Do-it-Yourself YouTube Uses Open Source Project Panda and Amazon EC2

For the do-it-yourself video software to do its work, you must register with Amazon's Web Services. Required are Amazon's Elastic Compute Cloud (EC2) and Simple Storage Service (S3). The Web Services provide storage and processing on a rental basis when combined with the corresponding software packages.

The Panda software comes free as Open Source and provides all uploading, transcoding and streaming services to the JW FLV (Flash) Player built into the web pages for playbacks (although you can switch to other video players). If Panda is bundled with the Amazon cloud, the software comes preinstalled with FFmpeg for video transcoding and supports formats such as FLV and H.264 for Flash Players and iPhones. Data for video, encoding, accounts and encoding profiles reside in the SimpleDB distributed database, while the uploaded and encoded video data is stored in S3.

On their project page, the developers describe the video integration using an "elegant" REST API as "completely painless," with implementation "in a matter of hours." The API documentation describes how to create, upload and clear the videos. Both YAML and XML are supported (with YAML recommended). Data is loaded into a separate window and a daemon takes over to encode the data in the required format. Panda then sends a message to the application that the video is ready to go.

Behind the open source project is the British firm New Bamboo, specialists in software development based on Ruby on Rails. As expected, you can integrate Panda in your Ruby on Rails applications; details are on a Panda webpage. The New Bamboo Panda source code is available for download under an MIT license at the github site.

Prices for the Amazon services depend on the actual monthly processing required and start at $0.10 for 1.7 GByte memory on a 32-bit platform. You can use Amazon's convenient monthly calculator to figure out the cost based on your anticipated usage.