IMHO, there is no need to have multiple listeners for context/configuration
changes. If at all, practically there will be only one listener each for
context & configuration management. So I feel we should either allow the
user to specify only one listener, or scrap the listener interfaces
altogether and handle functionality the listeners were supposed to do in the
managers.
As I see it, nodes N1 & N2 will both have Managers running on them. The
managers will communicate with each other, just like the TCP/IP stack
N1 N2
--------------------------
---------------------------
| ContextManager | <------------------> | ContextManager |
--------------------------
---------------------------
The Managers send messages & communicate with each other. So for each
function there will be a pair of methods.
e.g. Node N1 ==> updateContext()
Node N2 ==> contextUpdated()
Thoughts?
--
Thanks
Afkham Azeez
http://www.wso2.org
GPG Fingerprint: 643F C2AF EB78 F886 40C9 B2A2 4AE2 C887 665E 0760
