On Friday, 09/20/2002 at 07:58 ZE9, Andy Clark <[EMAIL PROTECTED]> wrote: > And even if we agreed that it was worth the > damage, your change then requires all parser > configurations to have a NamespaceContext object > in their configuration.
Hmmmm. I'd missed that point. Alternative suggestion: Make the data available via the callback (if the NamespaceContext exists; if not, return appropriate error condition), continue to issue the events -- but put in a parameter that allows us to turn the events off when we know that a particular set of code would rather rely on the callback. Would that give us the best of both worlds? ______________________________________ Joe Kesselman / IBM Research --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
