ceki, really, this is perfect example why no container uses commons-logging anymore. Do you like to repeat these errors?
I reiterate: there is a workaround by isolating this in ClassWorlds. It is not yet there, but ff we like to use slf4j then we will need to implement that. LieGrue, strub ----- Original Message ----- > From: ceki <[email protected]> > To: Maven Developers List <[email protected]> > Cc: > Sent: Thursday, October 11, 2012 12:25 PM > Subject: Re: SLF4J integration > > On 11.10.2012 11:18, Mark Struberg wrote: >> Oh I missed one more constellation >> >> a plugin could have slf4j-1.5.x + a logging backend we do not know of. >> >> I hope such things dont often exist, but in theory it could happen. >> >> For all of those cases we need isolation. > > A few months before creating the SLF4J project, I had a lengthy > conversation on the commons-dev mailing list about logging scenarios. > Richard Sitze (from IBM) offered a number of complex scenarios, and > that's when the penny finally dropped for me [1]. It is a waste of time > catering for all imaginable logging scenarios. The space of possible > logging scenarios is truly vast. However, only a very limited number of > these scenarios occur in practice. It makes more sense to concentrate > on the common scenarios to ensure that they are well supported. > > More concretely, I propose that we find one or two examples of plugins > declaring an slf4j dependency and test these plugins with the upcoming > version of maven (the one using slf4j). > > [1] http://marc.info/?t=110780972600001&r=1&w=2 > > -- > Ceki > 65% of statistics are made up on the spot > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
