In many ways Google I/O is the first, and to date, largest developer conference for the Internet. The attendance was impressive - over 3,000 developers were at San Francisco's Moscone Center to attend sessions covering a range of issues over two days.
Google firmly believes that the web has won as the future platform for application development. The company is now looking to woo developers to the Google cloud and at the same time make the Web a better place by adding more and more functions to it. To achieve this Google wants to make its cloud more accessible, keep connectivity to that cloud pervasive, and make the cloud's primary client (i.e. the Web browser) more powerful.
In other words it wants to make web browsers as, if not more, powerful and rich in functionality as its desktop sibling. That's easier said than done and means transforming web browsers from dumb information search and retrieval terminals to highly interactive application interfaces that deliver new levels of utility to end-users.
So what's Google doing to help‾ For starters it is inviting developers to mash-up Google services in their own applications. It's opening up a variety of Google APIs for content, search, authentication and so on that allow third-party developers programmatic access Google services (Gmail, Docs, Maps, Search, Picasa, YouTube etc) from the cloud in their own mash-up applications.
To showcase its client-cloud connectivity, Google gave several detailed examples at the conference. The first was the integration of Google Gears (a browser add-on in the Adobe Flash mould that allows for richer browser experiences) to improve search in MySpace email. Gears works by persistently storing, synchronizing and manipulating data locally in the browser, effectively allowing online applications to function offline.
Next it unveiled a new 'rent-a-cloud' pricing for its App Engine software that allows you to build web applications that run on Google's cloud infrastructure that is based on CPU use per hour - a model that sounds very similar to Amazon.com's Web services.
Google also presented a new version of the Google Web Toolkit which allows developers to build rich Internet applications in Java; the hosting of new Ajax libraries that lets developers improve the performance of their web applications using JavaScript tagging; and a strategy for transforming Web applications into mobile applications using a new SDK for its Android federated smartphone platform. The SDK basically provides WebKit views and allows mobile applications to be built using Web browser technology as the user interface.
Finally, Google also showcased interesting Web 2.0-like application development technologies for single sign-on for the Web (OpenID), universal authorization (OAuth) and a social network development 'standard' (Open Social). These technologies promise to connect web pages, applications and service to the 'sticky' social Web.
These diverse tools and technologies might seem to lack integration and to be targeted at different areas. In fact they're all cogs and wheels of a more meaningfully connected web that hosts Google Web services powered by the Google App Engine. Importantly some of these web services and applications aren't written just by Google, but by an entire market of independent developers.
These developers aren't just keen to build 'cool' web applications for the sake of 'coolness'; they also have an eye on tapping into Google's billion-dollar online advertising revenue stream.
Google's monetization strategy is simple. Invest in the advancement of the web by allowing users to do more on it, which makes the web a much bigger market for Google to monetize services like search. This is why we believe Google sees a pile of money in its web application development efforts downstream, even though for the time being it is focused on getting developers to build browser-based Web applications to thicken up its cloud.
Madan Sheina, Principal Analyst