Free persistence software for linux

Mica is a system for building network-accessible multiuser portable applications. It is a programming language and object environment designed to be accessible by more than one programmer at a time.
It includes an implementation of a mostly pure object oriented language. Automatic persistence, reflection, strong encapsulation, multiple dispatch, and run-time security are some of its features.
Mica is designed with extensible "Collaborative Virtual Environments" (CVEs) in mind. In particular I intend to implement a highly-collaborative authoring environment within which users and agents can cooperate. Other subsets of CVEs include MUDs/MOOs or massively multiplayer online games.
Main features:
- Long-running / persistent and extensible applications. (Mica provides transparent persistence, meaning the objects that you use in your application will be there the next time you restart Mica.)
- Networked applications. (Mica is designed to concurrently execute many sessions at once)
- Collaborative applications hosting multiple concurrent authors. (Mica provides the tools to make sure the objects you create in a running collaborative application can be made secure and wont be overwritten by others)
- Rapidly prototyped software. (Micas prototype inheritance allows object relationships to evolve over time in a more intuitive way than many traditional class-based object oriented languages)
- Software with lots of objects with complex, evolving behaviours. (Micas use of multiple dispatch often makes modeling the actions an object can take much simpler and more intuitive.)

The Castor project has been developed out of need: the need to get stuff done and the need to write useful code.
It has been designed and coded over many cups of Dunkin Donuts Hazlenut coffee, Starbucks Latte, and tall no-whip White Mochas, as proof that coffee beans do come in handy.
Castor includes software, such as Xerces, Jakarta ORO, and Jakarta Regexp, developed by the Apache Software Foundation. Castor also includes JUnit testing framework
Main features:
- Castor XML: Java object model to and from XML (XML databinding)
- Generate source code from an XML Schema
- Default introspection or mapping file for existing object models
- Castor JDO: Java object persistence to RDBMS Castor JDO is not the same-as or compatible with Suns JDO. We have a different approach to handling data object to RDBMS mappings.
- XML-based mapping file to specify bindings for existing object models
- Support for schema-less Java to XML binding
- In memory caching and write-at-commit reduces JDBC operations
- Two phase commit transactions, object rollback and deadlock detection
- OQL query mapping to SQL queries
- EJB container managed persistence provider for OpenEJB
- Ability to create base mapping from existing Java classes
- Ability to create an XML Schema from an XML input document
Enhancements:
- Fixed problem with TransactionContext.getObjectEntry(Object) that was introduced as a result of adding support for lazy loading 1:1 relations.
- Added F.A.Q. entry describing compatibility problems between various releases of mySQL 4.1.x and its JDBC drivers.
- Reverted various DOCTYPE definitions in src/tests/jdo to use DTDs.
- Fixed issue with GeneralizedFieldHandler not able to handle collections. Collections are now automatically iterated over during getValue() calls and each item in the collection is passed into the convertUponGet method. To retain the old behavior simply add a call to #setCollectionIteration with a value of false in the constructor of your GeneralizedFieldHandler. The behavior for setter methods is not changed since only one item at a time is passed into the setter already.
- Fixed whitespace processing bug, where a was actually getting appended to the end of a string value if the parser was calling the #characters method again with only whitespace following a previous call to #characters method where valid content existed.
- Fixed recent issue where ElementDecl#getType() method was always returning null for element references.

subPersistence is an abstract, light-weight yet flexible framework for working with O/R persistence mappers. It provides for independence of a specific O/R mapper such as Hibernate or Castor.
While you can use Hibernate or Castor (or other O/R mappers) directly, coupling your application too tightly to those can be a bad thing. subPersistence makes it easy to switch underlying O/R mappers, thereby decoupling the application layer from those.
It abstracts away differences between the mappers, while still making it possible to use their specific features if needed.
Enhancements:
- This version updates to Java 5 features such as Generics and is the first version to require Hibernate 3.

