> From: Stephen McConnell [mailto:[EMAIL PROTECTED]]
> 
> Can anyone provide a update/deescription/status and 
> differences between
> 
>    * Excalibur Datasource (which is used in Cornerstone) and in James
>    * Avalon-Sandbox Datasource

I started to have time that evaporated, but I wanted to build
a next generation version of Excalibur DataSource.  It was meant
to follow the javax.sql.DataSource and PooledDataSource semantics
much more closely.  I think I need to scale back from trying to
create a javax.sql.DataSource compliant implementation and just follow
patterns.

Excalibur datasource is already released and needs to be supported.
The Sandbox version is meant to correct some of the issues with the
older version:

* Interface assumes one to one relationship with DB and the rest of
  the system.  In the new version we have a DatabaseManager that
  will allow you to get a connection by name.

* Easier integration with javax.sql.DataSource implementations by
  DB vendors.  All the "big boys" have very well tested connection
  pooling code developed and tested.  We should be able to use it
  directly--without having to go through JNDI.

* Easier compilation--no need to distinguish between JDBC 3.0 and
  JDBC 2.0 because we will use dynamic proxies or PooledConnection
  style objects (references the connection instead of extending it).

* Pooling of prepared statements.  This is something that has not
  been done yet.

--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to