fis-gtm 6.3-003A-1 source package in Ubuntu

 GT.M is a database engine with scalability proven in large real-time
 transaction processing systems that have thousands of concurrent
 users, individual database file sizes to the Terabyte range (with
 virtually unlimited aggregate database sizes). Yet the light
 footprint of GT.M allows it to also scale down for use in small
 applications and software appliances (virtual machines).
 .
 The GT.M data model is hierarchical associative memory (i.e.,
 multi-dimensional array) that imposes no restrictions on the data
 types of the indexes or content - the application logic can impose
 any schema, dictionary or data organization suited to its problem
 domain. (Database engines that do not impose schemas, but which
 allow layered application software to impose and use whatever schema
 that is appropriate to the application are popularly referred to as
 "document oriented", "schemaless" or "schema-free" databases.)
 .
 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.
 .
 This metapackage always depends from the default fis-gtm version.

 GT.M is a database engine with scalability proven in large real-time
 transaction processing systems that have thousands of concurrent
 users, individual database file sizes to the Terabyte range (with
 virtually unlimited aggregate database sizes). Yet the light
 footprint of GT.M allows it to also scale down for use in small
 applications and software appliances (virtual machines).
 .
 The GT.M data model is hierarchical associative memory (i.e.,
 multi-dimensional array) that imposes no restrictions on the data
 types of the indexes or content - the application logic can impose
 any schema, dictionary or data organization suited to its problem
 domain. (Database engines that do not impose schemas, but which
 allow layered application software to impose and use whatever schema
 that is appropriate to the application are popularly referred to as
 "document oriented", "schemaless" or "schema-free" databases.)
 .
 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.