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

[Eric Crahen]

On 2/19/07, Ceki Gülcü <listid at qos.ch> wrote:

> One of the reasons why SLF4J is relatively successful is due to its
> simplicity. SLF4J only does static-binding and ito fails-fast. If
> there is no slf4j-binding available SLF4J prints an error message and
> dies immediately (it fails-fast).


Still possible with the Service API; except you could have much better error
messages than is technically possible with the current solution.

The ServiceProvider interface is only available in JDK 1.6.


It is available since 1.4 as sun.misc.ServiceLoader, even in 1.6

Moreover, it used the thread context class loader. SLF4J is not going to go
> on
> that road in the immediate future.


I really think it would be worth everyone while if time is taken at some
point to explore this solution. I get the strong sense there is an unfounded
fear of ClassLoaders due to past bad experiences with them; ClassLoader !=
problems. I look forward to seeing some serious consideration given to this
option in the future, instead of a knee jerk "classloader-bad!" response. I
think I've provided an enormous amount of information on the subject that
should be plenty to develop a working solution as soon as people are willing
to give it a try.

Thanks,

-- 
- Eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://qos.ch/pipermail/slf4j-dev/attachments/20070219/215298c2/attachment.htm>