware to OEMs. For instance, Chrome OS does not support mechanical hard drive; only solid state drivers are supported. As for other types of hardware, Google will for instance specify what wireless chips OEMs must use. Chrome OS will, however, run in a virtual machine.
The first Chrome OS devices are expected to ship late next year.
Different
Google calls Chrome OS a “fundamentally different model of computing”, and they certainly have a point there. However, I’m personally not very excited about all this. Call me a traditionalist, but I dislike web applications, and the idea of being limited to them doesn’t appeal to me at all. Google did state that web applications that support HTML5’s off-line capabilities will work without an internet connection too, but I have the sneaking suspicion that most of them don’t, and the prospect of having a dead machine every time the internet decides to call it a day isn’t a pretty one.
Then there’s the interface of Chrome OS. It’s all a work-in-progress at the moment, so things might still change between now and when Chrome OS ends up on the shelf. The current interface is very limiting, as the tab bar runs out of space after about 5-6 tabs open. Since applications as well as regular web pages take up residence there, it looks like a very inefficient and space-constrained environment for managing running applications and web pages. The panels, which can be tiny applications like a music player or a notepad, also look constrained and restrictive.
And every application full-screen, with no side-by-side abilities? I shiver at the thought.
Still, there’s a lot of innovative stuff here as well. I still don’t understand why modern operating systems do not stick every application in a sandbox, so Chrome OS has them beat there. The verified boot and read-only root file system are interesting too, as is the idea of encrypting all user files.
Overall though, Chrome OS looks like a very limiting environment to work in, and much will depend on the agility with which web developers will dive into HTML5 and associated technologies in order to create compelling and useful applications. Google also has a lot of work ahead to make the interface more appealing. I love Chrome’s UI as a browser, but I wouldn’t want to spend my entire computing time in it.
Google will also have to convince users that storing all their stuff on servers somewhere far away is a good thing, and I’m not sure people will accept that so readily – I certainly don’t.
Then again, I’m scared of anything new, so maybe I’m just too old-fashioned, and Chrome OS is for people who say “cloud” instead of internet. We’ll see.
In any case, the code is available starting today.
All looks very interesting. I’d love to give it a try… Thing is though, the only option for trying it out is to download and install Ubuntu, then download the Chromium OS source code, build that, and then shove it onto a USB stick.
Why can’t they just make a VM image available, like almost everyone else does?
Yea, it wasn’t so hard to create a VMWare image. I’m trying to convert it to VirtualBox now. Anyways, I don’t have the bandwidth to host it, 777MB, but if anyone wants it I can give it to someone else to host.
EDIT:
BTW, Apparently I missed the news, but VirtualBox just opens the vmdk… All that work to switch from vmdk to vdi, and it’s totally not important. So, yea, someone with magical hosting space, contact me.
Edited 2009-11-20 07:28 UTC
Yeah, didn’t look that hard to build an image…
Thing is though my primary OS is Mac OS X. I’d have had to download an Ubuntu image to do the build, and work out how to work around not having a USB key. Too much hassle for me. 🙂
Seems tho that some kind person has made a VMWare image and one can find it on TPB now. Handy.
I’ll happily seed a torrent
Unfortunately, the way it is tailored, it’s either a VMWare image or a VirtualBox image. I at least have it running well under VirtualBox, but it would most definitely break if someone boots it under VMWare. So, to keep confusion down, maybe I should just forget about it?
I think this is a wise strategic decision for Google
Check out the videos here:
http://www.chromium.org/chromium-os
While this is definitely not for me, it would be a great OS for my parents and something I can give them with certainty that it will always work.
Looks and sounds interesting and I am sure Google will contribute some interesting changes upstream.
It’s Chromium OS, not Google Chrome OS ~:D
Uh… no. Chromium OS is the open source project upon which Chrome OS is based. Kind of like Chromium Browser is the core for Chrome Browser.
Uh no … if you go to their site, everything is listed as Chromium OS. And then they take the source, and build Google Chrome OS out of that.
Just like Chromium vs Google Chrome for the browsers.
How is that not what I just said?
Oops, it is what you said. But it’s certainly not what I read, somehow. My bad.
I could use it as a second OS running on a hypervisor alongside Windows with disabled Internet. Or browse with it on a dirt-cheap ARM tablet. It would make a great OS for grandma.
But I don’t see it being used as a primary OS by anyone with some computer literacy.
You hit the nail on the head. Nobody is going to get one of these as the primary computer (expect the scenarios where you’d like to buy a zero-support computer for your parents).
It’s intended for secondary/tertiary “email/facebook” devices for living room, something you turn on in a whim and quickly catch up with whatever is happening.
It will also work fine for most common “entry level” offline needs (media browsing, ebook reading…).
If you mean no tech-inclined person, then you *might* be true (or not, let’s wait and see). But most people’s current use of a computer is for web browsing, chatting, emailing, media playing, some word processing… That’s already covered by Google’s services and other online services in the web. With audio/video plugins, you can easily have access to online/local multimedia with a browser frontend. I imagine rtorrent/transmission-cli/btd with their respective web based frontends for illegal downloading. Big trouble for Microsoft if this catches on (it’s Google after all).
Most people I know also use an mp3 player or digital camera.
It won’t, and even if it did MS could just counter with Windows Mobile + Zune store + local access to mp3 files which would be far more appealing to the typical consumer.
ChromeOS is too niche. People want to do more with netbooks than surf the web and open google docs. I don’t even know anyone that uses google docs.
A moblin netbook with alternative apps is a hard enough sale. But a Linux netbook that only comes with a browser? They won’t be able to get them cheap enough.
A background daemon that transfers images from your camera to picasaweb should solve that. You think that because it’s a web interface it can’t interact with other stuff than web pages?
You already know? How?
Typical consumer knows nothing about anything tech related.
Of course, it hasn’t been released yet, it’s not complete, who knows what it will be able to do?
I thought netbooks were primarily for net stuff.
I don’t either. But I find the lack of users has more to do with Google docs being too slow. Things are changing fast in the web browser market.
A netbook with no marketing at all. Tell me, who manages most ads on the web today?
I would expect the browser to be the GUI, like if you pick KDE it’s your GUI, with window manager, file manager and integrated apps. You certainly can do a lot with a browser (and more with google engine) provided you have a backend. Backends can be provided by a web service or a local service. You can fit both in a netbook. I’m not saying this is how it’s going to be, but it could be done for sure.
Just be realistic, lots of entreprise software are being redesigned as web services with web frontends, because it’s easy to deploy and manage from a central location without caring too much about the clients. Take random office worker: he would read mail, answer the phone, do some spreadsheet and word processing, and then use a job related software, like invoicing, HR software, etc. With a browser as a client, you as IT are freed from lots of expenses, IT contracts for all your clients, antivirus, etc. In the end, it makes sense, lot of sense.
No the point is that devices require drivers which is something Google hasn’t talked about. But yea uploading gigabytes of media to Google servers sounds like fun. So does having to download an mp3 every time you want to sync it.
So this is going to be sold on consumer ignorance?
With the Zune store people can buy music and rent movies. Consumers can understand that functionality. Why would consumers choose a device that only has a browser over a device that offers more functionality? I could see WinMobile devices easily outselling Chrome devices by simply touting their gigabytes of local storage. I’m really gonna laugh if Asian market OEMs end up going with Moblin instead of Chrome.
It isn’t being targeted at the enterprise, but more importantly businesses don’t need a new OS to run web apps. Keeping Windows means keeping printer and scanner compatibility as well as any native software that is needed or *might be* needed. Switching to ChromeOS is too much of a risky lock-in for the typical business.
Chrome OS devices will have gigabytes of local storage in the form of solid-state drives according to their announcement.
It’s intended for secondary/tertiary “email/facebook” devices for living room, something you turn on in a whim and quickly catch up with whatever is happening.
The Deities help us. Ultra quick Facebook appliances. So you can stop having a real social life even quicker(TM).
I haven’t seen a few people for over 6 months in the flesh, because they seem to believe that publishing every uninteresting little detail about their everyday lives and commenting on these posts from other people just like it, is adequate replacement for real life contact.
