On Mon, 21 Jul 2003 15:52:33 +0200 Ceki Gülcü <[EMAIL PROTECTED]> wrote:
> > At 07:48 PM 7/20/2003 +0200, Lucas Bruand wrote: > >At the very least, it is a heavy argument against people who think that > >Logging is a "retro" method to debug and claim that logging is hindering > >performance in Java... ( Probably preaching on this list is not very > >useful but you'll be surprised how often you hear this kind of thin) > > Logging maybe retro but its survives the life span of the application. > Application logging is analogous to fossils. Fossils are certainly retro > but it is the only way to learn about life forms long gone.
I completely agree with you... Some people think that given a debugger and the description of the problem, one should be able to reconstitute the whole process. Obviously this is true only in theory but it is somehow trendy for some decision makers...
Still, there is nothing like a well written logfile to debug a program...
Running a debugger when developing the application is quite different than running one at deployment time. When you deploy an application and it crashes, it is helpful to know what happened before the crash. AFAIK, no one uses a debugger at deployment time.
I guess that is the reason why airplanes have a black box and not a debugger. :-)
--
Ceki For log4j documentation consider "The complete log4j manual"
ISBN: 2970036908 http://www.qos.ch/shop/products/clm_t.jsp
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
