2009-07-13
Google announced its Chrome operating system on July 7 and developers have been weighing in on what it means for the developer ranks ever since the company said it was building an OS based on its Chrome browser.
Google announced its Chrome operating system on July 7 and developers have been weighing in on what it means for the developer ranks ever since the company said it was building an OS based on its Chrome browser.
Developer reaction has ranged from excitement over Google adding energy to delivering Google for the masses, to indifference over the Chrome OS being Yet Another Delivery Mechanism that doesn't really impact developers’ use of the Web as the real platform for development, to skepticism.
Indeed, one prominent developer, who asked to remain anonymous, shared what seemed to typify the skeptical developer's view. Said this developer: "Is this just another door into Google's roach motel strategy, where they let me in easily but make it impossible to get out? In other words, the question I ask is 'what does Google get out of this?' And I'd suspect that their motives are not exactly pure. Altruism has its limits, and at the core, Google is just another company whose primary motive is to preserve and grow Google."
However, Kevin Hakman. an Asynchronous JavaScript and X M L (AJAX) expert, said, "The browser as the foundation for an OS makes great sense and goes hand in hand with Internet connected devices and phones. Web-based software apps just keep getting better and better and there are millions of web developers who will be able to create more and more web apps for these devices. It's going beyond a 90/10 kind of thing. Today, 90 percent of what I need to do I can do in a browser. An OS with a browser at its core will move that towards 95/5 with the five being left to things like Photoshop, video editing and 3D gaming for the traditional desktop OS."
Meanwhile, like some others, Mik Kersten, creator of the open source Mylyn project and CEO of Tasktop Technologies, said he sees the chrome OS as a bit of a "back to the future" move. Said Kersten:
"On one hand, it’s history repeating itself from the Network Computer that Oracle and Sun were pushing in the 90's. On the other, the current combination of HTML and AJAX in web apps has been validated as a way more successful programming model to what existed on the thin clients of the 90's. And with netbooks needing to offload computation to the cloud, due to their limited processor and power limitations, this seems like a sensible business move if Google can pull it off."
Andrew Brust, chief of new technology at Twentysix New York, a consulting firm, tended to agree with Kersten. "It appears that Chrome OS will be little more than a Linux ‘distro’ with a tightly integrated Chrome browser, and that the apps it will run will be the relatively simple applications in the Google Docs suite, and ostensibly other Web-based applications that conform to specific APIs," Brust said.
Moreover, Google Gears will provide offline cache/storage capabilities, and everything will continue to run on a platform – the Web -- that was designed for rich documents rather than actual applications, Brust said.
However, "the various technologies that allow apps to run in the browser are essentially hacks and don’t constitute a stable, cohesive platform, much less an operating system," Brust said.
"And even if people think the browser as a platform provides 'enough' for 'most' users, the reality is that people rely on at least a few client-based applications and they need the kind of device/display/printer/other peripheral compatibility that Windows offers," he added. "So this really looks like yet another Google side show, that’s not aimed at the enterprise, and it won’t even be out until the end of next year. We already tried this in the 1990s with the 'NC' (Network Computer) and Netscape’s campaign to be the next OS. It didn’t work then, so where’s the threat with netbooks and Chrome now?"
For his part, Kersten said he believes one plus for developers from Chrome OS is that the Google “stack” will become more of a consideration. Google could make it easier to write Google Web Toolkit (GWT) applications, leverage Gears on the netbook for applications that need some offline persistence, and provide developers with other technologies that allow them to build full-featured applications that only require the Chrome OS to run.
"But what confuses me is that, for any value that Google adds to Chrome and the Chrome OS, they would fragment web developers even further," Kersten said. "For example, if Gears runs better on Chrome OS, a web app developer would need to figure out whether to primarily target the Chrome OS instead of optimizing everything for the still dominant IE browser. There’s already so much pent annoyance with needing to debug and develop for every browser that developers could shy away from leveraging any special features Google adds to their 'stack.' To address this Google could invest more in GWT to insulate the developer from the particulars of the browser, and have that be an on-ramp to using more advanced features in Chrome/OS."
Carl Howe, an analyst with the Yankee Group, said of the Chrome OS announcement, "Google just made this market more confusing for developers except for one thing: if you develop an application for the cloud (i.e. an online application), it will work anywhere. And maybe that's the most attractive solution."
Howe said the Chrome OS presents developers with different challenges in terms of choice, implementation cost, and marketing for the applications. Regarding choice, the Chrome OS will mean that Google developers will have to decide whether to write for Android or Chrome. And at least for the initial implementation developers will make that decision based on whether they're writing for a netbook or a mobile phone, he said.
Moreover, regarding implementation cost, Howe said, "Developers overall now have to start thinking about how many times they want to implement the same applications. Apple already makes 90 percent of the code the same for applications that run on both Macs and iPhones. It's conceivable that Apple could introduce a netbook that ran iPhone apps natively, making that process even easier. Being able to develop once, sell on many platforms is really a holy grail for software developers and may sway their choices of platforms going forward."
