When choosing a platform to power a new application that will be used by popular media sites, some of the criteria we look for are scalability, ease of deployment, and ease of management. As with the nature of news and media, traffic can go up and down without much warning and getting new content out of the door quickly is the name of the game. App Engine has greatly simplified development and deployment of YouTube Direct, a new tool built on YouTube's open APIs that allows media organizations to request, review, and re-broadcast user-submitted videos by embedding the upload functionality of YouTube directly into their own sites. By using the Google Plugin for Eclipse to "click and deploy" new code changes and App Engine's admin console to manage application data, our launch partners -- most of which had never used App Engine before -- were able to easily deploy their own instances of YouTube Direct.
To date, media organizations like The Washington Post, The San Francisco Chronicle, NPR, The Huffington Post and ABC's Good Morning America are using YouTube Direct powered by App Engine's Java runtime and we hope more organizations take advantage of the platform. Visit YouTube Direct's developer page to learn more about it. The project's source is also available on Google Code.
Here are some of the recent developments from the greater developer community.
Nick Johnson recently added a new module to the App Engine Python SDK which allows you to use the task queue service to execute deferred function calls. This library requires minimal configuration and makes it even easier to use tasks. Using it is as simple as calling deferred.defer with the function and arguments you want to call - for example:
from google.appengine.ext import deferred
deferred.defer(logging.info, "In a deferred task")
For more details, see the article.
The JRuby App Engine project has done a lot of work making spin-up time less painful with the most recent release. They've also had some great contributions from the growing community, such as an Image API that is ImageScience compatible. Ruby programmers can now build applications using familiar tools like Sinatra and DataMapper, while having access to the full set of App Engine APIs for Java. Follow future developments on the jruby-appengine blog and watch for talks by the JRuby App Engine team at upcoming conferences: "Scaling on App Engine with Ruby and Duby" at RubyConf and "JRuby on Google App Engine" at JRubyConf.
Open source efforts to provide alternate hosting environments for App Engine applications continue to expand, with the TyphoonAE project joining the existing AppScale project in providing a platform to run App Engine apps. TyphoonAE aims to provide a complete environment for hosting App Engine apps, and so far includes compatible datastore, memcache, task queue and XMPP implementations.
Juraj recently created and open sourced a library for the Java runtime which "generate[s] appropriate paging queries from a base query, and then using these queries to page through the data." This is intended to simplify iteration through all of the results in a very large result set.
If you love RESTful APIs then it would be worth your while to take a look at AEJ Tools. Their library provides "a server module which provides a Rest style access to your datastore data, and a client module which allows you to run queries remotely using the Groovy interactive console. It can be used to run queries, store new entities or as a data import/export tool."
The datastore related libraries just keep right on coming. Full text search is a highly requested feature for App Engine and those behind the open source appengine-search project have create a Python library which "[c]an defer indexing via Task Queue API. Uses Relation index strategy mentioned in Brett Slatkin's Google I/O talk."
Moving beyond the datastore, the developers at geewax have created an advanced testing framwork called GAE Testbed. It covers everything from isolated and quick unit tests to full scale end to end tests and builds on some of the great testing tools for App Engine which are already avilable.
In other news, the folks at techwalla have migrated their party chat app to App Engine and made the source code available. It uses the new XMPP API to simulate a chat room using your favorite XMPP client. You can read more specifics on their blog.
Check out the open source projects wiki page for more open source App Engine applications and utilities. Since the last posting, nearly 15 new applications have been added, so keep checking in as new projects are added regularly. If you have a project of your own to add, just follow the instructions at the bottom of the same page.
That's brings our community update to a close. We're always interested in hearing about new and interesting projects which use App Engine. If you have something to share, drop us a line.
Congratulations to Audrey Roy -- her "Price It By Phone" application won the Google App Engine + Twilio developer contest that ran from September 21st to October 4th. Audrey will receive $1000 in Google App Engine credit, and a Dell Netbook from Twilio.
Audrey hacked together her contest submission in under 48 hours after going to an art museum and discovering that the $60 box she wanted was only $37 on Amazon.com. She used the Amazon Product Advertising API to look up the best used and new prices for books by ISBN number, using any touch tone phone.
Look up Amazon.com book prices by calling (877) 265-8137. Enter the book's 10-digit ISBN number after the prompt. The voice will tell you Amazon.com's lowest new and used prices for the book. Then, to see the list of books that you've already price-checked, with links to their Amazon detail pages, enter your phone number at http://price-it.appspot.com.
You can find an interview of Audrey on Twilio's blog.
Thank you to everyone who participated in the contest.
Twilio runs weekly developer contests to encourage developers to explore the many use cases, technologies, and industries where voice applications can provide useful solutions. They announce a new category each Monday along with the previous week's winner. This week's category is "Twilio for Salesforce, with Appirio" and the deadline is midnight on November 19th.
Questions? Drop us a line at firstname.lastname@example.org