I'm not sure if I like implications of a web-centric OS, especially when its producer is as rich and powerful as Google. I see the market for it, and I think it will be a huge benefit for the average user. However, I see the transition from desktop-based software to web-based as a loss of user control. Today, when governments decide that mp3 sharing, porn, articles of political dissent, or news from non-national sources should be eliminated, they quickly realize, if they don’t already know, that elimination is impossible, because no agency can track down and deal with every user who has decided to install software allowing him/her to do these things. When software moves to the web, I think it will likely end up existing on the servers of one of an oligopoly of computing providers, who will be much easier to influence because of their small number, and because they will have no choice but to abide by legislation if they wish to remain in business.
This may not seem like an issue right now, because there is still choice. But with Google pushing a web-centric OS, I think within a decade, many people will have a hard time choosing something else, much the same way they had a hard time choosing something other than windows until the last five years.
Maybe I’m just a Big Brother fearing nut, but I don’t like where we’re heading.
I concur. I've become more and more aware of the looming Google benevolent giant, but one day that giant's going to get angry and by that point it may be too late.
That being said, I still use Gmail, Google Docs, Google Reader, Analytics, and other services. So I'm not exactly avoiding the apocalypse here.
What I think is interesting is that having a hard time choosing something other than Windows may have been quite benevolent in comparison with a Google-situation which turns ugly. What if there is an organisation that has all your email, all your documents, all your appointments and all your conversations and can follow the majority of your surfing via ads. What if this resource suddenly is too tempting to keep hands away from for "fighting terrorism" or something such.
Being forced to use Windows at work could be considered nice in comparison.
I don't think that anyone at Google set out to be evil. But power corrupts and the legal systems and governments have not exactly struck up a fight for privacy for the individual the last couple of years.
Yes! Thank you for capturing more of what I could feel in my gut, but could not bring to words when composing my comment.
I don't think that anyone at Google set out to be evil. But power corrupts and the legal systems and governments have not exactly struck up a fight for privacy for the individual the last couple of years.
I also see scenarios where Google never becomes corrupted by power, but by government mandate a la Patriot Act is forced to reveal content. I don't like that there appear to be many plausible ways where the whole situation could end up generally bad.
Indeed Google may never turn to the dark side, but to have that much information under one corporate entity's control is terrifying to say the least.
Unfortunately in the digital age, it is hard to achieve any semblance true security for your personal information without acting on what lay-people perceive as paranoia.
This is what really bothers me about it. If it weren't Google, it would be someone else.
What we probably have to worry about is that the base level of privacy that is easy to achieve is in freefall. The incentives for any individual organization are too strong, because of the level of service (to consumers or advertisers or whatever) can be so much stronger. It seems to me that compounding this is the fact that the 'digital age' makes monopolies natural in alot of places, even if only for a short time.
We're already rapidly adjusting to a loss of privacy, and maybe it's a necessary progression and privacy advocates will one day be a quaint historical footnote.
Even so, I'm definitely in the paranoid camp, and if we don't make a conscious effort to have a more active cultural attitude towards privacy, we might not be able to preserve any of it if we have a collective 'oh shit' moment down the road.
As long as Google doesn't do anything to limit my access to competitors or otherwise punish me for opting out of a Google service then I am not concerned at all.
The FUD is strong with this one. If you want to run your own apps, run your own server. Those who don't care will use the public cloud. Nobody controls the whole Web, and as long as your ISP is reasonable (which none are), hosting your own private app is still possible.
When we create new software that benefits from lots of networked users, we will be forced to make that software compatible with the people who do not have their own private servers. So users can run their own servers, but there will be those who cannot afford to do so or do not have the expertise. If there are enough of those kinds of users we will then have to bend to the needs of those users who in turn bend to the rules of the companies hosting their software.
I think the inevitable consequence of fears like yours is that you'll be able to pay different institutions, so that you can store part of your data here, and part of your data there. So that you can surf behind company A's proxy most of the time but use company B's when you feel like/need to.
