pdf2htmlEX renders PDF files in HTML. It aims to provide an accurate rendering while being optimized for Web display. It is designed for scientific papers with complicated formulas and figures, so a precise rendering is the first concern, but general PDF files are also supported.
pdirl (PHP Directory Listing) is a tiny directory lister program. It is meant to be better looking and more useful than the directory listings produced by Apache. It features searching, directory listing in subdirectories, breadcrumb navigation, translations, and MIME-type detection. It requires PHP 5.
pdnsd is a Proxy DNS server for Linux and FreeBSD that is designed to cope with unreacheable nameservers (e.g. because the dial-in link is not up) in a graceful manner to prevent DNS-dependent applications like Netscape from hanging. It has a permanent disk cache (i.e. the cache contents are saved in a file on exit) and supports parallel query and a wide variety of link uptests. It also has the ability to serve some types of locally defined records.
PECoMaTo stands for Picture-Embedded COntents MAnipulation TOol. It is basically designed to display any kind of information embedded in picture files, as well as checking, filtering, extracting, removing, adding, and fixing such information. In other words, it's a metadata processor. It supports the following file formats: JPEG/JFIF, Adobe PSD and FFO, and raw IPTC. It knows about the following metadata formats: JFIF, IPTC, Exif, Adobe, and Fotostation. More file and metadata formats might be supported later (TIFF, etc.).
peekabot is a versatile 3D visualization tool targeted at robotics researchers and developers. It's multi-platform, general purpose, easily extendable, scalable, and does not depend on any other frameworks to function. It also provides some functionality for interacting with your robots using the GUI. You can test your algorithms in both simulated and real environments. The only requirement is that the device can use TCP/IP.
pg-python provides access to Python from PostgreSQL. It contrasts with the built-in procedural language by interfacing with PostgreSQL types rather than converting them. This allows large instances of data types to be handled with greater efficiency while also allowing reasonable support for arbitrary data types without the need for a specialized implementation. Procedures are managed as entire modules so that initialization can be performed naturally by a stored procedure (each PostgreSQL function has a "main" entry point). Import statements may exist in a more natural position, and the "main" entry point may be decorated.