On Jul 17, 2013, at 8:55 PM, Gary Gregory wrote:
> Hi all,
>
> If you scan the source for all senders of "printStackTrace" you'll find 21
> references.
>
> Are all of those correct instead of logging to StatusLogger or throwing an
> exception
I would say "none of them are correct." Outside of testing, I can think of
exactly zero reasons where you would ever want to call printStackTrace in a
library or non-console application. They should all go away, IMO.
> or a different exception in the case where this is done in response to
> catching an exception.
>
> Also, in some cases like
> org.apache.logging.log4j.ReflectionComparison.setupCallerCheck(), both are
> done:
>
> } catch (final ClassNotFoundException cnfe) {
> cnfe.printStackTrace();
> throw new RuntimeException(cnfe);
> }
>
> In this case, should we not call StatusLogger?
>
> I'd also rather never throw a RuntimeException, instead an
> IllegalStateException.
>
> Also, this method is undocumented and uses a dynamic reference to the Java 7
> sun.reflect.Reflection class. How is that going to work on an IBM JVM?
It's not. I don't know what ReflectionComparison does, so perhaps someone else
can shed some light on it. I know that ThrowableProxy also uses
sun.reflect.Reflection, but that's for optimization and it has a "back-up plan"
that activates if sun.reflect.Reflection doesn't exist.
>
> Gary
>
>
> --
> E-Mail: [email protected] | [email protected]
> Java Persistence with Hibernate, Second Edition
> JUnit in Action, Second Edition
> Spring Batch in Action
> Blog: http://garygregory.wordpress.com
> Home: http://garygregory.com/
> Tweet! http://twitter.com/GaryGregory
