On Tue, Oct 11, 2011 at 1:26 PM, Raymond Auge <raymond.a...@liferay.com> wrote:
> On Tue, Oct 11, 2011 at 11:48 AM, Philipp Kursawe
> <phil.kurs...@gmail.com> wrote:
>> On Tue, Oct 11, 2011 at 2:20 PM, Neil Bartlett <njbartl...@gmail.com> wrote:
>>>
>>> log.log(LogService.LOG_ERROR, "blah blah");
>>>
>>> i.e., why do I have to type the word "log" FIVE times to output *ONE*
>>> log message??
>>
>> Thats clearly a design flaw. The log4j way is much cleaner:
>> log.error("blah");
>
>>> Another issue is that there is no way for a client to check if
>>> debug-level logging is enabled.
>>
>> Thats also something pax-logging provides just nicely.
>
> I suspect these are two of the main reasons why people are deriving
> their own logging service wrappers to begin with.
>
> Though I suspect that it's to remove the OSGi APIs from their code as
> well. e.g. such as bridging other more ubiquitous logging APIs onto
> the LogService via a separate bundle (not that this is the correct
> approach, but in the vein of DS, this makes not polluting your own
> code with OSGi apis a moot point).
My point is simply that the LogService makes writing code free of OSGi
apis (as is commonly preached) difficult on top of the other concerns
above.
--
Raymond Augé
_______________________________________________
OSGi Developer Mail List
osgi-dev@mail.osgi.org
https://mail.osgi.org/mailman/listinfo/osgi-dev
