qpid-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hiram Chirino" <hi...@hiramchirino.com>
Subject Re: A question for the ActiveMQ chaps on the list...
Date Mon, 25 Sep 2006 15:20:11 GMT
On 9/25/06, Gordon Sim <gsim@redhat.com> wrote:
> Hiram Chirino wrote:
> > On 9/25/06, Gordon Sim <gsim@redhat.com> wrote:
> >> When receiving an AMQP bind command asking that queue x be bound to
> >> exchange y, how does combining these into a compound destination name of
> >> some sort help us define a mapping for that command onto JMS semantics?
> >>
> >
> > Doesn't that information uniquely identify the destination?  If so,
> > then you can use jms semantics to send and receive messages from it.
> I'm not with you I'm afraid.
> On receiving the bind request I must ensure that any subsequent messages
> published to exchange y (that match the given binding details) are
> delivered to queue x.

So basically if the destination contained all of the binding
information.  The producers would be able to create the correct bind
request.  So that messages are delivered to the right queues.

And since that binding information specifies queues where messages are
being delivered, I think the consumer side of things should have
enough information to create it's subscription.

> A given JMS implementation can I'm sure find ways of doing that. My
> point was that I don't think we can define a generic mapping for this
> AMQP command onto JMS semantics.

Sure.. JMS is an abstraction after all and implementations will vary.
I'm just saying the qpid can come up with an implementation that does
do a good job of mapping the exchange/binding concept to JMS.


Blog: http://hiramchirino.com

View raw message