Tim Jowers <> wrote:
> Jon Skeet <> wrote:
> > Adam Maass <> wrote:
> > > > The 10 time probably refers to Vector rather than ArrayList. I can't
> > > > remember where I read this or the details. I just remember the number
> > > > 10. The penalty should be less now the synchronized has less penalty.
> > > >
>
> JBench did not compile due to needing some outside classes.
Did you not just download the jar file, JBench.jar? That contains
everything it needs to. I believe it also says on the website where to
get the other classes which are needed.
> I wrote a quick test. The whole issue is really how the arrays are used and
> maybe the best "test" would be one that allows this to be configured
> so we can test based on the expected/measured usage patterns. I heard
> a talk on a GC that used statistical analysis...OT.
Indeed, actual usage would (I believe) show interesting results - like
that the performance difference between arrays and ArrayList probably
has no significant bearing on performance in most circumstances.
--
Jon Skeet - <>
http://www.pobox.com/~skeet/
If replying to the group, please do not mail me too