Last week saw a steady stream of big announcements out of Google I/O. Of particular interest are the announcements relating to Platform as a Service:
API for Google Latitude: a programmatic way to interact with Google’s location tracking service and access location history. With Latitude expected to come under the Google Apps umbrella soon this will open up some interesting location based services for Enterprise at relatively low cost. I’ve been running Google Maps on my mobile with Latitude enabled for the last week or so and the background tracking is working well. All I need to do now is convince a colleague to enable it to and we can start working out some Enterprise use cases for this service.
App Engine For Business: a clear evolution of App Engine into the enterprise space. Roadmap plans for hosted SQL databases by the end of the year start to bring App Engine up to what Microsoft’s Windows Azure is offering. The lack of relational database support has always been a point of frustration for those of us “stuck in a relational paradigm” so it will be great to see this addressed.
Moreover, there’s indication of Java PaaS standardisation - this announcement outlines the upcoming ability to deploy the same Java web application across AppEngine, VMWare vSphere and the new VMForce from Salesforce and VMWare. This would imply an opening up of the current restrictions on the base Java classes that AppEngine supports and therefore hopefully an expansion to the kind of apps that can be run on AppEngine.
From a more general PaaS perspective Google has outlined some new offerings: Google Storage for Developers - which appears to compete with Amazon S3 / SimpleDB for storing massive amounts of data. But the most interesting is what’s also available of the back of this: Big Query and the Prediction API.
For me, Big Query is a killer application. From Google’s description it features:
- Speed – Analyze billions of rows in seconds
- Scale – Terabytes of data, trillions of records
- Simplicity – SQL-like query language, hosted on Google infrastructure
- Sharing – Powerful group- and user-based permissions using Google accounts
- Security – Secure SSL access
- Flexibility – REST APIs, JSON RPC, Google Apps Script
Business intelligence and data mining just got a lot more interesting! The infrastructure required to deliver this kind of analysis has always been beyond the reach of all but the biggest enterprises. Now, a small organisation can have the same analytical power as the big boys.
I’m sure there’s more to come…and I just wish there was more time in the day to try these all out.
Overall, this clearly indicates the inexorable shift of software development onto PaaS so that we, as application developers and integrators, can take advantage of the massive complexity and sophistication of these services, but only have to deal with them in a simple way.
