[slf4j-dev] Comments on scala-api

Hello Heiko, Looking at logger.scala, I was wondering why the Logger trait was a trait and not a class. I understand that as a trait any class can mixin Logger and methods such as debug(), info() would be available in that class. However, there is already a trait caller Logging which allows the mixing class to write logger.debug("..."). It seems to me that the Logging trait is sufficient in most cases. I would like to propose to make Logger a class instead of a Trait and to get rid of DefaultLogger. This would make the code a little simpler without loss of "desired" flexibility. I don't think we should allow classes to write just debug("...") or should we? WDYT? -- Ceki