Deploying to the Edge with CouchDB
Standalone CouchDB applications can be deployed directly to the client. This allows us to harness the power of full-stack view-source, along with the freedom of peer-based replication. We're building a vibrant ecosystem of code, developers and users. I'll share some of the challenges and implications of deploying to the edge of the network, and maybe even change the way you think about the web. I will focus on the affordances provided by CouchDB's replication, for building applications that take advantage of portable code and data. As a result, users can run powerful web-apps on their local machines (or remote CouchDB nodes) without any change in code. View-source on standalone Couch Apps gives users access to "server side" components as well as browser code. Accessing the full application stack locally has positive repercussion for open source. More importantly, it enables network effects as people build and share CouchDB applications. There are important constraints to follow when you're building apps that can deploy to any running CouchDB instance. I'll demo the CouchApp framework, which makes it easy to work with the grain of CouchDB, ensuring that your applications are portable across nodes, as well as scalable over large clusters. Applications deployed to the edge don't have to run in a vacuum. I'll also describe how CouchDB replication flows can be used to setup publishing, messaging, and other social platforms.