Michiel Meeuwissen wrote:
> Update of /var/cvs/src/org/mmbase/util/transformers
> In directory james.mmbase.org:/tmp/cvs-serv4014
> 
> Modified Files:
>       ChainedCharTransformer.java 
> Log Message:
> oops, that dit not compile. 
> 
> Index: ChainedCharTransformer.java
..
>  
> -import edu.emory.mathcs.backport.java.util.concurrent.*;
> +//import edu.emory.mathcs.backport.java.util.concurrent.*;

>      
> -    static final Executor executor = Executors.newCachedThreadPool();
> +    static final MMExecutor executor = new MMExecutor();


Actually, the change which I made here, and rolled back, was a good one, I
think. I tried to use a thread-pool, because this 'chaining' needs a thread
between chains, and that could be quite heavy on resources. 

ThreadPools are natively supported on java 1.5.

I think however that it is currently out of the question to demand java 1.5,
so what I tried here was to depend on the 'concurrency-backport'.  This is
an effort af the people (a.o.) who also implemented the java.util.concurrent
package of java 1.5. See 
http://www.mathcs.emory.edu/dcl/util/backport-util-concurrent/

Would it be a good idea to ship mmbase 1.8 with this 'concurrency backport
jar', so that we can use some of the concerrency related features of java
1.5 already without having to depend on it? We can drop it later, when we
drop support for java 1.4. 

Michiel

-- 
Michiel Meeuwissen                  mihxil'
Mediacentrum 140 H'sum                [] ()
+31 (0)35 6772979         nl_NL eo_XX en_US



_______________________________________________
Developers mailing list
[email protected]
http://lists.mmbase.org/mailman/listinfo/developers

Reply via email to