The Cloud Wars

This week, two major announcement have kicked off what I would call the cloud war: The announcement that Google will get into the OS business and the announcement that Google is launching its Google apps suite out of beta. Next week, at its Worldwide Partner Conference, Microsoft will stake its position when it comes to that new playing field.

A bit of history

In order to understand the importance of the current shift, one needs to study a bit of history. Since the dawn of the personal computer era, applications have been written and running largely on the user’s desktop. In the mid-90s, Sun Microsystems co-founder John Gage started claiming that “the network is the computer.” Marc Andreessen, co-founder of Netscape, the leading browser company at the time, was claiming that Netscape would “reduce Windows to a set of poorly debugged device drivers.”

However, due to limitation in terms of bandwidth and computer power, this vision didn’t come to be until well into our current decade. Today, individuals still mostly use Windows, even if most use it primarily to launch their web browser.

In more recent times, the availability of always-on, higher speed internet access, has allowed companies like Google to start offering more powerful websites, which took on features of full-fledged software applications. Leveraging technology that first saw the light of day in the 1990s (Flash was born in 1995 and XMLhttp, which powers AJAX applications was created by Microsoft in 1999), those applications started offering compelling competitors to existing products.

One the leader in that revolution has been Google. First with the release of Gmail and then with the release of Google Apps, the company has been working on offering online version of tools like email, word processing, spreadsheets, and presentation software. Leveraging its establish power in the advertising space, Google has figured that, by offering document and email management features to its users for free, it could create extra advertising inventory that it could then resell.

So Gmail, Google Docs, and Google Apps were born. Since they were consumer focused products, presenting them as products “in progress”, complete with a beta stamp and an advertising-based model. Jeff Jarvis warrants that such act was not only bourne out of humility but also as a  call to collaborate. This week, however, the company decided to shed the beta logo for most of its applications.

With its direct language to IT manager and its message emerging from the enterprise group, Google is making it clear that this announcement is not targeted at the consumer space. In a sign of growing business maturity (most software company attempt to appeal to the enterprise space as they get older and need to develop more predictable financial groundings), the company is now trying to appeal to the enterprise space, aiming its offerings towards a space that has traditionally been controlled by Microsoft (with its Office Suite) and, to a lesser extent, IBM (with its Lotus division offerings).

Poorly debugged device drivers?

But Google realizes that much of what it does is dependent on the continued goodwill of the different operating system providers and browser suppliers. Were it not for web browsers or operating systems, Google could not exist. Last year, the company started reducing that dependency by introducing its own web browser, named Chrome. Chrome was actually quite interesting in terms of browser development as it was the first browser to treat each window session as a separate application, ensuring that if one web page failed, the other tabs would not. This could be seen as something not completely unlike the way an operating system (or kernel, etc) doles out memory and CPU power to each of the applications it deals with and orchestrate who gets what.

The unstated strategic goal of the Chrome browser is to help reduce the dominance of Internet Explorer in the online space while providing Google with more of a say in terms of where web standards were heading (I’m sure some people will try to debate that point but, if Chrome is not intended as an Internet Explorer competitor, why is the only “official” version of the browser a Windows one, with no such offering on OSX or Linux?)

Chrome is not only an attack on Microsoft’s browser dominance in the web space but also an attempt at ensuring that neither Microsoft NOR Adobe get control of the future of web applications. Google’s CEO Eric Schmidt knows how trying to fight Microsoft can distract a company from very real threats by other unexpected contenders: he did come from Sun Microsystems and Novell before joining Google and saw, first-hand, how those two companies saw their focus on unseating Microsoft’s dominance in their respective areas blinded them to the threat that Linux came to be to both of them, ultimately dooming each of the companies’ efforts without Microsoft having to do too much.

So, as a veteran of the OS wars, Schmidt is now being careful in balancing its entry in the space. On one hand, he doesn’t want to offend existing partners like Apple and the open source community. On the other hand, he needs to ensure that his company’s offerings are actually going to appeal to hardware vendors. The OS will ultimately be little more than the minimum required to make the Chrome browser run. That means it will include an IP stack, some basic drivers to interact with the keyboard and screen (or a way for companies to offer those) and a UI that will be a full screen version of the Chrome web browser.

The description of the OS, as stated in the press release, describe it as such:

Google Chrome OS is being created for people who spend most of their time on the web… without wasting time waiting for their computers to boot and browsers to start up. They want their computers to always run as fast as when they first bought them. They want their data to be accessible to them wherever they are and not have to worry about losing their computer or forgetting to back up files… Even more importantly, they don’t want to spend hours configuring their computers to work with every new piece of hardware, or have to worry about constant software updates.

Put quite simply, this is a web browser with the basics to make it run online and offline (the offline components probably being based on Google Gears (already built into the Chrome browser) or some HTML 5 offline approach). Users will not really store much on their computer but everything will be sitting on Google’s servers, accessible from anywhere. Operating system upgrades will happen automatically in the background and everything will run in the browser. For those people expecting to run Firefox (or any other application) on this thing, sorry… it won’t happen.

