Author Jon Olick. @JonOlick on twitter.
A Fast, Embeddable Clojure language implementation in C.
Just started the project, nothing is done, everything is messy. Will clean up before version 1.0.
The near-term goal is to re-make clojure in an embedable form which is as fast or faster than the JVM implementation.
The long-term goal is to add types/functions for machine learning, matrices, tensors, etc... And specifically try to do automatic parallelization.
- Native implementation. Can be (in some cases) as fast as or faster than original Clojure (which uses Java's JVM).
- Parses code into native structures, then executes. Essentially interpreted.
- Lazy sequences
- Startup time is ridiculously fast by comparison
- Implementations of persistent lists, vectors, maps (WIP: set, etc).
- Not entirely complete implementation (yet)
- Native types and operations for Matrices and Tensors.
- Focus on automatic parallelization in general, and explicit when "you know what you are doing".
- Software Transactional Memory
Unix:
Run make to build and make install to install to /usr/local/bin (might require sudo)
Windows:
Run c.bat