Mailing Lists: Apple Mailing Lists
Image of Mac OS face in stamp
 
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Mac OS X Java Performance

"Hall, Michael J." <email@hidden> wrote:

>I am aware that there are tests that Java code must pass to qualify as
>Java releases.

AFAIK, the tests are for behavior, not performance.


>I am aware that there is a language specification that defines what the
>behaviors of Java classes should be.

The spec is essentially mute on performance.


>Should the JVM qualifications include performance requirements as part of
>the behavior spec?

No. For reasons much too long and far off-topic.


>So, is this acceptable and Windows should just be commended in this case
>for overachieving and providing a DataOutputStream that performs better
>than it needs to? Should the Mac JVM be questioned because it provides an
>under-achieving one?

What is the performance achievement standard, that one could be called
"over" and the other called "under"?

The only performance standard I know of for Java is one's own personal
expectations. Same goes for every other language I can think of.

If you want fast I/O, buffer it. This approach is literally decades old.
It was well known when C was still just the third letter of the alphabet,
and there is no reason to believe it will change in the next few decades.
Everything gets faster, but the relative gaps between CPU, memory, and disk
speed are at least as large as they've been for decades. They may even be
larger.

Of course, if you don't know to use buffering to obtain speed, or what the
tradeoffs are, or what the available I/O classes that buffer are, then
that's a problem. But I think that's a question of knowledge and
expectations more than anything else. The goal is more knowledge, less
expectations and tacit assumptions.

The knowledge (of Java, I/O, etc.) is readily available and fairly
manageable. Controlling expectations and avoiding tacit assumptions is
harder, especially when it depends on unintentional accidents of
implementation rather than intentionally considered design criteria.
Designs, however, are often victims of expectations as much as
implementations are.


>Again, doesn't a+b = c just sort of oversimplify?

Always.

But that's an oversimplification, too.

-- GG
_______________________________________________
java-dev mailing list | email@hidden
Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/java-dev
Be sure to read the FAQ http://developer.apple.com/java/faq/ before posting
Do not post admin requests to the list. They will be ignored.


Visit the Apple Store online or at retail locations.
1-800-MY-APPLE

Contact Apple | Terms of Use | Privacy Policy

Copyright © 2007 Apple Inc. All rights reserved.