[jboss-dev-forums] [Design the new POJO MicroContainer] - Re: JBMICROCONT-385 - Privileged actions for GenericBeanFact (Repost)

adrian@jboss.org do-not-reply at jboss.com
Tue Nov 18 10:37:51 EST 2008


"alesj" wrote : "adrian at jboss.org" wrote : 
  |   | We might want to create an interface for this contract since it is not good
  |   | that the GBF depends upon a particular implementation of the KernelControllerContext
  |   | interface?
  | Probably doesn't belong to KCC,
  | more like our InvokeDispatchContext?
  | 
  | What would be a good name? :-)
  | 

Don't really care about names, except when they obviously wrong....

anonymous wrote : 
  | Considering probable future usages/additions.
  | e.g. AccessControlContextAware?

"Aware" is usually for when you want a callback, e.g.
if the controller was going to do a setAccessControlContext()
Like having a bean that is KernelControllerContextAware
meaning there is a setKernelControllerContext().

Since this is just a getter interface, why not call it something like
ControllerContextAccessControlContextProvider? Its a mouthful
so you can probably shorten it? ;-)
Provider is the usual name if something is not
building, generating, loading, locating, etc.

Speaking of "locating", who is using the horrible KernelLocator
that has sneaked into the MC project despite all my explanations
about this being the wrong way to do it?

Basically, there never has been a singleton kernel in the MC (or the old JMX kernel
for that matter - although there we hacked it with the locateJBoss() method).

Having a class trying to emulate it, is bound to break
when somebody does create more than one.

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4190164#4190164

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4190164



More information about the jboss-dev-forums mailing list