trumpetinc

Yes - it is an optimization. I'm pretty sure you can re-use the cached value if there is only a translation. But please sanity check my logic on this. This...

I think that the key here is to separate the API into layers, and use naming conventions that indicate the layer. I wrote a little bit about this here: https://github.com/ghetolay/jwamp/issues/27...

Just noticed that the WampClientEndPoint does use the ExtendedWampEndpoint(WampDispatcher) interface... Boy, the JSR356 folks have created a bit of a mess here by preventing control of the instantiation of endpoints....

The equivalent race in the messageoverhaul branch was in the Promise class, which I pushed a fix for yesterday (thank you for spotting that). Would you be willing to do...

Can you do me a favor and take a look at the code in the messageoverhaul branch? I've got a complete rework of jWAMP underway that resolves a lot of...

Chuck - I have created a separate Issue for the Promise race condition you identified ( https://github.com/ghetolay/jwamp/issues/42 ). I have it fixed and will push that in a second. Can...

Hah - I looked at Guava cache also (and several others)! I would prefer to create the new TimeoutHashMap implementation - it'll be a very simple, elegant class (much more...

I've started putting together the alternative implementation of TimeoutHashMap, and I have a question: In DefaultRPCSender it looks for all the world like we are never removing result listeners from...

ok - I have comitted a new branch with a new implementation of TimeOutHashMap (and changing DefaultRPCSender so it uses remove() instead of get() when a response is received). https://github.com/ghetolay/jwamp/tree/timeouthashmap-redesign...

I just finished up a few finishing touches on the new TimeOutHashMap. I'm wondering if we should rename this to TimeoutCache or something like that - it doesn't fully implement...