> We want to fix the “Achilles heel of the cloud”, so that everyone has their data with them at all times regardless of Internet connectivity.

Isn't that what HTML5 offline storage and indexeddb are supposed to solve? Installing couchdb, spidermonkey, erlang and friends seems a bit overkill in my opinion, especially on a mobile device.

Yes that is what IndexedDB is for. We are working on BrowserCouch, a JavaScript library to "emulate" CouchDB on top of IndexedDB.

The big thing none of the HTML5 offline primitives provide is synchronization. Local storage is great, but in the HTML5 world it is still the application developer's responsibility to take data from the server and persist it to the client. For very simple use cases this is fine, but as soon as you want to allow collaborative edits to large data sets when users are disconnected, you'll end up implementing something that looks like CouchDB anyway. If you use the real thing (or BrowserCouch, when that is ready) you get synchronization for free.

Isn't that a bit "nobody else has tyres as good as ours, so buy our car?" Just write a good synchronisation library...

When are we going to stop grouping a million tools under the "NoSQL" moniker? By this standard, BerkeleyDB is NoSQL, it's a meaningless term. Redis is nothing like MongoDB, for example, yet they're lumped together.

PostgreSQL has a lot more in common with MySQL than redis and Couch...