Google’s view is that everything will run online and all data will be stored online. In technical terms, this is called sending information into “the cloud.”

However, there’s the question of how to plug components in there. I suspect that Google will lean heavily on its partners to release any device related drivers through the equivalent of an online application store, similar to the app store on the iphone, where Google controls the experience in terms of what gets installed on the user’s desktop and can recall or upgrade an install if needs be. The idea being that the hardware device does not need much power as most everything is coming from the web.

Developers will not be allowed to develop anything that runs on the machine itself:

For application developers, the web is the platform. All web-based applications will automatically work and new applications can be written using your favorite web technologies. And of course, these apps will run not only on Google Chrome OS, but on any standards-based browser

With these few words, Google is taking the same approach as Apple first did when itintroduced the iPhone: don’t look to us to provide you with any SDK, the web is the platform. Build your application using HTML 5 and all will be OK. This basically means that right now, Google either has no intention to provide an SDK or will keep it accessible only to select partners who want to integrate with their OS. They will first provide access to the device makers and then, over time, will create an SDK and an app store that they may even be willing to share with partners by white-labeling that store to sweeten the deal for any partner willing to install the OS.

The reason I suspect this would be part of the strategy is that pricing will not be a heavy deciding factor in whether partners will adopt the new OS and Google desperately needs the new OS to be implemented as widely as possible.

Many have said that cost was a large part of their strategy but I suspect it cannot be: Consumers have already been trained to consider the operating system as a freebie or low cost tool. On the windows side, consumers see the OS as something that comes with their machine, not something they buy separately. This effectively brings the price to 0. Even Mac users, who generally tend to be more willing to pay for products offered by Apple, were grousing at pricing on OSX, forcing the company to take a deeply discounted approach when offering the next version of its operating system for about the price of dinner and a movie. And pricing has proven to be a contrarian indicator in the netbook market, as consumers decided to pay extra for the Windows XP version of devices that also offered the same hardware at a lower price point with Linux.

Interesting timing

Having established that the company is looking to get more control of its end to end experience, one big question is why do it now? Why not do this, for example, at their developer conference, as they did for Google Wave? Why announce something that will not be available in the near term?

My suspicion here is that part of the reason for this vaporwave release is that Microsoft is about to unveil a series of cloud focused initiatives at its WorldWide Partner Conference next week: those offerings will include a major push for their cloud platform, Microsoft Azure, along with announcements regarding the Gazelle project (their own browser as an OS offering), and Office 2010, a substantially revamped version of the popular suite that will move collaboration and synchronization front and center. At its core, the revamped Office suite will not only include the existing components and features of older version but its guts will have been rebuilt with some DNA acquired as part of the acquisition of Ray Ozzie’s Groove Networks and its offerings.

I suspect that Groove and Ozzie have Google shaking in its boots. Much of Google’s strategic message is that it is more collaboration friendly than Office and, by leaving one’s documents on Google’s servers, one doesn’t have to worry so much about revisions and versioning. With Office 2010, Microsoft is fixing these problems and telling corporations that while Google’s message is nice, your proprietary information will be sitting on Google’s server. How about getting the same type of functionality but keep the documents on your own servers. Because most corporate IT department tend to be paranoid when it comes to their corporate data, the Microsoft message will resonate better.

So Google is not starting to position itself in the consumer market, hoping that applications which can run in the consumer world will eventually help tear down the corporate walls (to date, few corporations have adopted Google Apps and, if Microsoft offers a competitive product, I suspect it could remain that way for at least a decade). Having to do something, they have now decided to attack a core tenet of the Microsoft empire: its windows OS division.

The battle lines are now drawing:

  • Google is asserting that the world runs solely within a browser and all application logic is in the cloud; Microsoft will assert that substantial amounts of complex tasks require the power of the desktop and the cloud is there primarily as a tool for collaboration and synchronization.
  • Google is asserting that desktop PCs are merely thin clients; Microsoft is asserting that desktops are still the center of the computing experience.
  • Google is asserting that the net is safe enough a place to leave all your information; Microsoft is asserting its not.
  • Google is asserting that developers don’t want to run applications natively on a machine; Microsoft is asserting that the tightest integration happens at the OS level.

Each company is presenting a different vision of the cloud. I can’t say which is right as both offering compelling advantages and substantial flaws but I can highlight one important feature: in the future the software you are running will be connected to the internet most of the time and still be able to work when offline. And in that future, I suspect that the notion of software as a product you buy will probably disappear, with software as a rental model becoming the emerging approach. And I also believe that this is the beginning of the cloud OS wars.

Update: As expected, Microsoft sends out its reply.

About the Author

Tristan Louis

Writing and working on the internet since 1993, I've launched six companies, of which two went public and three were sold. This is my personal site and all opinions here are mine.

