RE: RemoteNotificationClientHandler fix

Simon,

> I have committed my fix, and the nightly build is
> running now (will run again tonight).
> Check it out and let me know what do you think.

I don't have live CVS access at work - and Web-CVS
must lag the real CVS by some amount of time.  

I assume that you moved start/stop into the interface.
 That works for me - I was cautious of stepping on
your architecture.  If we are saying that there will
always be a semantic start/stop on dealing with remote
notifications - the conclusion of start/stop in the
interface naturally flows.  If however, the
RMIConnector's fetcher ( for example ) used a callback
mechanism from the server to the client - perhaps that
would not require a local (to the client) thread that
is not managed by RMI.  Given the current
implementation however - I don't know of a way to
accomplish w/o a start/stop.  Even in my example of a
server/client callback, an argument could potentially
be made for the semantics of start/stop.

Brian

__________________________________
Do you Yahoo!?
Yahoo! Mail SpamGuard - Read only the mail you want.
http://antispam.yahoo.com/tools


-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click