On Wed, Jun 24, 2009 at 11:20 AM, John Tamplin<j...@google.com> wrote: > I prefer 4.
As I said, I'm happy with 4. > What will the fallback behavior be for runAsync calls without a name -- an > automatically generated name? >From Lex's original post, it seems the name will be used to manipulate the split point from within the running program somehow. It would seem to me that an anonymous split would simply not be manipulatable and the name could be null, if it needs to be anything at all. Ian --~--~---------~--~----~------------~-------~--~----~ http://groups.google.com/group/Google-Web-Toolkit-Contributors -~----------~----~----~----~------~----~------~--~---