One month seems like a good amount time to figure out some of these, at least just to formulate a plan. I am eager to see Derby client get rigorous testing that precedes a full release and to get it to derby users. I am concerned about potential resource leakage, platform specific, multi-threading issues that may or may not be present. These issues are typically found and addressed during a full release process. I also think current dependence on JCC client driver in 10.0 hinders wider adaptability of Derby.
You do raise some touch questions. :-) Like you suggested, we can work on getting a plan in place for these soon. Satheesh > I think aiming for a release in a month or so would be good. > > However, I do not want to rush out the first version of the client as > that will suddenly give us a legacy codebase. We have only just seen > this code and there are a couple of areas that we should consider: > > * which JREs will it run on > * do we need JSR-169 support > * what are the plans for SSL transport, including client auth > * what do we do with Referenceable > * what are our plans for connector support > * do we have the right properties > * do we change the commmand line processing for NetworkServer > > and probably a few more things as well. We don't need to resolve all > of these but at least figure out what the plan is. > > -- > Jeremy > > >
