RE: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Martin Gainty
Hi Chris- did you run the testContextReferences() and testArithmentocs from org.apache.struts2.uel.PerformanceTest for 1000 JEUL testcase iteration? here is what i have: Running org.apache.struts2.uel.PerformanceTest Eval: [#obj.age + #obj2.age] on 100 iterations OGNL: 4547 ms JUEL: 9750 ms

Re: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Chris Pratt
Unfortunately the same cannot be said for OGNL. I spent most of Friday going through the interwebs archives and finding out that one major sticking point in my architecture (and all of yours) is that OGNL synchronizes just about everything. I spent about 6 hours going through the code and removin

RE: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Martin Gainty
so shared variable-use synchronized not-shared variable-use threadLocal struts code has only a few synchronized blocks or synchronized methods but plenty of ThreadLocal specifically the all important Dispatcher instance private static ThreadLocal instance = new ThreadLocal(); thanks chris Mart

Re: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread musomesa
They are quite distinct -- ThreadLocal variables are not shared at all while synchronized permits sharing (but not concurrently). Suppose you have a background thread that does some calculation that you might need [in my case you have an abstract graph and a background thread is checking if the

RE: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Martin Gainty
Buona-sera Antonio ga is nice but i've heard italy has more sunny days is this true? Martin __ Prego non alteri/modifichi o interrompa questa trasmissione..grazie > Date: Sat, 21 Nov 2009 22:07:43 +0100 > Subject: Re: (clearly O/T) use of ThreadL

Re: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Antonio Petrelli
2009/11/21 Martin Gainty : > >  yes in all probability still at > > > pingback when you posted pics of "sunny and dry ga" when you get a chance Really? I like sun and dry ga. Do you like them? Antonio P.S.: I'm not mad, just curious... ---

RE: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Martin Gainty
yes in all probability still at pingback when you posted pics of "sunny and dry ga" when you get a chance interesting graphic btw Martin Gainty __ WARNING: Material contained within this transmission has been deemed "not suitable" for liberal art

Re: (clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Dale Newfield
I think that Martin is a turing test...and I'm beginning to feel that he might indeed be human...which can only mean the AIs are winning. Can Skynet be far behind? ObXKCD: http://xkcd.com/329/ -Dale P.S.: Even more clearly O/T. --

(clearly O/T) use of ThreadLocal vs Synchronized

2009-11-21 Thread Martin Gainty
Good Morning All- are there any instances where a factory use of synchronized keyword is preferred or considered more efficient implementation over creating a ThreadLocal object? http://www.javamex.com/tutorials/synchronization_concurrency_thread_local2.shtml any answers are appreciated! Marti

Re: Strange behavior after upgrade

2009-11-21 Thread test mail
Dear frnd, I have never used strut 2 but if your using hibernate then u r supposed to flush your session to update your database. Regards D'Souza On Fri, Nov 20, 2009 at 5:13 AM, Thomas Sattler wrote: > Hello all. > > I have upgraded a Struts 2.0.11 project to 2.1.8, and I ha