I wonder how “I’m sitting in the garden. Reading a book” [Random friend X likes it] –> “Enjoy your book.” –> “Nice having some leisure time. Wish I was there.” –> “Is the book any good?” –> “Barnes and Noble are having a sale.” is enough for some people.
There is no longer any exchange of ideas, nor talk about feelings or the deeper stuff of life. I guess I’m a dinosaur and I feel threatened by the birth of Homo Superficialis.
The only upside to this is that Google will probably get some serious revenue through shoehorning all those online time wastes into their adsense channel.
When your closest friends and family members have all moved away to different cities, it’s nice to be able to stay in touch at that level. It brings back a bit of “normalcy” since it makes it seem like they are still close by. It’s also easier than being on the phone with them for 8 hours a day.
Granted, I’ve never understood people who live in the same city, even in the same neighbourhood, doing the same. If you’re that close by, just pop on over for a visit.
and then I’ll later post the pictures on facebook.
Thanks for the heads up on the B&N sale.
If they’d include a terminal emulator, I could likely do most of my work on a “Chrome OS” device.
Everyone of us will give it a try but let me say a few things based on what I saw in those presentations:
1) stating that a computer is mostly an interface to the Internet is VERY reductive. Stating that most of us will only use their computers to access the Internet is reductive as well. Those concepts already failed in the past.
2) it’s amazing how we fought for YEARS IE and the way it was so tightly integrated with underline OS. We complained it was insecure, prone to errors, causing instability AND a problem under anti-trust point of view? No-one noticed that Chrome Browser is doing what IE 5.5/6 was able to do (directly or via ActiveX plugins) more than 10 years ago? The concept that EVERYTHING could live inside a browser and people could use that framework to develop extensions to their desktop (besides regular full-fledged applications) was part of the… Windows98 development process! Then we spent 11 years trying to remove IE from Windows and now… we build another browser-centric OS?? Life is irony for sure …
3) According to what I read, Google has made lots of compromises in hardware, just the way Apple did. Basically, they will only support new systems and, probably, specifically crafted PCs. This inherently means they don’t think they could aim to a large user base as they are kicking existing systems out as a start. Unfortunately, a PC is not a phone. And while a phone could be small PC, there’s a lot of assumptions to do when you behave like this. If any of those is wrong, you could fail. After all, Google is no Apple and hasn’t a loyal (yet small) user base.
Beside those considerations, I welcome this new system and I’m happy Google finally release its OS. Now they can prove what they can actually do 😉
Welcome guys !
Google does not hold majority share of the OS market so antitrust claims that the domination is being used to push Chromium on users is not equal to the past example. If Google had majority share of the OS market and was using that OS market share as a delivery method to push there browser then it would be a market concern. Where one company pushed a browser through a general use platform at the intended expense of other browser options, this is one company pushing an embedded browser for much less than general computing needs.
In terms of security, if chromium shows the poor decision making that most versions of IE have shown and Google can’t/won’t keep up with security patches then the deep integration of the browser may become an issue. The underlying OS layers seem pretty static outside of downloaded clean images/updates. Ultimately, we’ll have to see what the research geniuses can do with it though.
Just because it failed in the past doesn’t mean it will fail again in the future.
Looking at my wife’s computer usage, this would be almost perfect. The only apps she uses are Mozilla Firefox, WordPerfect, and Picassa. 95% of all time spent on the computer is spent in Firefox. The rest is typing up the odd document, and pulling images off the digicam to post online.
There are a lot more of these kinds of people out there then you seem to think there are.
Will this be perfect or even usable by everyone? Doubtful. But for a large portion of the computing public, it just might be.
The big difference is that this is not a general-purpose OS with a crappy browser pre-installed and “integrated” in such a way that a naughty website can control the host OS.
This is a special-purpose device. You don’t have access to “the OS”. You have access to a browser. Very big difference.
Again, this is not a general-purpose computer, running a general-purpose OS, that can be used to do anything. Why aren’t you complaining that this vehemently about the Kindle, since it’s really nothing more than Adobe Reader tightly integrated into the OS, only running on specific hardware?
Seriously, take a step back, and look at this from the viewpoint of a special-purpose device running a customised OS. Nothing too different from an iPhone or iPod Touch, except that instead of having native apps, it’s all web apps.
To all the critics of TBPrince, I think you are too quick to declare Google clean.
TBPrince has a point, it is just looked at from the wrong angle.
This is not about a General Purpose OS provider, trying to pull in all web traffic through their dominant OS by bundling a browser (inferior or not), in a bid to keep their OS relevant. (Although I can appreciate the astuteness of TBPrince’s line of thought that it is very ironic that people are cheering for a browser with integrated OS, while they vilified an OS with an integrated browser).
This is about a GIANT search and web applications provider, whose stated goal is to be the central hub in the organization of the worlds information, trying to push out new devices according to their own specifications which conveniently precludes doing any significant offline and local computing.
While right now Google’s Chrome OS is merely a whiff of smoke trying to materialize, I can see problems down the road. The worlds computing now is not done through Chrome OS, but what if we just keep on cheering uncritically and someday 80% of computing is done through Google’s devices? Let’s see, Google Chrome OS, Google search, Google Scholar, Google Mail, Google Docs, Google Blogger, Google YouTube / Video, Google Picasaweb, Google Maps, Google News, Google Groups, Google Books, Google Calender, Google Reader, Google Translate, Google Sites, Google Talk.
See anything disturbing? Google spans the pallette of what could be needed for computing. Right now Google has to go through other platforms like Windows + IE, Mac OS X + Safari, Linux/BSD + smorgasbord of browsers, to deliver their Cloud Applications. If Chrome OS devices become dominant, they don’t have to. They can even start augmenting the web in Google specific ways and not make the stuff open like SPDY. Although they could open it up anyways with impunity if they store all you digital life on their servers anyways.
I’m not saying “Google is Evil”, I’m just saying that concentrating too much power in one place is inevitably going to have negative consequences somewhere.
You know what is COMPLETELY different than what Microsoft tried to do? This project is opensource! I don’t think anyone appreciates how big a deal that little detail is.
I believe google is (cautiously?) gambling on the success of a new device, just a few notches above a phone.
Ain’t really what I spected.
A browser on top of a nearly naked kernel on top of a minimal bios. All software and non-volatile storage being provided by servers. It’s about what the rumors sounded like versus the wild claims that it would be the second coming of Linux backed by Google.
Now, let’s hope a VMware appliance becomes available, an ISO or something cleaner than a source build and storage juggle to get it into a VM.
To the extent that cloud-only OSes matter, this is clearly Linux’s presence there… to the extent that cloud-only OSes matter.
It’s just too bad that this couldn’t have been microkernel-based. Then, instead of one beautiful Ivory Tower stretching up into the clouds, we could have had two beautiful Ivory Towers stretching up into the clouds. Twin Towers, one might say…
Edited 2009-11-19 21:19 UTC
A world where all computers are managed as embedded os terminals for centralized services is a rather stagnant sounding place. I could see having a tablet at home that can be passed around with the home groupware interface loaded. For daily computing through, I won’t be giving up my traditional machine.
So true, as far as centralized computing terminals go, this has it’s place. Like the iPhone, it looks to deliver it’s intended functions well enough. Now to see what unintended functions it can delivery.
Just search for “chromium os virtual image”, you’ll find some, I’m downloading one right now.
Pretty much exactly what I expected… except they seem to be targeting a wider audience than would strictly make sense in the video presentation
There is nothing wrong with making educated guesses on stuff like this Thom, but it is somewhat dishonest to guess wrong and than claim to have guessed correctly. Yesterday Kroc posted it would basically be a web browser and that developers would already be prepared for it. You replied:
That was yesterday. Today you post this:
So which is it?
For me, it was _exactly_ what I expected and imagined it would be. Down to the small details.
*Finally*, some damn innovation in the user space.
Ditch the multiple splash screens, the multi-stage logins, the incompetent BIOS manufacturers (Google have supplied a custom firmware)
Version numbers are gone. Security is re engineered. Safe, silent updates and no more irrelevant nagging.
