Noms is a decentralized database based on ideas from Git.
Noms is different from other databases. It is:
Content-addressed. If you have some data you want to put into Noms, you don't have to worry about whether it already exists. Duplicate data is automatically ignored. There is no update, only insert.
Append-only. When you commit data to Noms, you aren't overwriting anything. Instead you're adding to a historical record. By default, data is never removed from Noms. You can see the entire history of the database, diff any two commits, or rewind to any previous point in time.
Typed. Every value, dataset, and version of a database has a type, which is generated automatically as you add data. You can write code against the type of a Noms database, confident that you've handled all the cases you need to.
Decentralized. If I give you a copy of my database, you and I can modify our copies disconnected from each other, and come back together and merge our changes efficiently and correctly days, weeks, or years later.
Noms is supported on Mac OS X and Linux. While Windows isn't officially supported, you can compile a Windows build from source, and it usually works.
The build contains the Noms command-line and some utility tools. You can use
tar -ztvf noms-*.tar.gzto view the contents of the tar.
Extract the Noms commands.
tar -xzf noms-*.tar.gz
noms dscommand to connect to the
./noms ds http://demo.noms.io/cli-tour chi-building-violations chi-building-violations/backup chi-building-violations/raw ...<snip>.. sf-film-locations sf-film-locations/raw sf-fire-inspections sf-fire-inspections/raw sf-registered-business sf-registered-business/raw
View the history for the
./noms log http://demo.noms.io/cli-tour::sf-film-locations
Next, you can explore a Noms database or take a tour of the CLI commands.
Explore a demo instance of Noms
What Noms is Good For
Data Version Control
Noms gives you the entire Git workflow, but for large-scale structured (or unstructured) data. Fork, merge, track history, efficiently synchronize changes, etc.
An Application Database with History
A database where every change is automatically and efficiently preserved. Instantly revert to, fork, or work from any historical commit.
An Archival Database
Trivially import snapshots from any format or API. Data is automatically versioned and deduplicated. Track the history of each datasource. Search across data sources.
TODO: Sample and video
We are fairly confident in the core data format, and plan to support Noms database version
7 and forward. If you create a database with Noms today, future versions will have migration tools to pull your databases forward.
We're just getting started. Some important features are not yet implemented including a query system, concurrency, auto-merging, and GC.
The Public API will continue to evolve. Pull requests which represent breaking API changes should be marked with
APIChange and sent to the slack channel and mailing list below for advance warning and feedback.