Enzyme is an architecture for creating user-searching-user applications. Right now, it is primarily useful as a resume create/search and HR management system. However, it's potential uses are wide-ranging, and it offers PHP developers an alternative to relying on simple "keyword" searching systems.
The OpenCA Project is a collaborative effort to develop a robust, full-featured and Open Source out-of-the-box Certification Authority implementing the most used protocols with full-strength cryptography world-wide. OpenCA is based on many Open-Source Projects. Among the supported software is OpenLDAP, OpenSSL, Apache Project, Apache mod_ssl.
TOra is a Toolkit for Oracle that helps the DBA or developer of a database application. It features an SQL worksheet with PL/SQL block parsing, a PL/SQL debugger and editor with syntax highlighting, a Schema browser, Unicode support, printing, and a full suite of DBA management tools.
QSD is a *nix daemon for administering game servers (like Quake, Half Life, etc). It uses a flexible plugin architecture for both authentication and for interfacing with the individual game servers. QSD has been designed from the ground up to relieve many problems game server admins deal with from day to day.
XMLDB uses an RDBMS to persist arbitrary XML documents. Due to its storage mechanism, searching for and recalling documents is extremely quick. You can also perform XSL translation on documents with surprising speed. The library can be used in any program to store libxml2 documents. A PHP module is also included, making XMLDB into a complete three-tier Web application development suite.
PostgreSQL BIND-9 SDB interface is an enhanced version of the original pgsqldb interface which is included with the BIND-9.1.3. This patch allows you to store all your zone data in a PostgreSQL database rather than in text files. This even makes changes visible immediately without restart/reload of the named server. It also includes PL/SQL functions and triggers to simplify reverse updates.
GNU EPrints is a solution for sites wanting to set up archives. By default it is configured to be an archive of research publications but can be easily altered. It is scalable and robust; all metadata is stored in UTF-8 rather than ASCII, and all "phrases" the system uses in user interaction are stored in an XML file for easy editing or translating. The system is built around Apache, mod_perl, and MySQL. Submissions are made via the Web and not entered into the main archive until approved by an editor.