This is a bold, unflinching statement of the future, and a breath of fresh air. Microsoft and the OEMs could never have done this in a million years.
Come on?! You call this innovation? We had this > 10 years ago, it’s a bloody thin client, except instead of the X protocol it uses http. That’s pretty much it.
How does this ditch the multiple logins? You still have to lock into every webservice you are going to use. I agree that getting rid of incompetent BIOS manufactures is a good thing. But multiple splash screens?
Security is not re-engineered, safe silent updates? You can do that already with most package managers just turn automatic updates on, it’s just not always a good thing, a lot of people actually like to have control over what gets updated.
If this is the future of computing I’d rather not be part of it. I’d like to have control over my information, I’d like to decide which applications I install (and I like to use native applications, they are and always will run a lot more responsive on my computer). This isn’t bold, it’s what was to be expected from google, and I’m completely underwhelmed.
Sorry, but that is a mistake and its a big one. There is a critical difference between this and the thin clients of yesteryear. Customers had to buy and maintain the servers that traditional thin clients were connected to – and upgrade/replace them when you they were outgrown. And outside of business environments thin clients served no purpose because your average Joe didn’t have a server to connect it to, and even if he did what would be the point of having a thin client for only one user? Not having to deal with a server is virtually the entire point of modern cloud computing (or SaaS or whatever buzz word is popular for it today).
Pick a provider for managing your credentials. If your a non business user this can be Google, otherwise its probably a internal directory server. Expose the directory through SAML
http://en.wikipedia.org/wiki/Security_Assertion_Markup_Language
Google already does this part for you, but if you run your own DS youll have to handle that part). Run cloud apps that support SAML (most do – by next year pretty much ALL will). BAM! instant Single Sign On for everything.
You just don’t get it. This is for people who DON’T like to have control of what gets updated… Its for people who don’t CARE what gets updated, as long as it works and works well and they can get along with their life without having to worry about it. You know, NORMAL people??? Its not for geeks, its for suits and grandmas and high school girls – there are A LOT more of them than of us you know.
And I am completely unsurprised. This is exactly the reaction I expect from most competent and enthusiastic computer users. And it simply doesn’t matter because this stuff is not for you. Its for everyone else. And you don’t have to like it but eventually you’ll have to at least tolerate it because your Boss will make you
Sorry, but that is a mistake and its a big one. There is a critical difference between this and the thin clients of yesteryear. Customers had to buy and maintain the servers that traditional thin clients were connected to – and upgrade/replace them when you they were outgrown. … … Not having to deal with a server is virtually the entire point of modern cloud computing (or SaaS or whatever buzz word is popular for it today).
Ooh, how innovative. Applying outsourcing to client/server computing.
And I am completely unsurprised. This is exactly the reaction I expect from most competent and enthusiastic computer users. And it simply doesn’t matter because this stuff is not for you. Its for everyone else. And you don’t have to like it but eventually you’ll have to at least tolerate it because your Boss will make you
It will come back to bite them all in the butt. Leaving all your information to third parties to be strip mined is a recipe for disaster.
On the topic of bosses. They can make you use this harebrained idea at work, but they don’t have control of the private systems of their employees.
Quite frankly, what do I care if my boss deems it a good idea that he should be screwed over by handing the keys to his kingdom to a third party? The skills I have now, can be applied at any employer…
*Finally*, some damn innovation in the user space.
Ahem, sorry… Client/Server computing is old. It is a return to the Mainframe model. Cloud Computing is a centrally controlled behemoth of computing power and the end user gets a dumb terminal (Chrome OS device).
I see no innovation.It’s technology spanning from the 60’s to the 90’s, with merely a fresh lick of paint. We even lose the control again we gained with fat client desktop computing.
Clue: everyone. The common thought was that it would be like what it is now. My personal expectation was that it would be more traditional.
Well you did say we, which traditionally includes the speaker, but I’m not going to harp on it. I’m actually genuinely curious as to what you think of it now that we do know what it is. I think a lot of people were expecting it to be a bit more than “just a browser” for no other reason than they couldn’t see the point of it all otherwise.
I think Thom is referring to the royal “we”.
I believe Thom was spot on. What is a traditional OS? A kernel, booting then presenting a user interface and presented with user tools ie: the Browser.
Storing everything on the Internet? I mean how can I trust Google, or anyone to keep my files safe. I know the web generation doesn’t care about this, but I certainly do, and I least of all don’t trust a corporation to keep my data private.
One thing that gets me is that most people have very slow upload, how could they benefit from something like this since it would take forever to upload anything. Its obvious what Google wants to do.. They want to charge you for storing all your crap on their servers either through ads or a monthly fee, either way it seems stupid to buy into a service that would be like that. The only way I would be willing to do this is if I could configure it to sync with my servers and not Googles.
Gee. You use a PC (or Mac) and think your data is private? who cares about your personal finance or video stash anyway. i much prefer handing my “so important” data to the safe hands of Google without thinking.
How about instead of releasing your data, they screw up and delete it or their servers crash like with the side kick. I would rather be responsible for my own data, and not have to worry about Google losing it.
that’s nice. Most people though have problems doing regular backups and when their laptop hard drive fails they are up crap creek. Google is far less likely to lose your data than your laptop hard drive is.
I get irritated with how slow hard drives are when they’re internal with SATA. Using external drives over USB is that much worse. Over the internet would be far worse than that. Yes, it would be nice not to have to worry about backups, and it would be nice to be able to get your data wherever you were, but the speeds would never be fast enough to be acceptable for me. Office documents might be okay, but multimedia stuff such as music and video? Forget it. The files are way to big.
If maintaining your data online didn’t cost any more than a local hard drive, and the speed was at least as good as a local hard drive, then I might consider it. Since that obviously isn’t happening (from a speed standpoint anyway, I have no idea about the cost), there’s no way that I’d even consider it.
What I want to know is what the use case for this is. Is this intended for netbooks? That’s about the only thing that I can think of. It’s innovative, I suppose, but I don’t really see the point (of course, I don’t really like the idea of netbooks either).
Personally, I hate the trend towards web-based applications. I _much_ prefer native applications. I certainly won’t claim that the current desktop model doesn’t need some work – innovation is still good – but overall, I really like the current desktop model, and I have yet to see any innovative stuff which tries anything different which I would consider better, or even vaguely close to it.
In any case, this is certainly interesting, but from where I sit, it seems rather pointless. Google obviously sees some value in it though, so maybe something interesting will come of it.
[q]What I want to know is what the use case for this is. Is this intended for netbooks? That’s about the only thing that I can think of. It’s innovative, I suppose, but I don’t really see the point (of course, I don’t really like the idea of netbooks either).[/quote]
On something like the 10″ netbooks, with a built-in cell chip, wifi, and ethernet, this could be useful. Without an always-on Internet connection of some kind, though, it would be pretty useless.
The use case is to wrestle Microsoft’s slaves away from Microsoft, and make them Google’s slaves.
See my take on the misleading term “netbook” for my personal feeling about that:
http://www.osnews.com/thread?395467
This association of small laptops with the cloud, through the misnomer of “Netbook”, is getting to be one of the most malignant mal-memes of our time. It’s getting to be almost as malignant as the “I” word.
Small laptops != Cloud Computing
Small, inexpensive laptops, with more computing power and storage space than the desktops of yesteryear give one the ability to take a desktop computer with them, with the *option*, not the mandate, of using the web as a balanced part of their mobile computing experience.
Edited 2009-11-19 22:06 UTC
Add a user-space web server, a user-space database engine, and voilà , you can have local applications and local data. I would also like a user-space IMAP server for local mail storage.
Maybe some terminal applications would be useful, too. The source is available, so hacking this in should be possible.
I really like the custom BIOS, the verified bootup, the encrypted user files, the speed 🙂
A comparison with the coupling of Windows and IE fails because IE was also tied with local applications (Word, for example) and – very important – IE used it’s own lock-in standards. Your Google apps will also work with Firefox, Opera, you name it.
