Reference implementation

In computer science, a reference implementation is software that implements a standard or de facto standard and is viewed as a reference for all other implementations of the same standard. The reference implementation is often provided by the creators of the standard. A reference implementation is therefore characterized by a high level of standard conformity . Another important function of a reference implementation is that the standard manufacturers can practically check their theoretical ideas, which in turn leads to a higher quality and consistency of the standard.

Reference implementations should primarily implement the features described in the respective standard as precisely as possible. The usability usually only plays a subordinate role. Since they were not developed for use in practice, reference implementations usually have a limited range of functions compared to comparable software and have poorer performance , which can be noticeable in slow processing speed, disproportionately high memory requirements and crashes .

Examples

• The Amaya web browser was developed by the World Wide Web Consortium from 1996 to 2013 and was used to examine web standards for their practicality even before they were implemented by other web browsers.
• The 7z file format and the 7-Zip packing program , which can be used to create archive files in this format, are reference implementations for the free Lempel-Ziv-Markow algorithm developed by the same author since 1998 .

See also

• Reference architecture
• Reference model

Individual evidence

1. ↑ Dirk W. Hoffmann : Software quality . Springer, 2008, ISBN 3-540-76322-8 ( limited preview in Google book search).
2. ↑ Berthold Daum: Java 6 . Programming with the Java Standard Edition. Addison-Wesley, Munich 2007, ISBN 3-8273-2468-8 ( limited preview in Google book search).