On 4/12/06, Adrian Tarau <[EMAIL PROTECTED]> wrote: > Well I'm not sure about those 5 seconds, I think it makes sense only if > the consumer starts consuming and the connector was not started.
Its perfectly legal in JMS to do this - many people do - so we shouldn't create a warning for perfectly valid JMS applications. > I agree, is perfectly normal to create multiple consumers for the same > connection and then start the connection, but if I consume before the > connection was started, a warning could help to figure out why nothing came. Oh, you mean consumer.receive()? Its starts to get a little complex now; as the thread doing the consumer.receive() could be different to the thread doing the connection.start(), so its still valid to do a consumer.receive() before the connection is started. Plus lots of people use the async consumption model - which this approach won't help; so I think i'd rather a more general catch all that consumers created on a connection that doesn't get started within some time window (1-5 seconds or so) we generate a warning. -- James ------- http://radio.weblogs.com/0112098/
