If you require the fastest database access off optical media or the Web, why not consider the custom Photon indexing and search system?

Most relational and non-relational database environments used today on mainframe, mini and micro platforms rely on binary-tree indexes for fast location of key data. The binary tree index design was first discussed in a paper by Bayer and McCreigh in 1972 and has since been modified for optimal performance within the limitations of its original design.

Photon does not rely on binary-tree or hashing methods. Photon's methods are better than all standard indexing methods, providing superior speed of search and retrieval, with extremely low-storage requirements. Under network conditions, and across remote links, our search speed remains consistently high where others falter.

For random searching of data in large indexes, Photon is 20% of the average size of conventional indexes such as those used in Oracle, DB2, Informix, Sybase, SQL Server, and all lower end database products such as Access.

From a speed perspective, Photon is 225% faster than the average of conventional indexes.

For locating identical search strings within an index, Photon is 18% of the average size of conventional indexes and 1,593% faster.

Photon has been used on a number of products, including Canadisk (Encyclopaedia Britannica 1989-97) and the Government of Canada's world famous Termium CD-ROM reference tool.