KIO-POV is a KDE IO-Slave for the Persistence of Vision Raytracer (POV-Ray).
It provides an easy-to-use interface to POV-Ray for rendering scenes, viewing render results, render logs and information about POV-Ray files. For example if you want to render the scene /usr/local/share/povray-3.6/scenes/advanced/biscuit.pov (which comes with POV-Ray) you just open:
pov:/usr/local/share/povray-3.6/scenes/advanced/biscuit.pov?render
in konqueror (or in a KDE image editor if you want to edit the image) and you get the image rendered by POV-Ray.
Render results are cached so when you access the same render twice without modifying the scene you will not have to wait for the render to finish again.
Apart from the basic rendering mode which returns a PNG image you can also access information summaries about the scene in HTML form (see screenshot) and view the log file of the render. If POV-Ray reports errors while processing the scene the error message will be shown as well.
KIO-POV also supports MegaPOV. MegaPOV scenes are automatically rendered with MegaPOV if it is installed.
Note KIO-POV requires POV-Ray 3.6 or 3.6.1 to function properly. It is strongly recommended to install the latest official version from the POV-Ray website (both the source and the binary version will work).
If you want to report bugs in KIO-POV or otherwise want to contact me about it you can do so via the povray.tools.general newsgroup on the POV-Ray newsserver. For KDE-related questions the KDE mailing lists will also be helpful.
Enhancements:
- adding animation support (using ffmpeg)
- adding inline render previews in the file information pages
- fixing various bugs

Molins is a framework for the development of applications with PHP 5. Its more important features are database persistence, logging, unit testing, dynamic content generation (PDF, Flash, graphics, XML), security roles, GTK+ wrapper classes, and a complete Web MVC model (using XSLT). It is fully object-oriented, with class packaging similar to Java.
Enhancements:
- This version is focused only on bugfixes.
- The most important bugfix is related to the exception handler.

LiteSQL is a C++ library that integrates C++ objects tightly to relational database and thus provides an object persistence layer.
LiteSQL supports SQLite3, PostgreSQL and MySQL as backends. LiteSQL creates tables, indexes and sequences to database and upgrades schema when needed.
LiteSQL provides object relations, in addition to object persistence which can be used to model any kind of C++ data structures. Objects can be selected, filtered and ordered using template- and class-based API with type checking at compile time.
Main features:
- SQLite3, PostgreSQL and MySQL - backend support
- object persistence layer with relation support
- automatic database schema creation and upgrading
- create complex SQL queries using compile time checked class API; minimizes need to write SQL query strings

Demokritos is a server implementation of the Atom Publishing Protocol.
Demokritos also provides a library that can be used for the development of other Atom-related applications.
Usage:
To run a test server, cd to test directory and run test_server.py
client_test.py in the same directory will run through a sequence of tests on this.
Enhancements:
- Duplicates are now handled in URI minting.
- A feeds updated property after a removal of an entry is now current datetime (not that of the latest entry).
- Posts missing a media type now throw an exception.
- A bug where entry collections were being retrieved from the persistence layer as media collections has been fixed.

Open Project Services (OPS) is a project collaboration platform, integrating server components such as e-mail, calendaring, timesheets, documents, knowledge, etc. with standards complient clients. Open Project Services consists of a project server and a set of client extensions.
We are building a project collaboration platform. It consists of a project server that integrates with services needed within a project, e.g. e-mail, calendaring, timesheets, documents, knowledge, etc.
The project server is a Webservices hub based on Java (J2EE). The persistence layer for project information is LDAP. We use OpenLDAP now, but others should also be possible. The e-mail backend makes use of a standard IMAP server. Our reference implementation uses Courier.
We are building extensions/plugins for clients, making them OPS-enabled. We support e.g. Mozilla Thunderbird (XUL) and Squirrelmail (PHP) for e-mail clients. For the calendaring and timesheet components we have a Mozilla Sunbird extension (XUL).
In short: we are building a project collaboration environment on an OSS stack, enriching services with project information, using COTS clients/servers and open standards.
Enhancements:
- This release includes some changes to the underlying LDAP schema to make it possible to disable tasks when they are completed.
- To make use of this new feature, the Web services and the sunbird timesheet client have to be updated.