[concurrency-interest] Changes since initial public review spec

Angelika Langer langer@camelot.de
Thu, 19 Feb 2004 07:19:59 +0100


Thanks for pointing to the piece of information I was missing.

Another question:  I've been playing around with the ThreadMBean and its 
deadlock detection feature.  It only works for the built-in locks, but 
not for any of the explicit locks.  Is this because we are talking about 
a beta release?  Or will it remain that way?  From a user's perspective, 
it's kind of puzzling that deadlock detection works for the built-in 
lock, but not for ReentrantLock.

Angelika Langer


-- 
----------------------------------------------------------------
Angelika Langer
Email: langer@camelot.de
http://www.langer.camelot.de/
----------------------------------------------------------------


David Holmes wrote:

>>In an early release of the concurrency utilities there was a class Locks
>>which had a method newConditionFor(Object o).  In the current beta
>>release of Tiger I cannot find it anymore and the list of changes does
>>not mention it either.
>>
>>What happened to it?  Has it been replaced by anything?
> 
> 
> It is gone. It turned out that the idea of adding multiple Conditions
> associated with the monitor lock of an arbitrary object was impractical for
> a number of reasons. Doug touched on these reasons here:
> 
> http://altair.cs.oswego.edu/pipermail/concurrency-interest/2003-June/000467.
> html
> 
> Basically such a change to the semantics associated with synchronized and
> Object would break every development tool out there that thought it knew
> anything about the built-in synchronization features of Java.
> 
> David Holmes
> 
> _______________________________________________
> Concurrency-interest mailing list
> Concurrency-interest@altair.cs.oswego.edu
> http://altair.cs.oswego.edu/mailman/listinfo/concurrency-interest
>