[slf4j-dev] Consolidating the LoggerFactory / A better plugin mechanism

Ceki Gülcü listid at qos.ch
Mon Feb 19 22:12:01 CET 2007


At 09:58 PM 2/19/2007, Eric Crahen wrote:
>Actually no one has presented anything concrete that could be demonstrated 
>as an actual advantage over what I proposed, which is basically the 
>standard Java way to do what you are trying to do. I'm not sure you can 
>claim your solution is an evolved version of the ServiceLoader approach as 
>you were unaware of it until I presented it.

Just to be clear, I am not claiming the current solution is more evolved. 
On the contrary, it is clearly more primitive. In the earliest version of 
SLF4J when it used to be called UGLI, we parsed a properties file and then 
instantiated a logger factory by reflection, *similar* to the way 
sun.misc.Service does it, but without using sun.misc.Service.

In practice, the sun.misc.Service may well give the same results as the 
current static approach. However, that is not the point. The point is the 
difficulty of debugging the logging configuration by the end-user who might 
not be familiar with sun.misc.Service or properties files or reflection. 
However, even junior java developers are somewhat familiar with Java's 
basic class loading mechanism.


-- 
Ceki Gülcü
Logback: The reliable, generic, fast and flexible logging framework for Java.
http://logback.qos.ch




More information about the slf4j-dev mailing list