Hi there - answers are inline...
On Wed, Nov 17, 2010 at 4:56 AM, ep <[email protected]> wrote:
> hello folks,
> whats the difference between the:
>
> <set-property name="gwt.logging.logLevel" value="OFF"/>
This will set the Level of the RootLogger to be OFF
>
> and
>
> <set-property name="gwt.logging.enabled" value="FALSE"/>
This will swap in Null Implementations for the logging related
classes, which will compile out.
>
> ?
>
> I understand I can use LogConfiguration.loggingIsEnabled() to wrap the
> code I want to optimize away during compilation, but I could also do:
>
> if(java.util.logging.Logger.isLoggable(Level.INFO)) {
> Â logger.info("this code will be compiled out if logLevel is below
> requested? complex result: " + doSomeComplexCalc());
> }
>
> couldn't I?
Yes - the LogConfiguration method is just something I added for convenience.
>
> or, I guess that gwt.logging.logLevel does not effect the compiler, so
> only gwt.logging.enabled will provide NULL implementation?
Yes - exactly - see my response above
>
> what about the loglevels - is there new permutation per each or are
> logging instructions not inlined during compilation? ( like in gwt-log
> project )
Not sure exactly what you mean here
>
> are there more papers on new logging I can dive in?
>
> thank you.
Not really, but you can look at the source code - it's actually more
straightforward than you would think. Note that you'll want to look at
the both the files in logging, as well as the emulation classes in
supersource. However many of the emulated classes just punt out to
Null or Regular Impl classes, which are in the logging directory.
>
> --
> http://groups.google.com/group/Google-Web-Toolkit-Contributors
>
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
