I don't think it's relevant who wrote the namespace.
I'm referring to the direction of dependencies.
It's Spring that's integrating HornetQ, since it's Spring that's configured to talk to HornetQ.
HornetQ knows nothing about Spring. But Spring must know something about HornetQ in order to talk to it.
Maybe I'm not being clear. Spring CAN work with HornetQ since HornetQ is just a bunch of POJOs. Its POSSIBLE, but what I would love to see is it make it EASIER.
For instance, you sent a link about the HornetQ/Spring integration, it was a very long and in depth read with a large amount of code and xml to be done.
To get a JMS ConnectionFactory with an in memory broker in ActiveMQ you add their name space to the top of the existing Spring xml for the application and you add one line of xml config. Thats it. Moreover, you can auto create destinations so there is no need for any other additional code. Just one line of xml and away you go. As you can see that is a significant difference from the blog post you linked to.
Finally, to be clear you could use ActiveMQ without its namespace support in Spring as well but it would require more work too, just like in the HornetQ case. However, they took a look at their Spring support story and decided to improve it. Its not about whats POSSIBLE its about what is EASY.
We already agreed it would be a nice feature to have
Clebert Suconic wrote:
We already agreed it would be a nice feature to have
+1.
No disputes about that.
I was simply responding to Benjamin's original comments that *HornetQ* was hard to configure, and the examples were useless.