Do we need to address the sandbox dependencies on impl before a release? We'll probably need to move some things from impl to shared to fix them. I'm pretty sure that I've identified the classes involved in the following issue.
https://issues.apache.org/jira/browse/TOMAHAWK-220 -Mike On 3/29/06, Sean Schofield <[EMAIL PROTECTED]> wrote: > I haven't reviewed those discussions. Are any of these issues so > problematic that we don't want to release for several weeks? That's > the real decision here. I agree that its a bit awkward to have to > configure the logging with shared_impl and shared_tomahawk but > ultimately I think the benefits of the shared solution outweigh the > problems. > > Whatever solution we come up to solve the shared problem, its likely > to have a few drawbacks. So my vote would be to go forward. > > As for the sandbox, that's not an issue right now since we are > planning on releasing the core for now (and then tomahawk.) So we can > cover that when the time comes. > > Sean > > On 3/29/06, Dennis Byrne <[EMAIL PROTECTED]> wrote: > > >Basically I think the plan should be to merge down this single change. > > > Create new branches for the core and shared. Release shared. Test > > >Core and release core. Before making the new branches I will change > > >the dependency to the latest shale-test which is now available. > > > > Were there any conclusions from the discussion Mario started a while back? > > There are a lot of classes that use scoped objects keyed by package name. > > > > What do we tell people about logging ? Currently people have to do things > > like this: > > org.apache.myfaces.shared_impl.StateUtils=TRACE > > org.apache.myfaces.shared_tomahawk.StateUtils=TRACE > > > > Any conclusions on Simon's static logger discussion ? > > > > I will help fix any of these. > > > > Dennis Byrne > > > > > > >
