Talk:Java Native Access
![]() | Computing: Software Unassessed | ||||||||||||
|
Moved for discussion
Moved from the article for discussion. The following appears to have been added against WP:COI and is unsourced. Perhaps some if it could be incorporated into the article if properly sourced. --Ronz (talk) 16:31, 8 August 2008 (UTC)
History
Fragments of the JNA project date back to a small shared stubs library originally written by Sheng Liang of the Sun Microsystems JNI team and demoed at JavaOne circa 1999.
Over the next seven years, Todd Fast (also of Sun) heavily modified Sheng's library to add full Win32 support, provide a friendlier API, and expand the ability to map native libraries, functions, and types to Java equivalents. He first published the project on java.net under the name of "Java Native Access (JNA)" in the fall of 2006. The name was intended to reflect both the similarities to and differences from the Java Native Interface (JNI), which is the standard (though relatively difficult) mechanism for accessing native code from Java.
Timothy Wall became the maintainer of the project in February 2007, and has contributed by expanding the feature set, adding linux and OSX ports, and making the library usable for more advanced function types.
In the first public release, cross-platform support was limited to Solaris and Win32 and there was only partial struct support for native calls. With the purported goal of providing a concise programming model for developers, the API took advantage of JDK 5's annotations and type safety features, though these features were later removed to support using JNA with earlier versions of the Java VM. Finally, as a result of Todd's work on the project at Sun, JNA is packaged under the com.sun domain, though most of the work on the project has since shifted to parties outside of Sun.