Arandomd is a network daemon that provides statistically tested output from a single ARC4 stream cipher generator to anyone able to connect to its listening TCP port. As multiple consumers cause it to reseed itself from a separate, strong random source, it is expected to produce consistently unpredictable results for cryptographic purposes. A configurable number of tests from the Statistical Test Suite for Random and Pseudorandom Number Generators must be passed before output data is sent to the client. It may be be useful in low-entropy environments, such as compute clouds, for generating high quality key material for strong cryptography.
directfb-lua is a Lua module which allows use of the DirectFB library. It is automatically generated from sources and is designed to maintain the DirectFB syntax, adding all the power from the Lua language. It eases a lot of DirectFB development and comes with a bunch of examples.
Piglet is a tool for parsing and lexing text for the .NET framework. The purpose of Piglet is to provide an easy-to-use tool for parsing text which can be easily included in any .NET project as a single assembly. In contrast to most parser generators, Piglet provides a fluent interface which enables you to express your grammar in a syntax which is accessible for users with no prior experience of parser generators. Piglet generates efficient, type safe, and reentrant LALR(1) parsers at runtime, which saves you from having a pre-compile step to generate your parsing tables. It also includes a lexical scanner generator which can be used independently of the parser generator.
giflib is a library for rendering and generating GIF image files. This is the GIF service code in C used for over two decades by graphics applications, Web browsers, game consoles, ATMs, and pretty much anything else that throws pixels on a display. It is simple, stable, and bulletproof.
libsas is a Python library to take the pain out of using SaaS APIs. It provides an abstraction layer on top of various APIs, taking care of constructing the URLs, serializing parameters, and authentication. You just call Python methods and receive Python objects. It's like an ORM for SaaS.