Its close to a one line change, so it doesn't need its own branch (I can
keep it local until its tested). I'm writing first because its a pretty
core change and wanted to see if anyone protests before proceeding down
this path.
Are you suggesting a branch for other reasons perhaps? Like pointing
the automated tests at it and running it into the ground?
Anil Saldhana wrote:
Why don't you create a branch for your experiments? When the time
comes, you can merge.
Scott Marlow wrote:
> Dimitris Andreadis wrote:
>> Please make sure you work on jboss head, Branch_4_2 is frozen right
>> now.
> Now that 4.2 has been released, is 4.2.x a good place to experiment
> with this change? I haven't been able to build head for a while.
>
>>
>> Scott Marlow wrote:
>>> Hi,
>>>
>>> A while back I measured the object lock contention in a JBoss AS
>>> 4.0.4 environment. I noticed a lot of time spent waiting for the
>>> internal JNDI collection lock. I think this is worth creating a
>>> Jira for and solving with a swego concurrentHashMap based
>>> collection (or java.util.concurrent when we use java5 as a base).
>>> Of course the actual contention is app specific but I started
>>> thinking about solving it.
>>>
>>> Is this worth pursuing? I'll create a jira and make code changes
>>> for it (I might need some help running the J2EE compatibility
>>> tests). Clearly, we want to pass the unit tests as well (thats an
>>> assumed requirement).
>>>
>>> Thoughts?
>>>
>>> Scott
>>> _______________________________________________
>>> jboss-development mailing list
>>> jboss-development(a)lists.jboss.org
>>>
https://lists.jboss.org/mailman/listinfo/jboss-development
>> _______________________________________________
>> jboss-development mailing list
>> jboss-development(a)lists.jboss.org
>>
https://lists.jboss.org/mailman/listinfo/jboss-development
>>
>
> _______________________________________________
> jboss-development mailing list
> jboss-development(a)lists.jboss.org
>
https://lists.jboss.org/mailman/listinfo/jboss-development