On 3/13/07, Christian Stein <[EMAIL PROTECTED]> wrote:
Trustin Lee schrieb: > On 3/13/07, *Ceki Gülcü* <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>> wrote: > > At 02:09 PM 3/12/2007, Holger Hoffstaette wrote: > >On Sun, 11 Mar 2007 22:29:03 +0000, Jeremy Booth wrote: > > > > > Is there a reason why the adapters don't implement serializable > other than > > > a lack of demand? Would it be possible to make them > serializable so I can > > > switch?, looking at the javadocs I assume it would be the > logger adapter > > > implementations, org.slf4j.spi.LocationAwareLogger and > org.slf4j.Logger > > > that would need the change. > > > >I don't think this is either necessary or desirable. If your > classes are > >Serializable, make the loggers transient and override readResolve() to > >recreate/set the log instance. Cuts down on useless serialization > overhead > >too! > > Sounds pretty reasonable to me. Jeremy, would Holger's suggestion above > work for you? > > > Implementing readResolve() for every class with a logger field is > painful. Would there be any easier way? Of course, the easiest > solution is not to use logger for serializable objects, which might make > sense. Or use the static logger per class approach. Although, not recommended [1], it could solve the resurrection problems. Or is static a no-go in your scenario?
It's OK in my scenario, but I just wanted to say that it's very cumbersome if there's such a scenario. Trustin -- what we call human nature is actually human habit -- http://gleamynode.net/ -- PGP Key ID: 0x0255ECA6
_______________________________________________ user mailing list [email protected] http://www.slf4j.org/mailman/listinfo/user