  • Michael Neubarth

    Good analysis Tristan. It’s a new version of the thin client vs. fat client model. Interestingly, while Microsoft will now argue that it is unsafe to store your data on remote servers in the cloud, that is exactly what Microsoft advocated when it launched HailStorm. A big question for Google Chrome is how good a product is the Google browser vs Internet Explorer and Firefox.

  • David W

    Very thought provoking, and spot on as always.

    The idea that the browser would become the OS? That Andreessen quote tells you how old it is. Kind of like my predicting “a big stock market crash” in 1995. No one would be calling me a genius. And the browser OS thing still hasn’t happened _yet_.

    Back then I knew the “browser OS” hype was a pipe dream for technical reasons. Networks didn’t have the capcity or the reliability. More importantly, it is a lot harder to do something on two computers with a big WAN in between, than to do it on one computer. Or in business terms :) it will cost a lot more, and not work as well. To a developer, writing a word processor this way sounds like the kind of idea only an old man in a suit could dream up.

    But the “Browser OS” term only sucks because it oversimplifies. There’s just a trend towards more complex client-server interactions in general. I think, because they’re becoming _possible_. Vastly stable, incredibly fast networks and banks of commodity servers are a good thing. :)

    Today Ajax and Flash have, in fits and starts, finally created what looks like a (barely) usable platform for developing on them, doing what Java had tried and failed to do. Note that Java failed for very specific, technical reasons – such as JVM startup time on 1990’s hardware. Though, I shouldn’t forget to give Microsoft’s successful sabotage attempt credit as well.

    Ajax is not new. People have been trying to design these platforms forever. Check out the NeWS framework (actually developed by long time “network computer” enthusiast James Gosling). Wikipedia has a good writeup (that I swear I didn’t edit before citing). ;) It seems clear Google is committed to evolving Ajax as a platform the way Sun has with Java. Their success so far is impossible to argue.

    The funny part is that I was wrong about web-based word processors. My revelation happened on the Obama campaign. Google docs was all they used. And it would have been impossible to run that campaign without it. Hundreds of thousands of strangers, with different computers and NO MEANINGFUL IT SUPPORT all flinging documents around and collaborating in real-time. That was simply impossible before Google.

    My guess is that these wars are going to keep being won and lost on the little technical details.

    For instance, it is _easier_ to write applications the way Google does it. Groove/Sharepoint requires that “MEANINGFUL IT SUPPORT” that Google docs does not. An expert will need to read manuals to install it. You will have to update it and patch it. For the percentage of admins who mess that up, it will go down and/or lose data. It will be expensive. It will fail with firewalls and break when you upgrade to Service Pack N. This may be exactly Microsoft will succeed in holding on to the enterprise for a long time to come. Big corporate IT departments love all that bullshit.

    Being free and open (or not) will contribute heavily. The marketplace has started to learn that lesson. We know who does that better.

    But you raise a big question: how many businesses will let their Secret Documents go onto a 3rd party’s servers? I actually have no idea. Maybe everyone will be afraid of google’s all seeing eye. I’m not saying they shouldn’t be. On the other hand, how many companies hire “3rd party” bulk document shredders?

    Oh, also, two nitpicks:

    When you say “This could be seen as something not completely unlike the task manager in windows, which doles out memory and CPU power to each of the applications it deals with and orchestrate who gets what.” I would put it slightly differently, “…not completely unlike the way an operating system (or kernel, etc) doles out memory and CPU power…”

    Also, I think it’s “Google Gears” plural.

  • http://www.tnl.net/blog/ Tristan Louis

    David,

    The substantial amount of support required by Microsoft tools will probably be offset by concerns relating to where the information is stored and instincts of self-preservation by most corporate IT departments.

    As for the corrections, thanks. I’m making them in the entry right now :)

  • http://www.hyperoffice.com Pankaj Taneja

    Wonderful analysis, and presents a very clear and panoramic picture of the Google and MS war, as well as specific battle areas. All that strategizing does certainly confuse you! One area you missed was Google’s recent launch of the Google Apps Sync tool, which it positions as a “Microsoft Exchange alternative” and will try to dent MS’s messaging monopoly, at least in the SMB segment.

    Google’s strategy has a hint of the “cool” factor, where the attempt is to make technology seem easy, accessible and hassle free (just pressing a few buttons is all it takes) while MS’s approach has to keep technology behind the IT department’s forbidding walls, behind which obscure tech language is used.

  • http://www.ciozone.com Ron Kost

    Really nice job Tristan of providing a basis for comparing Microsoft and Google’s approaches. We are currently discussing Chrome on CIOZone.com. For those who are interested in the conversation feel free to visit the topic thread in our TechThread Forum posts: Does the world need Chrome OS? = http://www.ciozone.com/index.php?option=com_fireboard&Itemid=431&func=view&id=1419&catid=116&limit=6&limitstart=6