Rich 'Shirley' Scheuerle wrote: >C) I think that it is okay for a tool to generate a full constructor if it >would like....its just not part of the spec, so the TCK should not depend >on it existing. > A comment on this - in the meeting with Sun last week, Eduardo took great lengths to educate Glen, Tom, and myself that the RI is not just any tool - in his mind, it must not become a vehicle for establishing defacto standards by including externally visible features which are not in the spec. When I heard that I took a mental note to look for opportunities to test the strength of Eduardo's conviction on this subject. This is an excellent test case.
Russell Butek wrote: >We all seem to agree. The TCK should not REQUIRE the full constructor on beans. > >So now what's the process? Should I start stepping through the "JAX-RPC TCK Test >Appeals Steps"? Or are we allowed to shortcut that process by talking directly with >Rahul and company? Sam? > Another thing worth noting - I am looking for ways to establish a dialog between Sun and Apache (as opposed to IBM or Macromedia). This being said, if product schedules are being gated by this issue, then the formal mechanism should be pursued in parallel - Sam Ruby
