A Day in the Cloud, new articles on scaling, and fresh open source projects for App Engine

The latest release of Python SDK 1.2.3, which introduced the Task Queue API and integrated support for Django 1.0, may have received a lot of attention, but there have been a number of other notable launches and events since our last update. Several of these are summarized below.

A Day in the Cloud

On June 24th, Google Apps and Virgin America invited people from around the world to participate in a one-day online scavenger hunt called Day in the Cloud. Competitors were each given one hour to solve various puzzles and find answers to a host of trivia questions, and prizes were awarded to the top five scorers. The site was hosted on App Engine's scalable infrastructure. For more information on the challenge, including photos and videos of participants competing mid-flight, check out the official Day in the Cloud Lounge.

Scaling series

App Engine enables you to deploy and run your Python- and Java-based web applications on Google's highly scalable infrastructure. There are a number of ways to optimize the performance of an application running on App Engine, some obvious and others more subtle. With this in mind, we recently released a series of articles promoting tips and tricks that you can use to make best use of resources like memcache and prevent potential issues such as datastore contention.

We also updated the articles listing since we recognized that the growing number of articles added since April 2008 (over 35 by last count) made it difficult to find content on a specific topic. The new listing allows you to filter the list of articles based on a particular tag or label, so you can easily find all articles related to the datastore, for example.

As always, we are interested in your comments, thoughts and suggestions for new articles, so please feel free to share.

GeoModel and geomodel-demo open sourced

Google Earth API support engineer Roman Nurrik recently open sourced his GeoModel project. GeoModel uses geohash-like objects called 'geocells' to provide a generalized solution for indexing and querying geospatial data in App Engine. GeoModel is optimized for the basic real estate finder/store locator use case, but can be adapted for use with large datasets.

Using GeoModel, developers can instantly geo-contextualize datastore models by simply inherting from the GeoModel class. Currently, entities can be associated with a single geographic point and subsequently indexed and filtered by either conformance to a bounding box or by proximity (nearest-n) to a search center point.

Other open source developments

• GaeVFS: an Apache Commons VFS plugin which implements a virtual file system on top of the App Engine datastore, using the datastore and memcache APIs.
• GraniteDS 2.0, now with support for App Engine for Java: Granite DS is an open source alternative to Adobe Lifecycle Data Services for Java EE application servers, and supports both server push and the persistence mechanism on App Engine for Java. For more information about the project and a link to a sample implementation, see the announcement post.
• jiql: a JDBC-based wrapper of App Engine's low-level datastore API which provides distributed database access via an interface familiar to many developers.
• pQg: a proof-of-concept PHP-based SQL to JDO wrapper plus sample application.

Posted by Jason Cooper, App Engine Team

Java is a trademark or registered trademark of Sun Microsystems, Inc. in the United States and other countries.