[
Lists Home |
Date Index |
Thread Index
]
> If a relational database has the capabilities to store XML documents without
> knowing the structure apriori, and lets you query (using XPATH and/or
> Xquery - not SQL), insert, update and delete nodes, subtrees or entire XML
> documents within and across all these XML documents stored in the DB,
There's little reason to compare relational databases with native XML databases.
A more valid comparison is SQL databases vs. XML data stores. First, there's an
ongoing debate about SQL and relational fidelity. Second, comparing
relational vs. object databases or native XML databases is a sleight-of-hand
used in marketing white papers -- compare against vintage technology instead of
referencing SQL products based on object-relational technology. (Today you
can develop SQL database plug-ins for types such as XML or video, just as you
can install a browser plug-in for reading Adobe PDFs.)
> I would put to you that you are not using a relational database, but what
Cache calls a post-relational database.
I concede the validity of an SGML content management system being described as
"post-XML" if it supports XML. However, that logic isn't the same as
describing COBOL as a "post-Java" language because there's a 2002
standard for COBOL.
The terms ISAM or hierarchical database have a basis in the literature and
industry acceptance, but "post-relational" is marketing-speak. There's no formal
definition. The intent is probably not to conjure up "post-set theory" or
anything about the formal foundation of the relational model.
There's also a problem of historical contradiction. Post-relational database
suggests something that emerged after Codd's research (and presumably
an improvement over the relational model). In reality, the core technology of
several "post-relational" products was developed years before the
seminal papers by David Childs (1968) and E.F. Codd (1969).
Cache, for example, is an extended implementation of M (ANSI X11.1-1995,
withdrawn in 2002). M is the '90s version of MUMPS, which was developed at Mass
General by Dr. Robert Greenes and company. It first ran in 1966 and was
presented at the Fall Joint Computer Conference in 1969. MUMPS was published as
a standard by NBS in 1975 and ANSI in 1977, years before the first SQL standard
(1986).
> Do any of the big 3 (Oracle, DB2, MS SQL-Server) do that yet?
Yes.
|