Yes, there are going to be less entities on the whole. But if the internet keeps growing, that might be a temporary setback--the infrastructure is going to hopefully become more spread out, more resilient to attack, and more designed to protect individual rights, and individual property. Think about it this way: as we move more and more pieces of our lives online, the value of our digital "stuff" grows higher and higher, and systems to protect that stuff, and our hopefully our constitutional rights (including the right to privacy), will appear.
At the very worst, you can always wait for some kind of massive privacy or fiscal (or even just data loss) tragedy would at least cause governments to act quickly to more clearly define the rights of an individual online :p
I agree that we have different options. We don't HAVE to choose Google, just as we didn't have to choose Windows; it was just the easiest/most convenient/accessible/popular way to go and so most people do it. Also, I'm sure more systems will be created that are designed to protect our data, as you say, though all you're doing is putting your trust in Corporation B over Corporation A.
Constitutional rights? Right to privacy? The government can listen in on any phone conversation they want (wire, cell, voip...), read any e-mail that gets sent, break into your home if they would like, et cetera. What rights do we have exactly?
Power corrupts. It's human nature. Your trust in the government to not abuse its' power is scary to me.
But what if there are web-centric OSes and desktop-centric OSes?
In a world where the OS market is fragmented, centralized control and opression becomes even more difficult than it was in a desktop-centric world where one OS had huge market share.
Now that's not the case if we end up with one web-centric OS, but I don't see that happening any more than I see Bing taking over 65% of the search market without any response from Google.
Maybe Google should enable the users to save their data to whichever place they wish. And the web OS, or the authorities who're running it should never have access to users data. The OS should just be the communication interface. Maybe this is utopian, but i guess its the right way to go from here.
If ARM and x86 are targeted, this means no Flash (not surprising, given how poorly it runs on Linux), which presumably means YouTube will be operating in full-on HTML 5 video mode by the time these devices launch.
If netbooks continue their meteoric ascent, and if Chrome OS grabs a significant share of the market, we may finally see the realization of the mythical Year of Linux On the Desktop. The ironic aspect, of course, is that even though a large number of non-technical users will finally be running on a fully open-source platform, they will use it to run applications whose workings are even more opaque than traditional closed-source, client-side apps.
I wonder what role, if any, Native Client will play on these devices. I understand Native Client to target only x86, but I find it difficult to imagine Chrome OS -- Google's first direct challenge to Windows -- not being able to run Native Client applications.
i agree with you. the comment with the large number of votes is a fearmongering comment, while the comment that you are referring to is actually a technical comment (the kind that i think is more relevant at HN).
I'll rather have some fearmongering (if you really want to call it that, there's another better term though "devil's advocate") from people to have a certain balance of opinion rather than just have people who go gaga over every google/apple announcement. I dislike MS as much as the next guy but those are genuine concerns.
Note, his comment is not like "sky is falling" but more like "beware, the sky may fall if we're not careful". I'll rather be cautious then stick my head in sand with my behind exposed.
My above comment "I enjoyed your narrative." is meant in a sinceren and non-cynical, non-subtext-whatsoever, non-ironic way. I wrote it to express a bit of appreciation towards Micand for crafting his imho good summary on the topic.
Can anyone explain to me, what is worth downvoting about it? Thanks.
So effectively your PC becomes a Google ThinClient.
Got all the Office Applications with Google Apps. Email/Communication with Gmail. The rest of the web. Multimedia with YouTube/Hulu/etc. Sounds like what majority of the people need. Even for work environments, would be pretty good (you can offer your proprietary application in the form of a web application).
What I want is a good enough "instant-on" that it's my only partition, and I no longer have to worry about hibernation or sleep mode ever again. Have orthogonal persistence for everything, so I never even have to save. Just open the clamshell, or hit the on button. Something comes along to distract me, I snap it shut, take care of whatever it is, then open it up again, wait a half second and continue on where I left off.
The ideal laptop/internet pad would behave like the Nintendo DS, if the DS could hibernate, and not just sleep when you close it.
If their approach to Android is any guide, much of the work will consist of eliminating any GNU or GPL components from the (userspace) runtime environment.
