On Aug 5, 2004, at 12:32, Shinobu Kawai wrote:

Without a benchmark, I would say that in a typical response there are
*lots* of method calls from the ones in your code passing through
Struts/Hibernate/etc. down to the Java library. Some of them
non-trivial as the ones regarding persistence. If the comparison was
going to be 1 versus 1.00000000000001 (if it is measurable at all) I
wouldn't see a real trade-off there.
I think it's not the method calls, but the creation of the log messages
that makes performance an issue.
cf. http://jakarta.apache.org/commons/logging/api/org/apache/commons/ logging/Log.html

Hmmmm, I am not sure how that page relates to your remark.

They are giving an exemple where the use of the explicit test is useful: when you need to do some expensive work _only_ to produce a result to be logged. With the use of the test you can avoid executing that expensive code that is not needed if logging is off.

-- fxn


--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to