Today Couchio, the corporate backer of the Apache CouchDB project, is heralding the momentous release of CouchDB 1.0. If you don't know already, CouchDB creator and Couchio CEO Damien Katz will tell you that millions of people already rely on CouchDB for non-relational data storage. The new release delivers several major new features including 300% faster writes.
Document-oriented CouchDB has proved its mettle quite well in production environments despite the fact that it has been in beta status and hadn't reached version 1.0. When DZone interviewed Damien Katz in May, he said that they just had to polish some of the APIs. Katz says the core storage engine has been solid and reliable for a long time.
For ease of development, the new release of CouchDB has a new authentication system that lets developers write Couch apps without having to make a user model first. This is all automatic in 1.0. You won't have to define SQL tables for users or write any potentially insecure code or plugins.
CouchDB finally has support for Windows machines, which is probably the biggest change in 1.0. Compared to the previous release, writes are also 300% faster for large documents. Finally, there are replicator options in 1.0 that facilitate system customization through replication.
CouchDB stores data using JSON and can be queried using HTTP. Couch apps do not require a middle tier, so instead of a browser, Java stack, and database, you just have the browser and HTTP. An implementation of Google's MapReduce is used for queries. Right now a version of CouchDB is being developed for Google Android. You can harness CouchDB's offline replication capabilities right now on Palm WebOS devices.
Now, I leave you with this hot Rap video about using CouchDB: