Sometimes an idea is just too obvious. In the last post here, I talked about solving problems in a general way and with enough flexibility and ease of use so that a user can bend things into whatever way he or she needs it. I found another great example.
When people ask what the best kind of setup CouchDB is used in, I have a hard time finding a straight answer. CouchDB’s power obviously lies in its replication abilities. But these can be used in various ways. The limits to the scale here are small peer-to-peer applications that are distributed all over the place and large scale, high traffic database clusters. This makes people sceptical: How the hell does this fit into a single database?
While going through my XTech talk for the last time before the actual presentation, my brain developed or connected the necessary synapses. Or something.
CouchDB’s replication solves the general problem of data synchronisation. And it gives you enough flexibility and ease of use that you can solve any specific problems you might have. Small or large scale. CouchDB does not care, it only does data synchronisation.