> It's not so wacky.  For log4j v1.3 I am adding "receivers".  
> Receivers are
> the opposite of appenders in that they can be configured to "receive"
> remote/eternal events.  They can be configured as part of a logger
> repository, same as loggers and appenders.  Once the receiver 
> has an event,
> it "posts" it to the logger repository where it gets handled by the
> configured appenders as if the event had been generated 
> locally.  So, it can
> be written to the console, written to a file, etc. v1.3 is 
> going to have a
> SocketReceiver (matching SocketAppender), SocketHubReceiver (matching
> SocketHubAppender), JMSReceiver (matching JMSAppender), and 
> maybe more as
> makes sense.

That's very much what I was thinking, but I hadn't thought about the even
more generic case of a Receiver, which you have.  Wow, that is cool.
 
> types.  Recievers make this possible.  Chainsaw does not need any
> appender/receiver specific code.  As new receiver classes are written,
> Chainsaw should be able to take immediate advantage of them with no
> modification.  Just configure it to use them via a (yet to be 
> designed)
> plugin gui or a configuration file.  

I like it.

> Also, Chainsaw should be 
> modified to be
> like LF5 in that the events displayed should be processed by 
> a Chainsaw
> appender.  

That's a good refactor of the concept of the internal Processor object,
would make TableModel code lighter weight, encapsulate the updating of the
model into a separate class.

> Does that make sense?  I can't wait to see some of these 
> changes folks are
> talking about.  I want to get my hands dirty too.

Me too.  I am too afraid to start anything until some of the changes are
committed.

Paul

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to