COMMENTARY

Peter Burgess

CHECK OUT THIS WEBSITE ... http://stigdb.org/

Stig is a new open source, non-relational, distributed graph database being developed here at Tagged.

Our goal is to make it easy for programmers to store and access large amounts of connected data in a way that keeps consistency, while still being fast and scalable. Designed for planetary-scale, with real-time social applications and web developers in mind, Stig is the ideal database for large social networks, real-time services, and cloud applications.

It's a graph. Stig stores data in nodes, and stores the connections between data as edges between nodes. This makes it easy to do Kevin Bacon computations and social networking queries that look for the connections between people.

Functional query language. Stig has its own native language that is flexible and powerful, but at the same time, it can emulate SQL and other paradigms.

Distributed and scalable. Stig is horizontally sharded, meaning that you can add as many machines to the system as you want. Stig works whether you have one machine, a thousand machines, or a million machines.

Points of View. Since isolation often comes at the expense of concurrency, Stig implements a kind of data isolation called a point of view. Private (one person) and shared (some, but not all people) points of view do propagate out to the global (everyone) database over time, but for fast communication, they only change the data for the people who need to see it.

Time Travel. Stig keeps a history of data as it changes, so you can easily go back and compare prior versions. You don't need to include history in your schema, and you can control how much it stores.

Durable sessions. Clients can disconnect and reconnect at will, while Stig operations continue running in the background.

© 2011 Tagged Inc. All rights reserved.