That article, of course, is only of use if your app is cpu bound
working over primitives and simple objects. That can be a
bottleneck, but I would not want to bet on it.
I agree with Greg: there is no simple answer. You absolutely must
test a sample that is representative of what your app really does.
If you app draws aliased graphics, Windows will run much faster, as
the antialiased graphics path is much faster on MacOS X than the
aliased one. Turn anti-aliased code on, as Java 1.6 does, and
windows gets trounced. I suspect, though I would not want to bet
lots, that under Vista, it no longer gets trounced thoroughly,
because Vista also uses antialiased graphics.
Scott
On 7/18/06, Greg Guerin <email@hidden> wrote: Bill Wagner wrote:
>With all that said, what do you fellows feel would be a set of
good low end
>(as in slow and old) machines to test against? The intent is to be
able to
>determine what the CPU/speed limits should be on our product.
There's no simple answer. (You knew I'd say that, right?)
You have to look at what your product is, who the target audience
is, and
then estimate what kind of cost constraints most of them live
under. This
is typically a marketing task, but since you didn't describe your
product
or your company, "typically" may not not apply to you.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Java-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/java-dev/email@hidden