Google did for Android so they could standardize around the Apache License for userspace. But who knows, they may not bother here. The netbook market is decidedly different than the mobile phone market, so they probably do not have the same constraints.
Yeah, they definitely prefer the Apache license, but the "userspace" in Android mostly consists of Dalvik and a bunch of stuff written for it/on top of it. Since they're trying to save space, they went with a smaller libc and don't include much at all in the way of 'standard unix tools'.
In other words I don't think they care too much about gpl stuff since they built their own UI and bytecode interpreter where most of the action is anyway...
This denies that Android runs on top of Linux, which is licensed under the gpl. Or that the Google search appliance runs on a ton of gpl'd code, all mirrored off of code.google.com...
You must be talking about some other comment, because I don't mean to deny any of that. I was merely pointing out that on Google's largest (public) OS project to date (Android), Google evinced a strong desire for a copyleft-free userspace and succeeded in building a non-GNU Linux distro. This is clearly a different project with different goals, so we'll see how it turns out.
The stated reason for that was that Carriers and Telcos wanted to build on Android and were put off by the GPL. (Though obviously not put off enough that they'd reject the GPL'd Linux that runs beneath it)
Google's approach to licences is far more nuanced than blind adherance or avoidance of any particular licence.
Webkit is LGPL, Chrome is based on Webkit, and the entire OS is built around Chrome. It's also speculated that their "new window system" is built on DirectFB which is LGPL.
I'd be interested in hearing your reasoning on that one. Lots of people complain about X in various ways, and while a couple years ago I would've agreed, these days, I'm no longer entering my monitor's vertical refresh rate manually, and X is pretty transparent to me.
And on the up side, it's nice to be able to run a GUI app remotely, though I'll agree that that's not winning over too many desktop users.
I wonder if people will be able to take this new windowing system and use it for their own purposes. That would be fantastic, if the new windowing system is good.
No doubt it will be a windowing system optimized for showing exactly one full-screen "window", and perhaps occasional modal dialogs.
I believe both Qt and GTK+ can be compiled to draw directly to the Linux framebuffer, if you don't want to use X (although they don't exactly provide much in the way of window management).
Yes, as soon as it becomes open source later in the year. I'm not sure how fantastic it would be, considering it would be Yet Another Windows System which only runs on these netbooks.
Instead, I'd be really impressed witha good windowing systems impemented using the canvas tag, something like:
I personally would never trust google with something as important and all encompassing as my OS. I value my privacy far too much for that, and seeing as how google collects usage data for just about every action you perform on services they offer, I'm not sure i'll ever go the google way on desktop software.
Google knows who my friends are (Contacts), what I'm saying to them (Gchat/Gmail), where I'm going (Gmaps), what I'm reading (Reader), what I'm writing (Docs) and what I'm looking for (Search).
As far as I'm concerned, operating systems are pretty interchangeable. The things I care about are already all going through Google. Does anyone besides Google offer a similar suite of apps? The only one I can think of is Apple's MobileMe.
How about Outlook/Thunderbird/Facebook (contacts) + Yahoo (mail) + OpenStreetMap (maps) + Sage/Friendfeed (RSS) + Zoho (Docs) + Bing (Search)?
While no-one may be offering a similar suite of applications, if you want to use them you can find different services that do.
Also currently even if google does know all that information about you, it doesn't know what files you have on your hard drive. In my opinion, giving carte blanche on data on my hard drives to google is a big no for a company that doesn't give two hoots about the privacy of their users.
The value of Google's services, at least for me, is that they all work together. I can login from an internet cafe in the middle of nowhere and have all my contacts in my email program, the same ones synced to my iPhone.
I don't know if I buy the claim that Google doesn't give "two hoots" about your privacy, as long as you don't count selling contextual ads as over the line. But anyway, my point is that the "files" I care about protecting are already with Google. For true protection of your local files, on any operating system, encryption is your only answer.
I believe with facebook connect, you can maybe hack those different services to work together (firefox extension?). Maybe, I should take a look at that idea.