This is the future, and the open source community could have done it technically – but not marketing wise. This will be a blow to Microsoft.
However, I agree with Thom that I cannot live without my local apps.
Edit: It seems like they use X.org !!! That will make modifications with additional local apps possible.
[chromiumos.git] / src / third_party / xserver-xorg-core / README.chromium
1 URL: http://packages.ubuntu.com/karmic/x11/xserver-xorg-core
2 Version: 2:1.6.1.901-2ubuntu2
3 License: Xorg MIT-like license
4 License File: LICENSE
5
6 Description:
7
8 The core X Server
9
10 Local Modifications:
11
12 Modifications are in the form of patches. See the README files for each patch.
Edited 2009-11-19 21:11 UTC
Gears, Google Gears… http://gears.google.com/
One only must wonder why Chrome (the browser) isn’t supported. Maybe the gears for Chrome are OS-only thing?
Chromium ships with Gears out of the box.
Local sql database is an html5 thing (possibly even more revolutionary than video tag?).
I may be wrong, but I believe the reason is that Gears is already built into Chrome – it doesn’t require a separate download.
This is the beginning of the end of the MS dominated era.
It is not made for a small minority of OS jugglers like us. It is targeted at the world majority of would-be users who are (mentally and financially) challenged by the present majority OS.
On the contrary, I think this is targeted at people with money to burn (that own several computers, and have no time to admin them all).
It is mostly what I expected. Moblin + Chrome
So no new windowing system. Plain old X11 is the windowing system (pimpled with OpenGL, Clutter etc.)
Things I like: Firmware, Security model, Open Drivers (I really hope they put pressure on the hardware vendors, that would give them A LOT of creds with the FOSS crowd.)
But I think they have to provide Native Client apps really fast .. people want native speed.
And btw. Litl is dead. ChromeOS seems to do the same (but more thoroughly executed) and will be “Less than free”(like Android)
No new windowing system… yet.
You might be right. I don’t _know_ anymore that you do, but I really don’t think that is going to happen. I think Google is going to intentionally a permanently avoid a native API for the following reason:
As long as the only thing Chrome OS runs is web apps, they can change virtually anything in it as they see fit – and as long as they can get Chrome (the browser) to run on it they maintain their ABI (essentially the browser itself IS their ABI). Without having to worry about maintaining binary compatibility with anything, they can effectively optimize the living sh*t out of it over time so that they can reduce their hardware requirements over time (or increase performance on existing hardware depending on how you look at it).
I think the whole point is to make the hardware itself as uninteresting as possible – Chrome OS is for running cloud apps, and running them as cheaply as possible.
So sure, it is little more than moblin + chrome NOW, but as long as they avoid getting trapped by having to support native apps, they can work on shrinking their OS stack. I predict X will be the first thing to go…
Having Native Client does not imply that they would commit to native app support. It just means you can run native code in the browser sandbox, for performance. Binary browser plugins as such are not sandboxed right now.
Maybe. They may support plugins. I admit browser plugins often able to fill the gaps in the HTML/CSS/Javascript stack, but from Google’s track record I think they will always choose to try and shore up the stack first. It just makes more sense in the long run. But I seriously doubt they will implement anything outside of plugins.
Open Drivers? I am actually somewhat worried about the OEM validation policies (although I guess they can be hacked out). I will put pressure on selected hardware vendors, but I’d call it partnerships.
Then, no hard-drives is another funny one… The vision if for light(weight) machines, that store as little as possible locally. I guess this can also be easily hacked around. But will the firmware be modifiable as to allow for a different kernel, with all the checksumming?
Finally comes into being what (I think) was Netscape’s biggest missed chance and the reason Microsoft crushed them.
Why have an OS when you have a browser?
Because the Internet wasn’t ready for it, back when Netscape was around. These days, web applications like GMail, or Google Docs are almost as good as their desktop counterparts – certainly, they’re good enough.
Also, the internet wasn’t so ubiquitous back then, and the target user base for a web OS – people who only care about web and email – didn’t really exist to the degree it is now. The people using Netscape then were the people now loudly proclaiming that web apps don’t do everything they need, or that they don’t like the idea of entrusting all their data to some remote service.
Yeah, I remember Bill Gates writing about it in his book “The road ahead” way back in 1995(?). He wrote that one of his fears was Netscape would create a browser based OS and negate the need for Windows for many.
Edited 2009-11-20 11:10 UTC
I wonder what Google’s solution will be for music libraries. Evidently, anyone with more than a middling music collection will need some local storage.
Also, I assume there will be copies of needed documents, etc. kept in cache for when working offline.
Oh, yeah. They were careful to point out that you could use it as a sort of oversized iPod.
Cached documents? For use by… offline web-apps? Nope. Your data lives at someone else’s datacenter. And what good would having the data do you without the applications… which also live at someone else’s datacenter?
You need to do a little research, methinks. It’s possible today to support offline web apps, and local storage of data. Funny enough, Google have done much of the work around that area… almost as if they thought it might be useful for a browser-based OS…
Did you watch the webcast?
There’s already Google Gears… I assume its functionality will increase before Chrome OS makes its official debut.
I wonder if the RIAA would accuse you of unlawful distribution via upload when your music collection syncs with the Google servers, lol.
In view of the judge’s summary judgment in the Psystar case, it wouldn’t surprise me if you bought a CD, and got sued automatically as it got copied into memory when you played it the first time. And again, after the second time, ad infinitum.
I hoped that they do something new. But thats something i can not use. In my country(germany) you don’t have fulltime web access. Like it was a hour ago, when the net went down.
What can I do with the OS when there is no internet?
And it does not look so new. It is a normal linux kernel with an chrome running.(chrome for linux is not released yet and the daily builds are not fully working)
I have data on my computer that is not for everyone( most of it is source code). How is it possible to combine Chrome OS and that?
Another thing for me is that I use a lot software paralell with the webbrowser and the web versions of those apps are not the bests.
So no Chrome OS for me, but I think I will play with it in a vm when it is ready.
Then it is not a good option for you.
I guess this will be like a small box with various usb entries, with a monitor or that can be hooked to a monitor or a tv set. Maybe it will also allow to make mobile calls.
Considering the marketing power Google has, and the utility of this set I’m not in doubt that it will be a tremendous success.
I’m sure microsoft will reply to this with some similar gadget.
I still can’t get into netbooks and they have gone and gone three steps lower. I just cannot imagine ever wanting to use this. Maybe I am just too paranoid to have anything store online.
At least with a netbook, you can dump pictures on it when you are out shooting pictures. Or have a bunch of music when you are in a net deadzone. Hell, it’ll be a paperweight in rural America.
Do you travel much? Because they’re absolutely brilliant for tourists – compared to a regular laptop, they don’t take up much space in a pack, nor add much weight. And yet they can still do almost anything a regular laptop can. I recently carried mine around South America for a couple of months, and it was perfect for email, for web browsing, for transferring photos off my camera, editing them, and posting them online.
Now, certainly they have downsides – a relatively small screen, and slow processor. But the processor is good enough for most purposes (and downright speedy compared to my parent’s old laptop), and I consider the screen size a reasonable compromise to get something so practically portable…
No, don’t really travel much. I can see the benefit of a netbook, I just can’t personally get into one.
Your example sounds like an itch that needed scratched. With this new OS, I just don’t see the itch.
I think that this OS has four things going against it, and that’s Windows XP, Android, the iPhone, and their own OEM policy.
Love it or hate it, Windows XP is a great OS for netbooks because it can fit on a small hard drive and runs almost all of the apps that people use on their desktop, plus there is a familiarity with the OS and its applications.
Additionally, all of the good remote connectivity options are there (Wi-Fi, 3G cards, 4G cards) already as almost every vendor releases Windows drivers and control applications for them.
Android and the iPhone, on the other hand, allow you to install your own apps locally. This OS apparently keeps everything in the cloud, and uses web apps for everything. Plus, these devices already have 3G connectivity. Android Market and the App Store also have thousands of apps.
Remember how well Steve Jobs’ telling developers to write their own web apps instead of native apps went over when the iPhone first came out? It’s one of the main reasons the Jailbreak community came to life in the first place. The same thing will happen here. It’s a matter of when, not if.
