Cameron Laird's personal notes on data management
The reference on data management I most use is
Jeffrey D. Ullman's
Principles of Database and Knowledge-Base Systems.
- RDBMS
[Explain ACIDity, SQL-92, ....]
- I'm still trying to figure out if
Caché
deserves attention.
- DB2 [offers single admin image for clustered servers]
- D3
- Gadfly:
Alex Martelli
literately
lauds Gadfly, and we also wrote
an
article on
its capabilities.
- Informix
- Ingres
- InterBase: Mozilla Public License 1.1; nicely mature,
with efficient query parser and seek algorithms;
not so hot for multi-user; available for Linux, Win*,
Solaris (not all same code?);
- MaxSQL combines
MySQL and Berkeley DB;
- mSQL
- MySQL: very fast at light loads; no transactions;
simplistic table-locking; not close to SQL-92; no
server-side cursors, no stored procedures, no SQL
views
- Oracle [explain good and bad]
- Paradox
- PostgreSQL:
not SQL-92 yet, also no declarative referential
integrity and outer joins until 7.0; shared cache gives
good multi-user performance. I'm thinking of writing a
long description of the prospects for replication.
- SQL Server [federates; relatively difficult management
for large installations]
- Sybase
- Titanium
- Visual FoxPro
LinuxCare has a serious-looking
comparison
of the leading RDBMSs. I haven't yet studied it enough
to vouch for its accuracy.
- LDAP
- in-memory and other embeddable
- db.linux
- dbm
- gdbm ... Bindings are available for many languages,
including
Tclgdbm
[Qgdbm]
- kdb is written in K
- MetaKit
- Microsoft thing
- TclVSdb
... an associated forms package called TclVSForm
...
Cameron
Laird's personal notes on data
management/claird@phaseit.net