GT.M High end TP database engine (Sourceforge)
Description from the commercial version:
The GT.M data model is a hierarchical associative memory (i.e., multi-dimensional array) that imposes no restrictions on the data types of the indexes and the content – the application logic can impose any schema, dictionary or data organization suited to its problem domain.* GT.M’s compiler for the standard M (also known as MUMPS) scripting language implements full support for ACID (Atomic, Consistent, Isolated, Durable) transactions, using optimistic concurrency control and software transactional memory (STM) that resolves the common mismatch between databases and programming languages. Its unique ability to create and deploy logical multi-site configurations of applications provides unrivaled continuity of business in the face of not just unplanned events, but also planned events, including planned events that include changes to application logic and schema.
There are clients for node-js:
http://github.com/robtweed/node-mdbm
http://github.com/robtweed/node-mwire
Local topic map software is interesting and useful but for scaling to the enterprise level, something different is going to be required.
Reports of implementing the TMDM or other topic map legends with a GT.M based system are welcome.
I manage GT.M and would like to make a minor clarification to the words “commercial version” on this page. There are no “commercial” vs. “non-commercial” versions of GT.M. There is only one version of GT.M. The same GT.M software on x86 Linux is available under proprietary as well as FOSS software licenses. The same support on commercial terms with assured SLAs is available at the same price regardless of license.
Comment by K.S. Bhaskar — December 10, 2010 @ 7:04 am
I used “commercial version,” note it is a hyperlink, to indicate the source of the description, which wasn’t the Sourceforge site.
Thanks for the clarification.
Comment by Patrick Durusau — December 10, 2010 @ 7:44 am