My prediction of how Chrome -- both the browser and the OS -- is going to flatten the playing field for programmers.
As I've said many times before, the biggest problem with web programming (and most programming is either on the web or going there) is the lack of a consistent programming platform. The amount of effort and money wasted to compensate for browser inconsistencies is phenomenal.
This is why Flex and Flash are doing so well. You write programs without even thinking about platform differences.
The price you pay for this is the insertion of a piece of code that checks for the browser and pops up a dialog that says "this application only runs under Google Chrome." Perhaps it automatically starts the Chrome browser if it's one your machine. If it isn't installed it takes you to an installation page.
This is no different than the experience people have with Flash or Java right now. Well, except that people don't hesitate to install Flash (the installations are virtually 100%, and the upgrades to recent versions have been happening faster than expected), because Adobe spent a lot of time early on focusing on the end user installation experience. Unfortunately Sun put the installation experience for Java on the back burner for so long that the window of opportunity may have passed (there's finally some motion in that area, so we'll see).
The vast majority of basic computer consumers trust Google, and will not hesitate to install the Chrome Browser in order to run an application, in the same way they install Flash to run YouTube.
Here is the Dell Mini-9 Netbook with Linux at $239. (If you just go to Dell and look up 'Netbooks' you'll end up on a page which has no choice but Windows XP, thanks no doubt to pressure from Microsoft; you have to actually search for 'Linux' to find the above link). Note that you can "personalize with Windows XP for an additional $150." Some of this is likely the additional hardware requirement to support Windows, because elsewhere on the site it appears that the incremental cost of XP is actually $60.
As the price of netbooks drops the Microsoft tax becomes more obvious, and more painful to consumers.
The early years of Linux was targeted to geek techno-nerds. Out of curiosity, I would try about once a year to install Linux, and give up when it started asking low-level technical questions I didn't know how to answer. Finally Red Hat turned the corner and made a Linux that would figure out how to install itself. And in recent years, Ubuntu Linux has made installation and use a pleasant experience for the average consumer. But it will be a few years before it shakes off its historical reputation of being only for uber-geeks.
The Chrome OS has no such baggage; instead, it has the shiny positive reputation from Google that will give warm fuzzies to the masses of netbook buyers. And, because Google learned so well from Microsoft's tactic of stealing wind from the sails of your competitors through timely pre-announcements (in Microsoft's case, for products that were only illusions meant to torpedo a perceived threat), those buyers are going to skip paying the Microsoft tax for Windows 7 and wait just a bit longer to get the free Google Chrome OS on their new netbooks, probably paying less for the hardware in the process because Chrome requires less resources.
Microsoft has a history of pushing out competition and pushing bloated and bad products down consumer throats. Steve Ballmer is the master of denial when real competition appears on the horizon, as seen in these two quotes:
Google's not a real company. It's a house of cards.
There can't be any more deep technology in Facebook than what dozens of people could write in a couple of years. That's for sure.
And of course there's the famous:
Developers, developers, developers!
He's right, developers are important. And if you only go to Microsoft conferences, it looks like all developers are using Windows. But if you go to any other developer conference, everyone is using Macs. I wonder if anyone has even told Ballmer this (who would want to? He throws chairs through windows when he hears bad news). By the time the board of directors replaces him, years of damage will have been infused so deeply into the company that I can't see how it will recover; if it does, it will be a long and painful process.
Windows 7 is going to take a big hit at the low end, where a lot of people are making their first entry into computing. "Windows 7 sucks less than Vista" is not going to be an effective campaign for people who are looking at paying an additional $150 on top of a machine that costs a bit over $200, and by next year could be under $200 to the point where the cost approaches double to have Microsoft (and just the OS, not Office). No amount of pitching can convince people that "a lot more money on top of your cheap netbook hardware" is more interesting than "free."
The experience is going to be a big selling point, too. Chrome OS will boot very quickly and get you onto the web in seconds. Even when it's asleep, Windows wakes up a lot slower than this, and people will see it, and the salespeople at the Buy More will discover that sales are easier when the customer sees a better user experience.
And we can't forget viruses. My brother makes about half his income cleaning viruses off of windows boxes. Everyone has heard about viruses; saying "Windows 7 has better virus protection!" (a different variation on "sucks less") is not exactly reassuring given Microsoft's long history of writing a virus friendly OS and not fixing it. Once again, Google's reputation will pay off -- whether Chrome is virus-proof or not people will have more confidence that Google will continue to improve it.
Once new computer buyers get started with Chrome OS, the baby duck syndrome sets in and you'll have a very hard time prying them away from Chrome and into Windows. No matter how understated Google's announcement of Chrome OS has been, it bodes very badly for Microsoft.
Chrome OS will need to connect to the hardware and control more than we're used to in a browser (the Chrome OS UI will basically be the Chrome Browser). We've seen Google creeping in this direction for awhile, with Google Gears to allow applications to more easily disconnect and reconnect to the web.
In order for you to be able to write applications that work on both the Chrome OS and the Chrome browser, the browser will need (sandboxed) functionality similar to the OS. The result is that you'll be able to write far more powerful applications without intruding on (or scaring) the end user. To do this in Flash, you have to install the AIR runtime.
