RethinkDB 1.4 (Some Like it Hot) is out. This release includes an improved wire protocol, updated drivers, support for query history in the data explorer, and a brand-new build system. It contains well over 100 bug fixes, features, and improvements (you can see the complete list of closed issues on Github).

Bill Rowan, engineer at RethinkDB, gives a talk introducing our open-source distributed document database– he discusses replication, sharding, writing queries with ReQL, and its internal architecture.

Michael Lucy, engineer at RethinkDB, gives a talk on how we built our query language (ReQL) as a DSL for languages like Ruby, Python, and Javascript and the challenges we encountered along the way.

Today we’ve updated the documentation page with a set of simple example apps that use RethinkDB as the backend database. These examples should serve as a quick way to figure out how to use stack X with RethinkDB. While very basic, the examples give a good idea of how to put all the pieces together.

So far we’ve published two applications based on the following stacks:

• A todo list built with Flask, Backbone.js, and RethinkDB (see it on GitHub)
• A simple Pastie app built with Sinatra and RethinkDB (see it on GitHub)

The code is available on GitHub under the MIT license. We really like the documentation format generated by rocco.rb, so we used it to annotate the source code. Documenting these apps has been a good exercise in eating our own dog food because we had to put the API to work and identify new ways to improve it.

Many people have been asking how RethinkDB differs from MongoDB and other NoSQL systems. Our first attempt to address these questions is a high level technical overview comparing RethinkDB with MongoDB. However, the overview is meant to be impartial, and it omits some of the more interesting bits about what makes RethinkDB special (such as our irrational love for Dota). In this post we wanted to describe the product from a more personal perspective.