A powerful replicable mobilized database: CouchDB

CouchDB is the document database from Couchio, and it's been getting rave reviews for its easy peer-to-peer replication across multiple platforms, and its tiny size—less than nine MBs. Unlike a relational database, a document database doesn't enforce a schema, so you can easily add new bits of information as your app evolves.

Already CouchDB has emerged as a great way to build task management apps, or anything else where sync is crucial. "The current app developer has to worry a lot about sync," explains Nitin Borwankar, VP of Business Development at Couchio. "With CouchDB, they just write HTML and CSS apps, and they're sync-enabled."

CouchDB has the potential to open up many possibilities for aspiring app developers, says Couchio CEO and CouchDB creator Damien Katz. "There are a lot of apps out there that just don't get built, because the difficulty of building them on a relational database is just too many man-hours," explains Katz. "One of the strengths of CouchDB is handling unstructured, document-oriented tasks. If you try to do that in a relational database, you spend all your time developing the app, and very little of your time actually getting to use it."

More info:

[gorillawack]

CouchDB allows replication of data peer to peer, or with the cloud, with a low file size 9mb install. No Schema is enforced, it is a noSQL database.
program in html/css/javascript/json and let the synching do its work. It is highly web scalable with regular http for its Api.

The view engine for querying and segmenting your document databases uses javascript logic which is more flexible and quicker than sql for pulling different data together.

the couchDB apps use very little battery life, and even a smaller memory footprint than the android keyboard at times. Resource usage wise its very efficient. Erlang was originally designed for phone switches, which is part of this efficiency on light weight hardware.

Moving cross database platform vendor for redundancy is easy, simply install couchDB and pull the previous data base as source, across geographical regions, as is done by bbc.co.uk using couchDB to replicate across geographical regions. Users maintain speed based on server/database location. Compare to google apps where dealing with transfer is an issue. CouchDB gives you this vendor independence. Like moving from amazon to rackspace for example.

couchDB is mesh based, so if the main hub goes down, the spokes can still communicate with eachother according to the data they store relevent to them. When the host is back up, they can re-sync.

databases can be selective in replication, only those that you are interested in seeing or accessing to save space and control user access.

[GraemeWolfendale]

via mapreduce which runs javascript code against the database on the server, for example looking for a particular keyword, as far as I understand it.

[sztoor]

One question ... if it works at document level then how you make the queries at different levels?