Google's new App Engine will let application developers outsource hosting and data storage for their applications by using key elements of Google's infrastructure.
As many people have noticed, the announcement just screams out for analysis in light of Nick Carr's new book The Big Switch.
]]> Carr outlines the history of electric power generation moving from an in-house operation of every business to its current position as a commodity produced by giant specialized power producers and sold at a metered rate. He argues that computing is undergoing an analogous process and that just as commoditized electricity changed the world, so too will commodified computing. The same industrial history has had to struggle with monopoly power, though, and that's something that has to be considered when looking at announcements like Google's App Engine.Richard MacManus provided an overview of the announcement last night and John Musser at Programmable Web provides a good overview of the technical details. There's an official Google video intro on the right. Most interested developers have already jumped in and looked at as much as they are able to access, though - so now it's time to talk context and consequences.
Scalability
Services like App Engine may very well represent history's next logical layer of abstraction, taking several onerous obstacles off of the to-do list of application developers. That means developers can focus on other things and leverage greater resources than they may have had access to otherwise. Google-sized economies of scale can beat just about anyone on price and in theory it bodes well for uptime. The all-to-frequent downtime experienced by customers of Amazon Web Services begs the question of Service Level Agreements for the Google App Engine - and there doesn't appear to to be any right now. Let's presume that all of that is going to get worked out in time, though. As Carr points out in The Big Switch, the early days of commodified electricity were also filled with worry about safety and reliability. There was an economic imperative to solve those problems, and except for externalities like damage to river ecosystems, asthma from coal plants and nuclear toxification of Native American land, the problems of safety and reliability have been solved in for electricity. Those aren't off-topic matters, either, see Carr's own coverage of Google's new giant server farm here in Oregon, where they want cheap electricity from our rivers to power the apps in App Engine.
Carr says in his book that progress towards technological advancement is more an economic imperative than it is a psychosis or a choice. See the above externalities and ask yourself about psychosis, but point taken - things like commodified computing make so much economic sense that they are in all likelihood inescapable.
Data Portability
It's up to you to decide to use Google Accounts as user authentication in your App Engine app, or to write your own authentication code. Developers tell me that porting these apps out of App Engine should require as little as rewriting some data storage code and perhaps a few webservices - but I'm not hearing people complain about the system being to confining. That's great.
It's very, very important that there be no barriers to leaving App Engine and that the service retains customers based on price and superior service. Anything else, any lock-in, will drive a stake through the heart of innovation.
I expected to hear about OpenSocial support for App Engine, but haven't yet. Hopefully standards discussions have been underway for some time, but it appears that other than being written in a single coding language - Apps will be pretty free to come and go.
It's not ala Carte
Amazon Web Services lets developers pick and choose between a handful of different services, including data storage and processing. Many startups, for example, just use AWS storage and do processing on their own servers. Google App Engine is an all-or-nothing offering and that's a little bit creepy. There may have been factory owners in the early days of commodified electricity who wanted to keep some parts of that world in-house too, so perhaps Google's all-in-one offering will seem reasonable in the future.
Advertising
The fact that Google is now offering to power and host web apps, many of which are only ever monetized by AdSense, is...convenient, for Google. Will other ad networks be allowed on Google App Engine servers? That Google is ultimately in the business of advertising against content and now will sell you a printing press may be nothing more than fair, but it certainly raised concerns about monopolistic conglomerates.
Google's dominance of online advertising is so severe, and the umbrella for innovators that such dominance affords is so large, that it can't help but raise concerns about a single corporate allegiance running between so many development teams in leadership positions in the web 2.0 economy. Now the App Engine feels a little like the minor-leauges, or a place for innovative sharecroppers, all brought together by the Google Dollar and the resources it can provide.
Competing Services
Competition is good. Amazon Web Services already faced some competition from services like Longjump (our coverage), Nirvanix and other cloud computing services (see Larry Dignan's discussion yesterday). Adding Google to the mix could turn up the volume on competition for customer loyalty.
While Google App Engine is only being previewed right now, there's already some people asking wether comparing it to Amazon Web Services is a matter of apples and oranges.
Bret Taylor, the former Google App Engine Project Manager and now founder of red-hot startup FriendFeed, couches his questions in some very nice words about the project.
I am impressed. The App Engine team has done a fantastic job, and I think they have already changed the way I do hobby projects.The next logical question is: would I run a real business on infrastructure that is so different than everyone else's? If I change my mind about App Engine, what are my options? I am hoping a number of open source projects spring up as alternatives to lower the switching costs over the next year. I will be very interested to see how many startups take the leap and run on App Engine entirely in the meantime.
Taylor posted about his experience testing out App Engine on a lightweight blog that he wrote on the platform itself.
Taylor isn't alone in saying that so far, Google App Engine is a relatively simplistic offering. There's a substantial number of people feeling unsure about Python being the only language being supported at launch, as well, though Google says that other languages will be supported in the future.
The point is, Google App Engine may be neither competitive nor monopolistic - it might just be trivial as Google Pages or Google Base.
So far it seems pretty simple and useful, though. We'll have to take a deep breath, hope that Amazon and others step up their offerings a notch or two in response, and see what the future brings.