And if google did care about my privacy why read all my mails, and keep track of every webpage i visit from their search engine. I don't buy the "we'll only take a peep, just for advertising" logic.
Also, google already having all the files you care about would be more an exception than the norm. I would presume a majority of people keep their files stored on their hard drives.
In my opinion, keeping track of my actions in any way goes against google's self stated "don't be evil" policy. Or did Paul Buchheit take that policy with him to friendfeed???
Microsoft knows who my friends are (Messenger/profile), what I'm saying to them (Messenger), where I'm going (Bing Maps), what I'm writing (Office Live), and what I'm looking for (Bing). Granted Office Live is a beta and there is no Microsoft-branded online feed reader. But a lot of the checklist is covered.
If you start checking off every box, you'll find that Yahoo has a fairly similar total suite of properties, too. Google just has the most mind share.
The key to supremacy in the enterprise market is to seize control of the entry point.
IBM did it in the 1960s and 1970s with the entry point being a highly skilled personalized sales force in an era where relationships established the trust needed for enterprises to commit to large systems.
Microsoft did it in the 1980s and 1990s with the entry point being its proprietary API for its desktop OS which, with the help of OEM hardware partners, it used to choke out anyone who sought to compete with it in the desktop applications market. It then tried (and failed) to use its desktop monopoly as a vehicle by which to dominate the back-end of enterprise computing.
Google now seeks to gain similar supremacy with the entry point being superior proprietary products ultimately derived from or at least capitalizing upon Open Source in a web-centric computing environment. What form this will take, I don't know. But it will take a form by which Google tries to seize and maintain control of the entry point and ultimately bend the world's primary computing environments to its proprietary advantage. The Chrome OS is one piece of this effort.
The question is one of which company will become the ultimate gatekeeper to all or most of enterprise computing, which is where the money is. Google wants that role and that is where its fight with Microsoft will lie.
As they say, "The law of human nature - it ain't been repealed yet."
Microsoft's success was based almost entirely on lock-in: they made everyone depend on something that only they could provide. The equivalent strategy for Google would obviously be to hoard the world's data in their private silos, which they are arguably already doing. I would find that scenario more plausible if it weren't for Wave.
Wave, by its nature, will obsolete or radically transform, many of Google's existing web apps: gmail, docs, blogger, picasa, calendar.. anything related to communication or collaboration. Yet, it seems as though they've gone to great lengths to make Wave decentralized, even as far as open sourcing their server implementation. They've done this even though a silo style Wave app would likely be a huge success anyway. I can't fathom why they would allow Wave to exist if they were hoping to control everyone's data, as it clearly represents a threat to the control they already have.
I'm not dropping my guard, but at this point my best guess is that Google is just trying to create an open and fair web ecosystem, in which they will compete based on their momentum and massive resources, and in which their major competitors lose their existing advantage. Either that, or they don't have much of a plan at all and the company is really just a bunch of smart people doing cool stuff with a huge pile of money that fell from the sky.
Yes, that would be a perfect device for it. Because, when we talk about netbooks, I am not so sure I'd like to have Chrome OS on it. Ok, it would boot faster and maybe also run slightly faster, but the boot time isn't that much of a concern for me. On the other hand, I have a half dozen windows apps I want to run on it, since I don't live on the Internet only.
They have one of the best chances to pull it off out of all the attempts over the last 25 years.
I really hope they try to go big -- bury ALSA cleanly, use Skia for all drawing as vectors, and have an event model intended for Cocoa/WPF/GTK# (not the crapulence of Xt).
It's taken 14 long years -- and it's Google-Chrome-plus-Javascript rather than Netscape-Navigator-plus-HotJava -- but the browser has finally become a complete alternative to a desktop OS.
The article's final sentence: "In the age of the hollowed-out computer, the king of the desktop rules an emigrating empire."
What if one device has a switch on the side: 'locked-down terminal' or 'open general purpose computer'? And via virtualization, the two sides are sealed from each other, but instantly switchable? Why not the best of both worlds?
Im convinced it's not a coincidence that this week they also announced dropping a load of beta tags, hid the free apps sign up page a little more and made a number of other tweaks.
