IGD converts schema and data of component (relational, ASN.1, and tagged-ASCII) MBDs, into a common ACeDB format. Subsequently, the results of these conversions are converted to the global IGD schema also expressed in ACeDB (see below). The IGD global schema is expressed in a system-dependent DDL and therefore cannot avoid being affected by ACeDB system considerations. For example, certain features of the IGD schema reflect ACeDB limitations for large (approaching one gigabyte) databases rather than domain modeling requirements.
Conversions in IGD are expressed using rules that can be stored and maintained in a database. Most changes regarding the structure of component MBDs entail only updating these rules.
Schema and data converters in SRS and LinkDB take the form of parsers that are used for retrieving data elements from ASCII files. The SRS parsers are specified in an SRS-specific parsing language. Indexed data elements and their offset positions are extracted and stored in index files, with minor format changes. In a sense, index data can be perceived as having been converted into a common DDL for query and navigation purposes, but the original MBD ASCII file (or file libraries) are not disturbed. SRS and LinkDB do not support a global schema of their component MBDs.
Docking-D converts schema and data of component MBDs into
a common format supported by the object-oriented VODAK DBMS.\
Entrez converts schema and data of component DNA sequence
(e.g., GenBank, GSDB) and protein (e.g., PDB, PIR) MBDs into
local ASN.1 databases.
We are not aware of the conversion language and mechanism
underlying the Docking-D and Entrez converters.
We are not aware of any effort of ensuring the accuracy (e.g., information capacity preservation) of the schema and data converters involved in the systems mentioned above. Anecdotal evidence suggests that information loss occurs during some data conversion processes, but the causes and extent of this problem information loss have not been examined.