The gist of it seems to be that each file is stored as a series of binary chunks inside RethinkDB. [1]

[1] https://github.com/internalfx/regrid-spec#chunks-record

The basic idea is taken from GridFS. But we've reworked it from the ground up for RethinkDB.

The spec is also currently out of date. It needs to be updated to match the current state of development.

Mostly, just method names have changed.

This is the sort of interesting technology that RethinkDB makes possible. I'm the next months/years I think we'll see more projects that use Rethinkdvs scaling and realtime capabilities to make some very useful tools.

One nice thing about GridFS is that the spec is maintained by MongoDB and consistently implemented in all its major DB drivers.

Any plans to port this to other languages, or get the spec blessed by RethinkDB?

It is already ported to the .NET driver.

https://github.com/bchavez/RethinkDb.Driver

Slava Akhmechet is aware of the project, and Daniel Mewes has helped me adapt GridFS to RethinkDB.

https://github.com/rethinkdb/rethinkdb/issues/1523

Daniel from RethinkDB here.

The answer is yes, that's definitely on the radar. I can't give an ETA for the official drivers yet, but I hope it's going to happen relatively soon.

Out of curiosity, why would you prefer this sort of implementation to something more in line with MogileFS? [e.g. Metadata storage with the actual file stored independently on the local file system of multiple physical nodes]

https://code.google.com/p/mogilefs/