Their own OEM validation policies, while good in theory, are going to have the effect of keeping people off of the platform. This isn’t due to Google, it’s due to the fact that there isn’t a review process to install Android, Windows, Windows Mobile, or Linux on another portable device.
In short, I think that Google should have just tweaked Android for netbooks and leveraged what they already have.
Google has been doing a lot to get Web Apps behaving like normal apps.
Look at what they did to GWT to get Wave nice and shiny.
Its still no replacement for native code though.
The value of your market just got a 1000% boost and I wish you all the venture capital you deserve. People bitching about native apps have likely no clue about modern web development and what era we just entered.
These guys are off to a great head start:
http://www.tonido.com/
http://www.amahi.org/
In three years time your parents will have gotten themselves a $100 Linux home server plug at the local super market to go with the variety of web clients they have kicking around the house. And will you care about those gadgets just as much as you care about their VCR tape collection today.
Perhaps four years 😉
You’re dreamin’and smokin’ some seriously bud.
Lol. Yeah well, I welcome the potential of this platform and hope that it changes the playing field a bit. We shouldn’t have to wait for Zorry Zebra to experience some truly different computing paradigms on our devices and computers. The tricoders of tomorrow aint running no GNU/Hurd 🙂
Edit: It should read “And you will care […]”
People bitching about native apps have likely no clue about modern web development and what era we just entered.
The vast majority would rather pay a premium for Windows than have an OS that only provides a browser. I think you have no clue as to how many people are tied to at least one native app.
I really think Moblin is a better direction.
Oh yes, I think they do. And the number of people who earn their living developing native apps that couldn’t be replaced by cloud apps or other web apps. There are much more uses of a computer than running twitter, you know.
That’s cool stuff, but somehow I don’t think that’s the kind of stuff Google is looking to target. See, they want you connecting to their web properties, not storing stuff locally in your own network…
… and power users around the world shed a tear
Why not just use G OS and its cloud version. It is more finished, polished and it works right now.
Google should have just bought them and continue their work.
Take a look at it:
http://thinkgos.com/gos/index.html
http://thinkgos.com/cloud/index.html
And in the other and (smartphones), one are doing the exact opposite; Using native apps to access specific sites on the internet, instead of the browser.
I like it. People afraid because it’s dependent on internet needs to move on. It’s not 1980 anymore and that’s the whole point. Sure if you have a dial up that’s not for you but more and more people have internet connection up 99% of the time.
What I am really hoping is that it could finally push computer makers to get rid of the BIOS antiquated code for all PC not just Chromium OS based ones.
How about being afraid because it is dependent upon Google, Inc. That’s a different thing.
Really? I travel by car a lot. And I certainly don’t have Internet access anywhere near 99% of the time with my mobile broadband. And when I’m on the highway, I’d prefer the reliability, relatively low latency, and speed of 56k dial up, to what I observe with mobile wireless.
Don’t tell me when I need to move on. I recognize the practical limitations of Google, Inc.’s “vision” immediately. If your Internet experience is stationary enough that you are comfortable with complete dependence upon the cloud, then perhaps it is you who need to “move on” in a somewhat more literal sense.
Edited 2009-11-19 22:52 UTC
But they demonstrated a Gmail on the iPhone that works without connection by caching and then automatically sending/receiving mail as it detects a connection. Why can’t this be done with a netbook too? How is this different than a local mail client that does the same?
I live near a large city and yet twice today I didn’t have internet access for my laptop.
At this point the iphone is far more useful to me as a mini computer since I can play games on it when not connected.
Sometimes you want to go to a cafe and just use your computer quickly without messing with their network.
I really think it is too early for a web-only OS. It’s too niche.
HTML 5 and Google Gears let you create web-apps that can work offline. You don’t need a network connection to use your GMail, for example. I think we’ll be seeing more and more of this in the next few years.
a letter while waiting? That sure beats having local games and movies like with the iphone.
Hey kids get yourself a copy of Google chrome, it allows you to write a letter to grandma when waiting for the bus.
I was believing in the revolution. Although this looks very interesting, I’m afraid it’s too much of a simplistic approach though. Now I don’t know if it’s going to really catch. One thing for sure: Finally someone beat the hell out of Microsoft and GNOME treating users like complete dumbies…
I am sure it’s going to catch among 90% of users, because you know… the majority wants it easier and easier.
Another vote for the saying “Power Users shed a tear”.
What? This thing is targeted at the Facebook-only crowd who’ll then settle on Google Docs because it’s all that is available. Who’s treating whom like complete dumbies? Who are the real dummies?
Microsoft and Gnome users are the progeny of Einstein, by comparison. Chromium OS is AOL all over again.
Edited 2009-11-19 23:16 UTC
Anyone wanting to try out a more “traditional” OS with the browser as the main application should try out Browserlinux (http://www.browserlinux.com/) based on Puppy.
It’s main apps are:
Firefox (with Flash and Java available)
Geany text editor
xterm
Utilities
Hi,
Yep, pretty much what everyone expected, but I still don’t really get it.
How does it differ from just running a browser on a ‘real’ OS – expect there’s a bunch (and a big bunch at that) of stuff you can no longer do?
Saves the price of a harddrive I guess, but you can run a lite Linux on flash drive’s already can’t you?
To me, it’s all start to feel like some kind of ‘lock in’ strategy.
Oh well, I never drank the ‘do no evil’ koolaid anyways!
When MS can also just provide a browser with Windows Mobile and give access to the same apps but also provide a local version of office as well.
I think most people would actually pay a premium for local storage. I know I would.
Local storage means nothing. If you get your device stolen, you are screwed. If it burns, you are screwed. With cloud storage and caching, you don’t have to worry about backups, synchronization or anything at all. What’s upsetting people here is the possibility of lock-in (and most of them are already locked-in by Microsoft, Apple or ideology).
Edited 2009-11-21 13:51 UTC
Except that with local storage I can make my own backups. If someone is too naive to do that then they deserve what they get, but I’d rather trust my own backups than a central server over which I have absolutely zero control. If something goes wrong on the other end, I’m still screwed but unlike with local storage and/or backups there’s not a damn thing I can do about it. I’m fscked and Google wouldn’t give a shite about it would they? Sure, they’d apologize, but a whole fracking lot of good that would do me after the fact.
And as for security, you trade one kind of security (knowing exactly where your files are) away for another less-tangible kind (supposedly better security against attackers). You’re trading away your freedom, and I’m not talking about the GPL or any other license here. Honestly, have we forgotten the Amazon Kindle incidents already? It’s just waiting to happen here, on a much broader scale. The RIAA doesn’t want you having that album? Gone. The MPAA doesn’t think you should be able to watch that foreign movie? Snip. I’ve heard of forgetting history, but we’re talking about a period of a few months when content was removed from Kindle customers without their cconsent. Anyone who is willing to jump on this, in a way I feel sorry for you not being able to tell what’s going to happen if this gets big. At the same time, I’ll be laughing my head off at all those who managed to help kill the Microsoft monopoly only to have ushered in the era of total Google domination. Somehow, I think that would be worse than anything we’ve seen in the tech field so far. Microsoft had control over your data indirectly via file formats, Google will have total control over what apps you can run and your data on top of that. I don’t even think Apple, control freaks though they sometimes are, dared dream of that.
otherwise you wouldn’t see laptops advertised with 320gb drives.
Telling people to just upload their image and mp3 collections over to Google servers is not going to go over well. Consumers won’t consider that to be a feature.
Getting a laptop stolen is only a problem for people who don’t keep a backup. I not only keep a backup of my files but a backup laptop.
Would You trust your life/identity on the cloud?
If I didn’t have such major privacy concerns about everything that comes out of Google then I would probably give this a go.
I hope they can allay my fears, though Google does not have a great track record in this regard.
Edited 2009-11-19 23:17 UTC
Seems pretty much like a browser based Linux. A bit more extrem though than most recent approaches (cloud *only*), but generally pretts much like e.g. LimeOS http://usa.mtcera.com/os/index.html
Well, there is one big diffrence thoug: While most browser based OSes yet failed more or less I am bedding my right arm and my both balls off that this one will be rather successful, because it is driven by Google.
