AlgART Libraries is a collection of Java libraries for processing arrays and matrices. It features63-bit addressing of array elements (64-bit long int indexes), a memory model concept (allowing data to be stored using different schemes, from RAM to mapped disk files), wide usage of lazy evaluations, built-in multithreading optimization for multi-core processors, and a wide set of image processing algorithms over matrices. Almost all classes and methods are thoroughly documented via JavaDoc.
MLPACK is a C++ machine learning library with an emphasis on scalability, speed, and ease-of-use. Its aim is to make machine learning possible for novice users by means of a simple, consistent API, while simultaneously exploiting C++ language features to provide maximum performance and maximum flexibility for expert users.
Pascal-P4 for Free Pascal and Delphi is a port of a Pascal compiler written at ETH Zurich in 1976. The source code of the compiler is documented in the book "Pascal Implementation" by Steven Pemberton and Martin Daniels. The purpose of this project is to make Pascal-P compilable by Free Pascal and Delphi while keeping the changes to a minimum and preserving the line numbering as much as possible.
QWebSockets is a pure Qt implementation of WebSockets, both client and server. It is implemented as a Qt source code module (.pri file) which can easily be embedded into existing Qt projects. It has no dependencies other than Qt. It features text and binary sockets, frame-based and message-based signals, proxy support, and strict Unicode checking.
Opensort is general purpose sorting software that aims to be a fast and easy solution for the sorting of large or small data-sets and data manipulation in general. It is still in the early stages of development and lacks many of the advanced features its commercial counterparts have. For the moment, it only provides a simple command line interface and a C library other programs can use to cover some basic needs for data sorting.
Capsule Tree is a general purpose, self-balancing tree data structure for large, ordered data sets. It is designed to provide the same characteristics as B-trees and B+trees, but built from the ground up for in-memory usage. In other words, there are no provisions for “slow” I/O cases. The original motivation for this tree was a better backend for memory managers. However, the end result was a new sub-category of trees.