Differences between Connection Pool managers in Weblogic 6.0

I'm trying to understand the difference between two ways to get a connection from a Weblogic 6.0 connection pool. I've come across two different ways that we can get a connection from a pool (after it's been created thru the console).
1. Have a connection pool manager with a getConnection method that uses the following code:
Connection conn = null;
Driver myDriver =(Driver)Class.forName("weblogic.jdbc.pool.Driver").newInstance();
Properties props = new Properties();
props.put("connectionPoolID","PoolName");
conn = myDriver.connect("jdbc:weblogic :pool", props);

2. Using JNDI and providing a URL (such as PROVIDER_URL below) for the machine the connection pool is on. The following code could be used in a connection pool manager class too:
Context ctx = null;
Hashtable ht = new Hashtable();
ht.put(Context.INITIAL_CONTEXT_FACTORY,"weblogic.jndi.WLInitialContextFactory");
ht.put(Context.PROVIDER_URL,"t3://hostname :port");
try {
ctx = new InitialContext(ht);
javax.sql.DataSource ds =(javax.sql.DataSource) ctx.lookup("myJtsDataSource");
java.sql.Connection conn = ds.getConnection();
} catch (Exception ex) {}

Is either method preferable to the other? Or does it not matter?
Are there better methods?
Thanks for your time
Owen

[This message has been edited by Owen Nishimura (edited August 01, 2001).]

hmmm!
just sharing some thoughts.
Let�s suppose that some day you want to use your ejb�s with some different DB from different vendor, if you got the connection from the first way you post you�re gonna have to change the code from your EJB ok, now if you use DataSource all you need is binding the dataSource environment with the new connection pool you�ve just created and dont need to change anything in your code.
And if you have CMP you don�t even need to use code inside your EJB to get a connection all you have do to is bind your dataSource with your ejb at the weblogic-cmp-rdbms.jar
as I said just my considerations I�d love to hear more opinions about that.