At 09:16 AM 6/24/2003 -0400, Berin Loritch wrote:
Ceki Gülcü wrote:
In my experience, and I have only been doing high volume server apps for about
four years now, you can never rely on theory. Always err on the side of
caution.
The core of log4j is quite small. So what applies for large pieces of software, does not necessarily apply to the case in hand.
Perhaps.
Erring on the side of caution is perfectly reasonable if you do not want to spend the brain cycles to understand the problem in all its details. However, if you have the luxury of thinking about the problem long enough or if the problem is particularly simple, you can come up with an approach that nicely fits the problem.
:) I guess you are calling me lazy ;P (Like all programmers)
All I was saying is that experience has taught me that when I think I have it right and thought through things properly, I'm proven wrong with cold hard facts.
I'll defer to your experience in this area, but if Log4J borks in one of my high throughput server apps can I say, "I told you so"? I promise to have a patch along with it.... ;P
The category class is often read but seldom modified. Unfortunately it cannot be cloned. So the ReaderWriterLock seems to nicely fit the bill.
Hmmm. <thinking-aloud>Do I want to explore the guts of Log4J or just remain an informed user....</thinking-aloud>
Where did you get the code for the ReaderWriterLock? I remember in the now defunct Excalibur Concurrent utilities we had trouble with ours. We decided to go with the much better tested Doug Lea utilities which will be incorporated into JDK 1.5 along with generics and attributes (the other goodies I have been salivating for).
As to adding objects to arrays of vectors, then may I remind you that some
collections will bork on "null" objects?
Good point. The code in question (in Category class) checks against null objects.
Ok.
--
"They that give up essential liberty to obtain a little temporary safety
deserve neither liberty nor safety."
- Benjamin Franklin
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