Wow, an os that just runs a browser. How very exciting. You know, we had a name for this and still do: “thin client.” It’s not quite a traditional thin client, as it’s not simply a bridge to native apps running on a server, but the overall concept of having essentially a diskless local system that does nothing but interface to remote services is the same. Boring. Oh, but this one has Google behind it so they can constantly serve you ads and push their services. The funny thing is, though, that an iPod Touch has more capabilities than one of these ChromeOS-based machines. Sad, Google. Very sad, indeed.
And for those claiming this is a revolution… Well, let’s think. How many areas in the world have constant internet access versus those that don’t? This revolution will never happen until the infrastructure is there to support it, if it ever is, and further this revolution will be the death of your personal freedom regarding your data. Before we jump head first into this cloud computing buzz (hint, thin client again), perhaps we should stop and think just a bit about what will happen when someone else holds control over your data. I’m not just talking about privacy either, I’m talking about screw-ups (see Sidekick) and what happens if your important data is lost because of a mishap on Google’s part? That’d be a class action waiting to happen. At least on my own computers, if I lose data it’s my own damn fault for not keeping a backup. The thought of having no control over our own information, of what we want to keep and what we want to delete or what is lost because of a mishap, scares me a great deal (Amazon Kindle, anyone?).
I hate to go off on a rant, but here goes my Dennis Miller on this whole thing
Chrome OS (and other systems to come that are modeled after it) WILL be the next big thing – it is virtually inevitable. It wont happen immediately or all at once, but it will happen. Heres why:
1. The rate that cloud based applications are being released is accelerating dramatically – and many of them are becoming as good or better than their desktop counterparts as time goes by.
2. Browser capabilities (with HTML 5 and modern JS engines) have reached the critical point that allows elegant, functional, and fast user interfaces to cloud services. See gWhatever, salesforce.com, workday, etc, etc, etc. And there is no reason to expect this trend is going to slow down.
Why?
The fact is it is EASIER to build BETTER applications as services over the long haul. If for no other reason than this: the implementer has complete control of the entire stack from the hardware up. This is not true of any other development paradigm. If you control EVERYTHING, including the hardware itself, the OS, the database, etc., you can replace anything in the stack with something better/cheaper/more efficient as needed. You also optimize your expenses because you can determine extactly how much hardware you need, how much memory you will need, etc. All the variables are under your control. And everything is STILL under your control AFTER deployment to customers. The better you get at it, the smarter you become, the more adept your engineers get, all of this allows you to work toward effectively reducing your cost – and anything and everything is a knob you can tweak if you choose to, there are no limits to the scope of changes you can make.
That is a HUGE competitive advantage when competing with conventional software. Think about your typical enterprise application (i.e. Exchange, or take your pick). Want your data to be truly secure? Extra Customer Expense. Want data replication? Extra hardware expense. Want to support more users? Extra hardware AND software expense. And as everything grows it goes LESS reliable and slower because it simply isn’t built to truly be scalable… Of sure, it CAN scale, but as an install gets bigger is costs MORE per user to keep the same level of service, not less.
Are people REALLY paying attention to what Google actually does??? They build hardware and design software infrastructure for one and only one purpose – MASSIVE scale-out with MINIMAL expense. That is the key to their entire business. Sure, not all of their stuff starts off so hot from the user point of view, but it does start off as scalable and they can grow it with comparatively little expense…
This lets them do things people in the conventional software world CANNOT do – they can build applications that are truly appropriate for enterprise businesses AND individual consumers, because they are not deployed into a closed network on their customers hardware. Gmail is a perfect example. Gmail is single handedly retiring large enterprise installs of Exchange and other email systems because it is:
1. Cheap (really cheap compared to Exchange).
2. Capable – maybe not as good as Outlook in some areas, but it is actually better in others (search, tagging, etc.) and it is only getting better with time.
3. Requires no capital outlay.
4. IT staffing requirements are minimal.
5. No server-side hardware needed AT ALL.
6. Runs on anything with a decent browser.
7. The big one – storage limits so large they are virtually non-existent. And no charge for storage, its built in.
8. Optional long term archival of ALL email (up to 10 years)
9. ISO compliance already baked in.
10. I could go on… really…
And Google’s expense to run the whole thing is amortized by all the people using it for free. People often think it is the other way around – i.e. the paying customers are carrying the free users. That is not true, it is the other way around. All those free users represent potential converts to paying customers, because most of them actually work for someone. It is self advertisement, and it pays off in the long run.
They use it, they learn to like it, it almost always works. And then they go to work and their Exchange server is down… Alot… ??? Hello Boss, get a clue? You can see where this is going I hope.
Anyway, getting back to Chrome OS. The point isnt to replace ALL conventional desktops – there is no need to do that. But creating a platform with _really_ minimal hardware needs will allow the creation of _really_ inexpensive devices that are just as capable of working with cloud apps as a desktop is. Chrome OS isnt for your home computer, at least not yet, but for the time being it WILL be attractive for many businesses. Yes BUSINESS – that is who is going to gobble these things up at first.
I think the perfect hardware for running Chrome OS will be very small and light netbook with essentially the equivalent of the iPhone SoC in it. 128MB of ram should be enough for it considering how much crud has been cut out of the OS. And it wont need nearly as much flash storage (if any at all) – local storage will only be needed for caching… Id be willing to bet that the first Chrome OS device will almost be EXACTLY this…
Anyway, the point of Chrome OS is not to compete with other Operating Systems… The point is to create a hardware/software stack that can run cloud applications using as little hardware as possible. It is getting users comfortable with running cloud apps that is truly the point – Chrome OS is a mechanism for creating hardware as a loss leader for cloud computing.
And a way to stop wintel platform from stagnating the whole industry. There I can see good use.
They mention the 5s start as something exceptional, but really the OS is not usable after 5s you still need to connect to the network, on my home network that’s often >3s and this is not even taking into account loading the webapp you want to use and suddenly the start-up time is really not that much better than a normal OS. Also Karmic brought a 5s boot time when using a SSD and that’s a full OS, so the Chrome OS time is not impressive.
There is no reason at all to think that the 5 seconds doesn’t include network connection setup. Kinda the whole point of putting the browser on top of a custom kernel is to make it possible to do this kind of stuff earlier in the boot process – you can eliminate the 100s of dependencies Windows for example has to manage before it can initialize the network stack. Chrome OS won’t have most of these dependencies because, well…, the only thing it runs is a browser. Thats it.
Second, the 5 seconds will probably be on hardware that is 1/10th the speed of your home computer… Probably an ARM or Atom core with maybe 128-256MB of ram. In that light I think it is pretty good.
when everyone uses sleep/resume.
But more importantly I could care less about instant boot if it means booting into a desktop that requires internet access.
ChromeOS has to be one of the most overhyped operating systems of all time. The new graphics stack turned out to be a rumor.
It’s a Linux distro that locks you into the browser. Amazing!
You do realize sleep/resume could just as easily be implemented on Chrome OS as anything else… And it could be implemented so that if you switched computers you could resume your last session on a different machine. Try that with a conventional OS. People’s inability to fathom how far this paradigm can actually go amazes me…
Then stick with your existing desktop/whatever. It isn’t going anywhere anytime soon. But many people (including me) really don’t see the point of even using a computer without the internet.
Everyone says the next big thing is overhyped prior to it becoming the next big thing… We’ll see. And PLEASE dont refer to it as a linux distro – it is NOT a linux distro. The Linux kernel is an implementation detail, it could just as easily be run on top of BSD or something completely custom built. It has no application binary interface – it is an OS designed to run a browser stack and that is all it runs. I’m not trying to dis Linux by saying that either, Linux happens to be the best available option. But an OS that only runs a browser only needs to care about running the browser – the stuff under the hood is subject to change. As time goes by I predict the stuff under the hood WILL change – probably drastically. Google will hack it down to the bare minimum as they figure out which parts they can whittle away… X is a convenient starting point for a graphics stack – by not committing their ABI to it they can hack it off at will.
