Vadim,
Is it really bad enough to hold the release? Why can we not fix it in
the next release?
Shash
PS: Admittedly, my opinion is colored by the grunt-work of making
repeated released, but, still...:-)
Vadim Gritsenko wrote:
... or ThreadLocals are EVIL! If you have not heard the word yet,
ThreadLocals
tend to prevent (web) application(s) from being garbage collected when
thread
locals are not properly nulled out.
So in order to allow web applications using LogKit to be properly
unloaded,
LogKit's ContextMap has to have one more method:
/**
* Remove current ContextMap.
* This method removes a ContextMap associated with current thread,
* if there is any.
*/
public static void removeCurrentContext()
{
c_localContext.set(null);
}
Consistent usage of this method, like:
try {
... ContextMap.getCurrentContext() ...
} finally {
ContextMap.removeCurrentContext();
}
will guarantee that ThreadLocal is properly cleaned up and web app
will be
properly unloaded. I'd like to include this addition into the release;
and IMHO
there is no reason to do yet another RC build just because of it :-)
because it
is simple addition which does not alter any existing behavior.
WDYT?
Vadim
PS http://www.jroller.com/page/tackline?entry=fixing_threadlocal
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
