Ask YC: Anyone using CouchDB?
If so, what do you like about it? What problems does it solve for you? What does it not have that you really wish it did? How is this better than a traditional DB?
Link to project: http://code.google.com/p/couchdb/
Link to project: http://code.google.com/p/couchdb/
For what it solves for me: It's schemaless structure allows me to develop and deploy much faster, because I don't have to spread my data across tables and find the correct and fastest JOINs to get the data back. Online replication makes high-availability a snap, the same goes for load distribution. Offline replication saves me from DB syncing issues when there's no network.
Being alpha software, it does not yet have a full fledged permission and authentication system (though planned and prepared for, it is not yet implemented) and I really wish it did, but that's going to be a matter of time.
Is it better than a traditional RDBMS? That's the wrong question in general. It is another tool and if it fits your problem is for you to decide. I still use plenty of relational databases.
"Due to a current limitation in the Erlang/OTP filesystem driver a CouchDB database file can not grow beyond 2 gigabytes in size. This is not a limitation in CouchDB itself, but in the standard library it uses. This problem will be addressed in a future release."
This seems like a serious hurdle for major adoption.