Ideas (just ideas) list:
--------------------------------------------

Whenever it's done. Maybe utopic. This section is open (well, all others are, too :-)

- Across-connection transactions.
- Asynchronous logs.
- Configuration management without service restarting.
- Remote and asynchronous state management (HTTP, SNMP, CORBA).
- Remote and asynchronous configuration management (HTTP, SNMP, CORBA).

- More databases (MySQL, Oracle).
- Communication between daemons (1 r, 1 w). Maybe CORBA.
- Service redundance. State transmision. CORBA (SNMP?)

- Host several pools within a single daemon.

In 1.0.0 version:

- Full failure detection and recovery, maybe transparent to the user.
- External program to check and get in sync databases.
- Support every authentication method supported by PostgreSQL.
- Comprehensive regression test suite.

In 0.5.0 version:
- Troubleshooting section in documentation.
- Logging section in documentation.
- Recheck logging policy.
- Make logs cleaner when connections to backends succeed or fail.
- Define what to do with pending transactions (reset?) when serving a connection.
- Could this be the first BETA?