And regarding marketing, Howe said, writing an application is just the start; someone has to sell it as well. "Google has not yet become a powerhouse of application distribution, while competitors like Apple have services like the iTunes App Store that are setting the world on fire," he said. "Again, developers have to weight whether they want to follow their platform allegiances or the money."
Dylan Schiemann, CEO of SitePen and co-founder of the Dojo Toolkit, said of Chrome OS:
"It reminds me somewhat of Palm's WebOS. I think it's currently a confusing message with regards to the differences between Android and Chrome OS. At the end of the day, it really depends what features they make available to web application developers. One of the reasons pure web apps aren't enough for the iPhone, for example, is that there's no mechanism for web applications to access features like the camera.
"PhoneGap does an interesting job in trying to fill that gap, giving web applications access to native features and then being installed and deployed as native apps wrapped in a browser. Microsoft's Gazelle also offers an interesting fit in this space, somewhere between what a traditional browser does and what Chrome OS proposes to become. If anything, the market is becoming much more interesting!"
Moreover, Brust said Google has shown that people love Web applications, and the machine- and OS-independence that they afford. The V8 JavaScript engine in Chrome, and its very notion of script compilation, shows that the Web can be a real platform and run-time, with optimized execution, just like Java and .NET, he said.
Further, the notion of a lightweight “un”-OS that boots quickly or, more simply put, the idea of a computer that’s just an extremely simple appliance, is compelling to people, Brust added. "Outsourcing complexity to the server is just fine by most people," he said.
Yet, Brust, who also is a regional director for Microsoft , said:
"Even if Google OS is not a commercial success, it will still be a very powerful proof of concept for what is an ‘aspirational’ product for many people: a computer that gets out of the way and lets people do their email, document, Web and social networking work without any friction at all, and at extremely low cost. The very fact that people yearn for this -- and I think they do -- should scare the stuffing out of Microsoft, whether or not this particular implementation is successful. When you add this threat to a growing list of others (loss of market share for IE, loss of consumer enthusiasm to Mac OS, loss of developer enthusiasm to LAMP, the uphill adoption climb for Silverlight and the near abdication of the smart phone market to iPhone, Android and Palm Pre), you can see Microsoft has a very formidable challenge to its market position. And I think Microsoft needs a fundamental shift in order to respond. That can only come if they take Google’s announcement very seriously."
Microsoft regional directors are independent developers, architects, trainers, and other professionals who provide a link between Microsoft and the developer community.
John Cromartie, a developer at Mobomo, said, "Considering the shift in the market towards netbooks and the ubiquity of the web, it really makes sense to have a web-centric OS at the core of the experience. It will be a great thing for netbook consumers if they can get a stable, coherent, focused and well-designed OS for the kinds of computing people are doing with these devices."
However, "The web-app-only model for application development wasn't exactly a hit for the pre-App Store iPhone, but if Chrome OS offers a good API for building web apps that take advantage of OS and device features then it could be a nice platform to work with," Cromartie said. "I'm excited, but we'll just have to wait and see."
In other words, this is yet another validation of how far web app programming model has come, developers said. People will happily open their netbooks to use the big web apps like Gmail and Facebook, and Google could get some OS revenue out of Microsoft’s hands in the process. But unlike something like Android and the iPhone, which represents less of a shift in the programming model, the Chrome OS seems like a logical extension of it to devices that don’t need a full-blown OS, Kersten said.
However, citing the Apple experience, "we have a more recent bit of history repeating with that,” Kersten said. "Apple first proposed that developers simply build web apps for the iPhone. That failed. Now all the popular App Store apps are more akin, in terms of programming model, to their desktop app cousins, since they use the OS’s widget toolkit and resources. So as a developer, my main take back from this is that if Google wants the Chrome OS adopted by the developer ecosystem, they will have to be more clear about what that Chrome OS stack is, how it will enable developers to access local resources and hardware on the netbook, and how the browser and thin OS combination will provide a rich enough programming model to compete with Microsoft’s inevitable response."
Meanwhile, the ever-prescient Peter Coffee, director of platform research at Salesforce.com and former technology editor at eWEEK, linked back to a May 2003 eWEEK column he wrote describing what essentially has become known as the Google Chrome OS.
Yet, in a more recent post, Coffee brings the cloud into the equation, saying:
"If you set out to design an operating system today, I'm sure you'd treat access to the assets of the cloud as a first-class function -- not as a secondary feature of some optional application called a 'browser,' suggesting that the user is casually exploring cyberspace in hope of running across something interesting. Further, you'd design for a model in which the cloud is at least as interesting for what it does as for what it knows: a world of active content, not just static pages or even Web 2.0's interactive communities."