I get the impression they are trying to repackage their cloud services together: it would make sense if the plan for Chrome OS is to ultimately tie it all together.
If you haven't tried out the chromium nightly desktop builds, they're actually quite nice. The team is still nailing down the plugins, but the browser itself is wicked fast and pretty stable (given its alpha nature)
And yes, it is wicked fast. On my system, it doesn't even bounce in the dock before opening. It just opens. Plugins aren't supported yet, but I just open any sites that require Flash in Safari.
> On my system, it doesn't even bounce in the dock before opening.
I think it doesn't bounce for a different reason. For example, Firefox doesn't bounce when it autoupdates. Everything should bounce for at least one time in normal conditions.
I think stable for a web browser these days requires flash/java support, I don't know if this has been fixed or not in the builds though. Probably less on the java side and more on the flash.
Even for a minimalistic web-browser Chromium is still rather lacking. While what used to be butt ugly font-rendering has gotten better, at least to the point of readable, as far as I can tell there is still no plugin support.
The only way you can fully configure the browser is still trough running gconf-editor for some parts, and configuring Firefox and then importing settings from there.
That being said, my biggest issue with the current chromium builds is that performance is truly horrible and majorly unstable. Sometimes things just works, snappy, as expected. But in more than 50% of the cases when opening a new page the entire browser locks, for what can be periods of up to 5 to 10 seconds. Wasn't process isolation supposed to fix this?
If this browser is going to be the "OS" for the web, in it's current state its closest equivalent would be Windows 3.11.
IMO Firefox 3.5 is a much better browser, while peak performance being somewhat slower, at least it performs predictably.
Wow - completely opposite from my experience. What are you running on? I'm on Ubuntu 8.10 right now, using the nightly builds.
I've had a tab lock up once or twice, but in general it's very nice, way faster than FF 3.0, and pleasant to use. The key bindings are the only thing that seems a bit wonky, but I'm generally happy and at some point may start using it as my main browser, if FF 3.5 isn't significantly better.
Also the complete opposite of my experience. I used the nightly builds almost exclusively for over a week on Ubuntu 9.04 and I had no performance issues. Things are certainly incomplete but I was slightly disappointed to go back to the poorer performance of Firefox. I should probably place that blame on the Flash plugin but I think Chrome will handle this problem much better.
I'm running this on a Asus EEE 901 with Ubuntu Netbook remix running the nightlies.
My biggest issue is how the latency of the browser seems to vary between milliseconds and two-digit seconds. Performance is randomly varying over three orders of magnitude and that kind of performance instability is worse than performance being consistently "slow".
I had problems where after a few seconds it would completely crash X, drawing garbage all over the framebuffer, but it turned out that it was just triggering some bug in Compiz+Nvidia on Ubuntu.
Though you'd think if they're announcing a linux-based operating system, centered on their web-browser, making their web-browser run on linux would be trivial. OTOH, could it be the diversity of linux distributions that's making it difficult for them?
I don't get the impression that is actually a significant problem for anyone. Most distributions include the same libraries, recent kernels with similar options and even put things in similar places. Packaging is the major difference - though that's a job best left to the distributions anyway.
They have been working on the linux port for only 9 months porting what was essentially a native windows application written in Visual Studio. The fact that they got this far this fast shows that they have a pretty good team.
Since they say they're implementing a new windowing environment for this project, there may not be a lot of porting overlap between Chrome for Chrome OS and Chrome for your normal Linux distribution.
I'm glad Google has identified a need for something like this, but I don't really trust that Google has the right talent to build an operating system. They have serious problems with user experience and interface design (see: Doug Bowman). The mere idea that they are attempting to encroach on an aspect of experience as large as the operating system is to me somewhat scary.
I do hope they prove me wrong. But based on their past performance, I am not holding my breath.
Also: Google, if you're reading this, please at least make the typography render properly in this OS. You've already failed at this with Chrome, which has some of the worst antialiasing I have seen in a browser (surpassed in crappiness only by IE).