Perhaps we could have the first real use case for Wayland here?
Ive thought the same thing. I would not be at all surprised. I’m actually more surprised that it is not already running on Wayland or some other custom stack – I honestly thought they would have forgone X, but I see it as a pragmatic decision to get to market faster.
Edited 2009-11-20 09:25 UTC
Yes, there are definite benefits to thin-client computing or remote desktop connections like RDP/VNC/NX.
We implemented NX on all of our school servers. This allows staff and students to access the same desktop and applications from home, from a laptop, from a hotel room, from Beijing, etc as they access when sitting in front of a computer at school (we use a diskless setup for the computers in the schools, though, not thin-client).
It’s actually starting to give us a bad name, though, as staff and students move to other districts and realise just how limiting “local apps” can be. No longer can they access their school account from home. No longer can they login to any computer in the school to access their account. No longer can they use a laptop to access their account. Everything is tied into the one computer (unless their new school is fortunate enough to have a network share for files).
That’s one of the nice things about web apps: ubiquitous access. You can access your webmail account from any computer, anywhere in the world, so long as it has an Internet connection. You can access your web photo album from anywhere. You can access your web documents from anywhere. You can access your music from anywhere. And so on.
As long as the web servers support standard protocols for accessing your data (and this is why FirstClass is doomed to failure in the web world), does it really matter where it is located? If you are paranoid, then just encrypt everything you store on remote servers.
Do web apps appeal to geeks? To some, they do. To others they don’t. Do they appeal to non-geeks? To some, they do, to others they don’t. Just like everything in life, to each their own.
The point was that boot times are not as relevent as they used to be when people rarely cold boot with laptops.
I’m amazed by people that don’t realize this is just Google trying to push a Linux distro that is just a browser built around their web apps.
Because most applications don’t require internet access.
But it isn’t BSD, it’s the Linux kernel and until they fork it I will refer to it as a Linux distro. Why shouldn’t I? Because they have a custom interface?
It has an abi but only Google is allowed to use it. It’s a Linux distro that locks you into a custom browser. Sorry but this is not a revolutionary OS.
Sleep/resume would be pointless on ChromeOS. Firstly you would disable the builtin security. Secondly, there’s absolutely no point in restoring the data which was in memory. All data is in the cloud already!. Finally I doubt that it would actually speed up the boot process.
Yes thinclients can have interesting applications, ChromeOS is a crippled thinclient, which can only connect to webservices.
But the thing is browser only linux distros have been around longer than ChromeOS. What Google is doing here is nothing exceptionally new, but somehow the same people who were previously saying “who would ever want something like that” are now falling all over themselves on Googles “paradigm-shift”.
Why? What does loading state data have to do with security?
Data != machine state. And just because sleep/resume on your typical OS involves saving memory and CPU state to disk, that isn’t the only way to implement it. The PURPOSE of sleep/resume is to allow saving your working session after a power off so you can resume your work later, not to increase boot speed.
Chrome OS would probably simply write out what tabs (apps) were launched and reopen them upon restart (admittedly it isn’t quite that simple, as there are some details that would have to be worked out so that apps could store their internal state if needed, but it is doable non the less). It could save this state data locally and/or in the cloud, so that it would be possible to resume the session on a different machine.
And no, this does nothing to speed up the booting of the machine, but that isn’t the purpose of the feature to begin with – people only see it as a way to speed up machine booting because cold booting is so slow…
You see that as crippling. Traditional (modern) thin clients are simply mechanisms that move all the processing to a server – the server is doing most of the same work the client would be doing otherwise. A particularly archaic “feature” of this is that the server is doing a boatload of RENDERING work, it has to run the damn graphics stack. Even with X, it has to run a substantial portion of the graphics stack, it may not do the actual rendering, but it is still attached to the graphics stack at the hip. I see THAT as crippling.
The beautiful thing about web services (if they are built correctly) is that they are completely ignorant of even the existence of a graphics stack…
But it is NOT a browser only Linux distro… It is a browser that essentially runs on a firmware. Linux just happens to be the firmware for the time being. The browser IS the OS, no one is seeing the big picture because they are stuck on the whole “Linux Distro” bit – Linux is just a cog in the wheel that the user cannot interact with AT ALL. I expect Google to rapidly whittle down the kernel to only the absolutely bare minimum required to run the browser as efficiently as possible.
Remember I said this: It will NEVER run a piece of native code compiled for Linux outside of what Google ships for it… It may have a sandbox at some point for running plugins – but I doubt even that.
The fact that it is limiting is the whole point… That is what makes it a paradigm shift. If it was made capable of running native code then people WOULD run native code on it – and that would defeat the whole purpose. Google is trying to sell the world on running cloud apps – why would they then create an OS for running conventional ones…
It may fall flat on its face, who knows – but I think it will catch on. Remember, as long as it never runs native code (it won’t) it will remain 100% compatible with existing fat clients running a browser. If you are happy with a traditional OS there is virtually no reason to run it, other than it (hopefully) will make it possible to build extremely small, light, efficient, and inexpensive devices that you might find attractive as alternatives (or in addition) to your existing computer.
Its for companion devices now – it wont replace a conventional machine for most people. But if you wake up one day in the future and everything you need to do on a computer can be done through a cloud app well…
But it is NOT a browser only Linux distro… It is a browser that essentially runs on a firmware. Linux just happens to be the firmware for the time being. The browser IS the OS, no one is seeing the big picture because they are stuck on the whole “Linux Distro” bit – Linux is just a cog in the wheel that the user cannot interact with AT ALL. I expect Google to rapidly whittle down the kernel to only the absolutely bare minimum required to run the browser as efficiently as possible.
So if I slap a custom browser on top of X Windows and use the same libraries that exist in all the other distros I no longer have a Linux distro?
The browser is not the OS, it couldn’t interact with the hardware if it didn’t have the kernel and those standard distro libraries. The browser is just the only interface provided to the user.
I don’t care if they lock you into a custom browser. A two-man distro could do the same thing.
Until they fork the kernel or use something other than X Windows I’m calling it a distro.
Chrome OS WONT run user binaries… The point is it does NOT have an application binary interface at any level below the browser. It could run Windows 7 under the hood and it wouldn’t make ANY difference to the user – they CANNOT run anything directly on it.
I’m sorry but I just don’t understand your point of view. I’m really not trying to minimize the importance of Linux here – it IS an important component of the OS, but I don’t see how you can call it a distro when it by definition cant run a user binary…
True. I never said no one else could have done the same thing… What I said was no one else HAS done it before. Name a Linux distro that you cannot install a binary on? And I don’t mean just by default – I mean it offers no mechanism to install one and actively protects itself against ANY modification. It may as well be a black box in my book – it isn’t Linux anymore.
I define a distro as a package of selected components for running binaries targeted at the underlying kernel’s ABI… Whether it is Linux, BSD, Solaris, Windows, whatever – they all share one thing in common – they are built to run software that adhere to the underlying ABI – and generally the kernel is what defines the ABI.
If you made system that used the linux kernel to bootstrap java – and the only thing it could run was java applications… Is Linux its defining characteristic anymore? I mean it just a JVM sitting on top of something, but that something is inconsequential to the user. It could be BSD or Windows or a custom firmware that does nothing but act as a hardware abstraction layer. The point is it no longer MATTERS.
Edited 2009-11-22 05:09 UTC
I’m sorry but I just don’t understand your point of view. I’m really not trying to minimize the importance of Linux here – it IS an important component of the OS, but I don’t see how you can call it a distro when it by definition cant run a user binary…
So if I took Ubuntu, stripped it down to just a browser and prevented users from installing new software it would no longer be a distro?
Why wouldn’t it be a distro with limited functionality?
When I hear Linux I think Unix work alike… If you remove all ability for it to function as what is commonly accepted to be a “Unix”, it is no longer Linux to me. It may use a Linux kernel, but what difference does that make to me as a user? I cannot see it, I cannot use it, it may as well not be there…
But I give up trying to explain myself at this point – if you want to call it a distro be my guest.