I suspect that Chrome will take the same conservative approach as Adobe AIR did, adding OS connectivity carefully on an as-needed basis.
An interesting step on the "sandboxed OS functionality" front that hasn't attracted quite the degree of attention that it maybe should have is Google's "Native Client" project: http://code.google.com/p/nativeclient/ The system uses a re-targeted version of gcc/g++ and a partial implementation of the system libraries that is pretty straightforward for developers to use (no special compilers or setup required) and supposedly quite safe (not my area).
I once installed Chrome browser on my xp machine (still waiting for the linux version), run it and did not go to any web address. Then I run wireshark and was amazed with the amount of traffic associated with the chrome process. That's why i use firefox + Ubuntu linux. I will probably install Chrome OS out of curiosity, but I guess I'll stay using Ubuntu.
I think you left out the role of office from your analysis. I don't like Ballmer, but I think it is honest to say that the real advantage of Microsoft is not Windows but Office. It is true that many netbook users are only interested in social-networking, reading news, etc. However, a great number of users will need to read/write Office documents. And in my opinion, the replacements for Office applications that Google offers are still far away from providing one tenth of the quality of MS Office applications.
I know many people that will happily move away from Windows to Linux also on Desktops, but they just don't do that because "Open Office sucks". To not speak of other more complex applications: for example I use Matlab, and I very happy that Matlab is available for Linux too: but the GUI of Simulink for Linux is almost unusable, while the GUI for Simulink on Windows is very user-friendly.
So, to overcome Windows, Google needs to make a considerable effort in improving Office replacements to a point that people start considering dropping MS word. However, it is a lot of effort, risky effort, and I do not know if google is willing to take the risk. At the end, my opinion is that Google is still very far away from breaking Microsoft monopoly.
Google already has a set of office apps - GoogleApps. These are already actively promoted to corporations (with significant drops against existing TCO). ChromeOS is just the piece of the puzzle that allows us to almost see the whole picture...
> I know many people that will happily move away from > Windows to Linux also on Desktops, but they just don't do > that because "Open Office sucks".
I haven't had any version of MS Office at home for 4 years or so and haven't missed it once.
I actually use Open Office at work because things that don't work properly in Word (e.g. bullets) work perfectly in Open Office. That and it converts Word documents into pdfs much better than anything else I've tried.
> Re - "Windows wakes up a lot slower than this": > One of your main arguments in favor of Chrome OS is that > Windows boots up slowly. The old windows vista/XP do boot > up slowly. > > But you are wrong regarding Windows 7. Windows 7 beta > actually boots up in literally 5 to 10 seconds. You > need to fact check these things. > >  I used the Windows 7 beta for months. The fast booting > time is the most amazing part of windows 7 for me (because > it is by far the most visible improvement).
I think it's interesting that after years of denying that Windows is bloated and doing almost no housekeeping on the source, MS took a whole release cycle to almost nothing but cleanup on Windows. For once, I'm kind of looking forward to trying a new version of windows.
The real lesson here is that all these managers from MS that used to go around telling people that you should only focus on new features and not worry about cleanup were jackasses. It worked for a while but Vista has to be the clearest example of technical debt coming due ever.
The deeper OS connections really scare me. If MS has taught us anything it is that tying the web browser too much into the OS is a recipe for disaster. Yeah, in theory everything can be effectively sandboxed, but...
I don't believe anything of this. I do not believe that HTML5 will kill Flash, I do not believe that netbooks + ChromeOS will replace notebooks + Windows 7, I do not believe in thin clients and the extinction of desktop apps etc.
Honestly, I wished Bruce would blog about programming again...
A lot more is required from the web apps side in order to reach the same functionality and performance as native apps. For example, Google docs offers the basics, but it is no match for Microsoft Office. Same goes for other apps (Photoshop, for example).
Give it another 10 years, and perhaps with 4G networks we may start seeing web apps that are equal on performance and features with today's apps.
One thing is for sure: "As I've said many times before, the biggest problem with web programming (and most programming is either on the web or going there) is the lack of a consistent programming platform. The amount of effort and money wasted to compensate for browser inconsistencies is phenomenal.". I couldn't agree more. The solution lies in a Java-like virtual machine that allows lazy downloading of its components and the components of the application. Web applications are doing that, but they are slow and running inside a browser.
> Honestly, I wished Bruce would blog about programming > again...
Ditto that. At the very least, Bruce, get someone else to edit your stuff. Your last few postings have been largely incoherent and often self contradictory rants about subjects that change, apparently randomly, from paragraph to paragraph.
> The deeper OS connections really scare me. If MS has > taught us anything it is that tying the web browser too > much into the OS is a recipe for disaster. Yeah, in > theory everything can be effectively sandboxed, but...
Erik, Microsoft "tied" the two together by taking some IE function exports and moving them into kernel/user/gdi and moving some kernel/user/gdi functions into IE so that if you deleted the IE files Windows wouldn't work. It was purely an artificial move so that they could claim that Windows wouldn't work without IE (and thus it's not anti-competitive "bundling").
Regardless, if you think that ChromeOS is anything more than a kiosk approach, you've drunk too much Googlaid. (Sh*t, I thought I just made up a new term, but Google itself proves that "Googlaid" has already been used many times :-(