As I said below (http://news.ycombinator.com/item?id=693339): you've not got what they're thinking of as the OS. The browser is the operating system, and the apps are all web apps. The core of the OS is Linux, but the interface is all browser -- you don't install apps, you just type in their URL. There's hardly any OS UI they need to invent that we haven't already seen as part of Chrome itself.
Not really. You still have to be able to log in, store data, set up parental controls, deal with power management and screen savers and wifi networks, choose desktop backgrounds (assuming they don't force you to use a single tabbed window), choose keyboard/language/sticky keys, define trackpad gestures, accessibility settings, etc...
I am curious to see how they deal with this stuff.
My guess is that you will be in a single tabbed window, with special tabs for things like wifi, login, file browser, settings, etc (like Firefox's about:config).
At least that is how I'd do it. Just type config:wifi in your location bar to change your wifi settings. Visit config:files to browse your local files, etc. Of course, all of these can be bookmarked by default so the end user can click rather than type.
Chrome already does stuff like this with about:memory, about:network, etc.
Interesting. The one missing piece--a RAD development environment that's as efficient as Visual Studio is at producing fat clients. The business world will only embrace the webcentric model when it becomes as efficient to code in as .net click-once fat clients. The idea of simple, cheap, rugged hardware is very appealing. But programmer time is the #1 cost--businesses go for easy drag/drop fat client apps. Software written for 5 users has to be thrown together pretty fast or it never pays back...
I don't think that Google Chrome OS would be suitable for most corporate shops that are generally Microsoft centric (and have been for quite some time). I think this free OS will be geared towards light users that primarily use their computer for surfing the web and checking their email.
RAD = rapid application development, the business term for "drag-n-drop GUI builders" that we all know and hate. Not having one for the Web is a valid criticism, but if GWT can handle Swing API input then we DO have one.
I'm familiar with RAD, although at least for my employer it's more than just drag-and-drop GUI builders.
It sounds to me like jeffspost is over generalizing and suggesting that Google Chrome OS will only succeed if it's embraced by large corporations. The software hasn't even launched and jeffspost has already figured out it's "one missing piece".
And generalising that it will need to replace Windows.
They are launching it as a netbook OS for a reason. Netbooks is a loose term. I'm not exactly sure what the line is, but one definition might be a small laptop that doesn't replace your main computer.
I daresay that whoever buys this will either have another computer or have no need for corporate apps.
That sounds cool, but I don't know exactly what you mean... I think you might mean that you could create a Java applet front-end (using "Swing API"), and drive something in GWT with that, if "GWT can handle Swing API input". But how would you create the Swing front-end? What would driving GWT do you? Please join the dots - doing this often reveals gaps that weren't clear before, so it's a good exercise.
Taking another view: any Java app can be an applet (if you take care of things like access to the file system). Therefore, we already have then, since Eclipse and Netbeans can be an applet, and thus a web app. Theoretically, anyway. :-)
GWT compiles Java to JavaScript, right? So compile the Swing calls into calls to equivalent JavaScript UI library calls. Barf if they try to do stuff you can't do in JavaScript (filesystem), as GWT surely has to do already. (disclaimer: never coded GWT in my life)
Being Microsoft-centric, in itself, is inertia - it can change if there is a sufficiently superior alternative in terms of businesses' actual needs.
Google OS can start off in low-end markets (like Toyota did...), and then move up.
jeffspost notes a startup opportunity: webapps for creating webapps (actually webapp clients, but that's less memorable)
They don't need application vendors to target Chrome OS in order for this to be a success - the whole point is that it just runs web apps, period.
Its targeted at consumers and netbooks, not those that want to run heavy duty, CPU-intensive applications. The whole idea of a netbook is to be small, light, quick, and primary a web browsing device.
Google has already indirectly "extended" Native Client with native 3D graphics using their O3D API (http://code.google.com/apis/o3d/). It's just a matter of time until someone codes up a library to expose that API through JS interop in Native Client.
I think we can expect very close integration of Native Client into the Chrome OS as soon as or right after it goes public. To the level of accessing hardware ports to control external devices right from your Native Client web-app.
