Hi Bob, extending will be fine, but it should not be necessary. Can you open a jira issue to track this and attach a Junit test case.
On 22 February 2010 15:04, bob.deremer <bob.dere...@burningskysoftware.com>wrote: > > Hi AMQ experts, > > I posted a question last week, but haven't had any replies yet [ > > http://old.nabble.com/URGENT-QUESTION:-AMQ-5.3.0-bug-or-configuration-error------ObjectMessage-is--still--being-serialized-when-using-setObjectMessageSerializationDefered-and-setCopyMessageOnSend-td27654579.html > > http://old.nabble.com/URGENT-QUESTION:-AMQ-5.3.0-bug-or-configuration-error------ObjectMessage-is--still--being-serialized-when-using-setObjectMessageSerializationDefered-and-setCopyMessageOnSend-td27654579.html > ]. As a workaround, I've decided to implement my own ActiveMQ message that > extends ActiveMQObjectMessage. My goal was to override the default > serialization that is happening when an object message is sent via vm:// > transport so that only the object reference is passed around. > > Can someone tell me if there are any inherent problems with this? So far, > my solution seems to be working functionally, but my gut tells me that > there's a lot I am probably not aware of happening inside AMQ(s) message > dispatching. I've been debugging the AMQ 5.3 source over the past couple > days - which is how I found the issue in the related link. > > Ideally, I'd like to know if what I mention in the other link is a bug, or > if there is a reason ObjectMessage(s) are being serialized - even when I > set > the flags to not serialize them. > > Thanks for any help/clarification, > Bob > -- > View this message in context: > http://old.nabble.com/Is-it-safe-to-extend-ActiveMQObjectMessage-to-work-around-the-default-object-serialization--tp27688649p27688649.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com. > > -- http://blog.garytully.com Open Source Integration http://fusesource.com
