On 12/12/2010 05:59 PM, Niclas Hedhman wrote:
The original intent of the Jini specification was that only the net.jini.* namespace was supposed to be used by application built on top, but very early this was breached both by recommendation and by lack of enforcement. So, at the moment, there are a handful of classes in the com.sun.* space which is considered part of the 'official' API, or at least recommended as 'work-around' for certain recurring issues people have.
Looking back in history, Jini v1.0 had some classes in com.sun.* which in subsequent releases got 'promoted' to net.jini.* namespace, and that should be the recommended move forward here too.
It was discussed on river-dev and the idea it got from this was that we still believe in a separation between specification and implementation. The implementation that resides in com.sun could be renamed into the org.apache.river.impl namespace without causing to much conversion activity with the users of river. I believe there are strong feelings about keeping to the original specifications. My personal view on this is that the original specifications should not stand in the way of new developments, but since this does not block new developments at this time i have no objections to following the consensus.
I know this is the prerogative of the IPMC to determine this, but your remarks deviate a bit from the consensus formed on river-dev IMHO, should we go back to the drawingboard and discuss your proposal within the PPMC?
Gr. Sim --------------------------------------------------------------------- To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