Imagine drivers that get loaded and updated on-the-fly. Of course, Google will have to implement a smart security mechanism (along the lines of RSA authentication for dynamic device/driver coupling) and provide a way for device manufacturers to register their driver's public keys with Chrome OS.
I think Native Client's environment is constrained enough that it would be feasible to implement JIT translation between architectures (i.e. x86 -> ARM).
ISA translation tech is already widespread: a company called Transitive has been quite successful in offering translation products for Unix vendors who have needed to transition their customers to a new architecture (most visibly they provided the PPC->x86 translation engine for Apple's "Rosetta").
With Native Client, the translation ought to be made easier by the structural limitations that are imposed on NaCl modules primarily for security purposes -- to use a natural language analogy, it's easier to translate a text if you know all the words originate from a certain dictionary.
I think this is great. Most people don't know what a browser is anyway, so for them it probably won't make much of a difference. It will be interesting to see how much of an OS you can strip away.
I don't see that anyone has mentioned Apple in any of this. They certainly aren't just going to roll over and let Google take this entire space. They will launch a netbook at some point with the iPhone OS, some storage, the app store to get everything you want, a superior user experience and industrial design, and oh yeah, the browser to also run all these other apps...And it will be this kind of weird hybrid old model vs. new all over again.
I think Google will continue to "win" and we will use their applications/reader/ecc until they will have the dominant search engine. If you look at what they offer (with the exception, maybe, of excel-like webapp), they are not killer application. We are using them because our online-life is Google-centric.
So we should remeber that everything else they do is buld on their search engine dominance (think to the omnibox)...
For the webos is the same. It's fine, for a netbook IF Google will remain Google.
If you look at what people do with a Mac, they do far more than just using web application. Far Far more.
The danger is and always has been centralized control (and not just in software). No one should own the web OS — the web is ours.
Stallman & Co. foresaw this (at least) a decade ago. He's right to be worried. The real issue is: what are rights in the digital age? Who has control of what?
I might not back everything he's done. And he sometimes dismisses technical issues as trivial over political ones unnecessarily. But he is having the conversation at the right level of abstraction.
We'll keep facing the pitfalls of centralized power until we collectively decide to kill that beast permanently.
I've an AA1 running Linpus, which is buggy and slower than I anticipated. It's bloated down with superfluous software that I don't need, and that you have to follow a series of complex guides to remove. (Even then you risk screwing up parts of Linpus which are dependant on it.)
All I use it for is browsing and light coding on the move. Windows, and the likes of Ubuntu Netbook Remix are all too 'big' for that.
It sounds like the Google Chrome OS is exactly what my netbook needs.
I get the impression that the only native application it will run is Chrome. So how are you going to edit your source code, push/pull your changes, test it, run it?
I've seen web-based ssh clients, and someone might be able to write a web-based vnc / remote desktop client, but I can't see it being useful for offline coding.
None of my websites (that I would edit on the netbook) are business critical / make money / have any real world importance whatsoever. I can log in via cpanel, roll out some coding amends live and worry about breaking things as and when it happens.
Is this any different than someone else taking Ubuntu/Firefox and calling that the Firefox OS? Chrome isn't an operating system, linux is. This is a distribution and pre-installed browser. How is this anything different that what we currently have in the netbook market.
The real player in the future of netbooks is very likely Apple, not Google, because any netbook running OSX will also be able to not only run all the usual OSX apps but Chrome too...
2 years ago I made cl33n, a Linux live CD that boots Linux, connects to the net, runs the Matchbox window manager and launches Firefox- that's all it does. ChromeOS is "Google Chrome running within a new windowing system on top of a Linux kernel." cl33n is "Mozilla Firefox running within a little-used windowing system on top of a Linux kernel". cl33n isn't unique, Webconverger and other great live CDs do a similar "make it easy for anyone to browse safely" job.
Google cancelled cl33n's AdSense account 8 days ago, right before they were due to make their first payment. The site and its small traffic & AdSense revenue hasn't changed for 2 years (am working on an update now, I got kinda busy with a startup ;)). If cl33n did Bad Things then it's been doing them for 2 years now. I wonder if Google will refund the money to the advertisers?