Linux distro = Linux kernel + GNU userland and usually X Windows
I don’t see how limiting a Linux distro makes it any less of a distro. There already are distros that lock you into a browser.
http://webconverger.com/
I call BS on that and Citation needed.
Which machine starts karmic that fast from power on??
Most take longer for BIOS messages, Then those times are just to login/GDM. Gnome takes even longer to start. And then Firefox ,,,
So what they showed today (10 sec to a broswer window) is way better than Ubuntu/Mac or Windows.
EFI perhaps? That can connect to WiFi even before the kernel has loaded.
Edit–I remember Google saying in the announcement video that if the checksums don’t match up, they redownload the latest version of chrome to restore the machine. This is all before the system has loaded (because it failed checksum)—and possibly before the kernel loaded (because that’s checksummed too).. ergo, EFI firmware connecting to the wireless.
Edited 2009-11-20 07:44 UTC
Hmm, really not sure about that. From the security docs they put out it seems that when something fails there is always a backup partition which is used to download an uptodate one, so EFI (which according to most Kernel devs is total crap, just like BIOS btw) may not be necessary.
A quick google would have found you the citation
http://arstechnica.com/open-source/reviews/2009/09/ubuntu-910-alpha…
I don’t think it’s including BIOS time though
Well a lot of phones boot faster and are more functional than ChromeOS
Anyone remember thing called Javastation? Ah yes it was supposed to be the great thing that would bankcrupt Microsoft, unmelt the icecaps and bring Dodo back alive. Well it failed miserably. ChromeOS is much like JavaStation with small twist that might save it. JavaStation was aimed on corporates, ChromeOS is for consumers. It’s intresting to see if 10 years have matured technology enough and attidutes changed.
Well, the only thing that saves a bad product is marketing (Microsoft does it constantly). I doubt Sun did enough of it to make it a reallity. But Google, on the other hand, has the power of the web to make it well known. So wether it turns out crappy or not, they will make sure the world knows about it. How do we know about it and why is it already in everybody’s mouth (tech guys at least) even a year before release?
…I like it. It would be pretty useless to me tho’. Maybe good for the kids.
I have Windows 7 on my netbook and I run JBOSS, Java, etc. on it so I can do some development when I travel… which is why I bought it in the first place.
Edited 2009-11-20 16:07 UTC
I’m not sure why so many of the computer/OS experts are so excited about “cloud” computing. Aren’t these the same folks that fuss about security? What is less secure that trusting someone else with your data? Data that you cannot access without the internet. This is crazy. I don’t trust MS, Apple, nor Google with my data, why should I? I have been to quite a few places where the net isn’t accessible, what do ya do then? I’m not sure I’d trust a company with my data that can’t seem to “search” w/o ensuring certain companies always popup in the results. I was hoping that Google would be a real player in the OS world…how sad, nearly as bad as their browser….which isn’t as bad as IE, but still pretty horrible. Geez!
Someone breaking into your computer and stealing your data is an embarrassment/setback to you and/or your company. Someone breaking into google datacenter is a cataclysmic disaster for a multibillion dollar company.
Guess which one gets secured better?
or truecrypt if your version of Windows doesn’t come with it.
Edited 2009-11-21 06:08 UTC
I guess that won’t help if your machine is on, and someone uses an exploit to gain access.
but most exploits involve getting people to follow links in emails or install software. Widespread system exploits became rare after XP got a decent firewall. Conficker would have been killed off quickly if it wasn’t for all the pirated XP installs that have auto-update turned off.
Google’s security plan may be good for grammy but I’d rather encrypt and backup my own data. What happens when Google’s servers go down and you need an old document? Wait while playing some online flash games? No thanks.
Does this mean someone can hack your computer even when it’s not on? Kinda like Sara Palin losing her email acct. So if a hacker came up with a 0day or just started bruting Gmai.. I mean Google OS, he could potentially own everyone, regardless of your Firewall, security software etc? Also as far as security goes, why is your data any safer in the cloud? I’m assuming if I break your silverlight (shutter) session I can still get to your data (just on the server now, @ http://fjkfsdjksa instead of C:\my do~) (what this won’t run silverlight? but I need to…)
I wonder what is “data” for most people here? Your mp3s? Your movies? Your pictures? Your documents?
Most documents are sent by mail in the clean, and thus can be potentially sniffed, stolen by bots, etc. If you want to be sure nobody can read a private document, encrypt it. The place you finally store them is irrelevant if you use encryption.
And for those that care a lot about their financial documents, I’m sure most of them don’t have an encrypted partition in their laptops, they don’t use any technology that prevents eavesdropping when editing documents in a public cafe and they don’t even use a virtual keyboard when accessing their online banking. Most people talking about “their data” don’t realize (as I didn’t) that nobody cares a shit about what they do with their family pictures, and that just by using Google, Yahoo, Microsoft, etc, you are being tracked by IP (unless you constantly use anonymizer), browser, OS, language, and profiled so they can sell you better ads.
So, what’s your “data” that it’s so important as to be held only locally, encrypted and completely secured from spyware?
What do you have that’s so important anyways? No one cares about your family photos.
That along with Do you really need itunes? is going to make these things fly off the shelves.
And yes many of us do work with sensitive information, especially those of us in the proprietary software industry.
I also know quite a few people in the medical industry and they would never store their documents on Google servers for the simple reason that they don’t have to. They operate on a need-to-know basis and Google doesn’t need to know about a word document that details your venereal disease. It’s better kept on the local intranet.
I also know someone who works for the feds and they don’t even connect their computers to the internet. To get something like winzip installed requires going through a chain of command and filling out a form.
There will also be people that oppose the idea simply because they feel it is not respectful of their privacy. Google should have built a hybrid system, going web-only will make it instantly unappealing to large sections of the population.
Why isn’t this a phone OS? Of all things, my phone is connected to a network 99% of the time. It does small apps and wouldn’t need to download Photoshop etc. Small uploads. It could be quick to respond to some new innovative tech as long as the hardware supported it.
I read the Google announcement, saw they are working with hardware vendors to have netbooks at this time next year, and that they have released the source code. Unless I missed it, I saw nothing about betas. Are non-source code/compiling type people like me going to have to wait a year to try it out? I know there has been discussion pro and con about driver issues, but will that or something else preclude public betas?
what a retarded video. Guy screams like a girl that it takes a whopping 45 seconds to boot yer PC, and that he could “make a sandwich” in that amount of time.
What. The. f–k.
Yah, I’m sure his (and most pc user’s) time is SO valuable, that they can’t wait 45 seconds.
Chromium OS: Solving problems nobody has.
I love it. When Microsoft had the idea way back when the desktop was an extension of the net (Active desktop), I was stoked. But then the idea seemed to have died down. Why? Anti-trust cases, the decoupling of the internet from Windows itself. Now Google realized the idea much the same but took it to a deeper level. Again moving back to the idea of very thin clients much like what Sun used to laud. I thought it’s a nifty idea. But it never caught on. Of course for now I would like to see this on an ARM based system. Light and savvy. Obviously this system is not for heavy applications but time will tell, with the advancement of the net.
Thin clients have been mentioned before in this discussion, but if we concentrate on the software rather than the hardware, in my opinion the model that springs to mind is SaaS.
Of course Google is not going to ask for money as such, and I expect the markup on the official netbooks for Chrome OS will be microscopic, but an ever increasing audience for targeted advertising is still worth a lot of money.
Anyway, I wrote a lot more about it here — take a look if you wish:
http://www.zitellotonnato.com/2009/11/chrome-os-looks-cool-but-smel…
RT.
Edited 2009-11-23 08:36 UTC
I’m really excited about Chrome OS. I really hope it is able to take a large bite out of Microsoft’s domination of the desktop. For years, I’ve been wanting to try different OS, but because of compatibitily with work, I’ve always relegated myself to Windows. So, I’m hoping others can jump on the bandwagon. There’s a lot of useful information that are starting to become useful to explain exactly what this OS can do and what it has to offer such as…
http://www.allchromenews.com/2009/11/22/google-posts-several-chrome…