I've appealed but have not received a response from Google. I still don't know what cl33n did to incur Google's wrath.
For application developers, the web is the platform.... Google Chrome OS is being created for people who spend most of their time on the web, and is being designed to power computers ranging from small netbooks to full-size desktop systems.
Does this mean an operating system that runs chrome and little else or an operating system that runs mostly chrome?
Using Google Chrome OS, means that we throw Windows and OSX away. Windows or OSX (Desktop-application) are not good, flexible because they are dektop applications, but because they have dozen of years of user experience. Throwing this user experience away, means we are starting the OS history again, which is a lose of time and money. In my opinion!
It is also interesting to think about this news in the context of FaceBook buying Parakey (http://en.wikipedia.org/wiki/Parakey), though it is a web-based OS. IMO, given Google's array of applications, they are better suited to write a web-centric OS than FaceBook.
we believe choice will drive innovation for the benefit of everyone, including Google
This is the kind of competition that I would like to enjoy (i.e. between products that both belong to me). I wonder if it actually will work out for them, and if so, how I could use that strategy myself.
I switched back to Firefox for security reasons and because Chrome turned out to be really slow re-rendering tabs that haven't been visited for awhile. Not too enthused about a whole OS but I suppose that might be one answer to the security problems.
I can't wait to try an actual version of this in a virtualbox. Betas should be available for geeks way before the netbook manufacturers get to installing it on their devices in 2010.
It's going to be interesting to see how committed Google is to this project. Google seems like a company these days with more ambition than substance sometimes. They have some really good services but they also have a lot of services that have seen very few, if any, significant improvements lately. It does remind me a bit of Microsoft when they had to get their fingers into every category of software even if they weren't offering a particularly good product. It was more about being in that space than actually being the best at it. I hope Google is not trying to go this route because it almost always ends up in mediocrity.
You have to re-think what is meant by an "operating system".
What is Windows? It's that thing, in the background, that lets you run all the applications you want to run.
But what if all your applications were written in Javascript and HTML, and had URLs instead of .exe files? With all the extra power built into Chrome, they could still write to file systems, store data, work offline, do sophisticated graphics and parallel processing -- these are not hypotheticals, this is what Chrome can do right now. We're just beginning to scratch the surface of what an app written from Chrome can do.
When the iphone was first released, development was going to be via javascript and safari. That didn't go too well with neither users nor developers. Isn't this going to be the same thing? And for the so called windowing system I would not hold my breath and except anything like quartz. Its probably going to be just enough framebuffer code to drop X11 so the browser window can run in fullscreen. The window manager will then be implemented in javascript. Wee!
When the iPhone was first released, development was restricted to web apps and they didn't have access to any API. The fact that it was javascript and HTML was mostly tangential to the complaints of people who wanted to install (and sell) local apps, with access to the unique hardware, though obviously existing Cocoa coders, 3D game programmers and anyone needing to get close to the metal weren't going to be happy even if these other elements were taken care of.
On the other hand "fart apps" and things that are basically light wrappers round existing web apps have no such problem.
See the Palm Pre for someone taking this to it's logical extreme.
If they depend on Javascript as an applications language then I hope they add goto to the language. At least that's what another HN discussion has led me to believe.
You're forgetting, though, that it is an OS that (appears) to consist solely of an environment to boot a browser as rapidly as possible. And, really, it consists more of linux kernel modified for rapid boot, probably optimized for low battery usage etc. They are not building an entire OS.
They will probably build something like fbdev wrapper to run their browser along with Native Client and GPU-accelerated (even on ARM) ffmpeg's codecs for html5 support, and it could be a real deal. They will reuse 80% and build 20% as usual.
This may not seem like an issue right now, because there is still choice. But with Google pushing a web-centric OS, I think within a decade, many people will have a hard time choosing something else, much the same way they had a hard time choosing something other than windows until the last five years.
Maybe I’m just a Big Brother fearing nut, but I don’t like where we’re heading.