> > I've a question about this code since it was necessary to create a second > event for the ie object even though the original object and it's event > should (I believe) be available. Am I doing something wrong here?
By setting ie.event, you are adding a new attibute to the IDispatch wrapper provided by Python. When you marshall it to another thread, you get a new IDispatch object, as the marshal process provided a new pointer. There is no magic that associates multiple instances like that. Try just passing the event as a normal param. > Second when I test this code I see that the thread runs ( trace messages > thread: tic occur). But events are not being pumped since none > of the event > messages occur. Eventually the thread times out. By running a message pump in the thread (*and* the main thread), I expect to see all events always be delivered to the *main* thread, regardless of what thread you use it from. The reason is that IE itself is not fully free-threaded, so events etc will be delivered to the IE "main thread", not to the thread it is being currently used from - that is the point of the marshalling calls you made - to give COM the chance to marshall back to the correct thread. Mark _______________________________________________ Python-win32 mailing list Python-win32@python.org http://mail.python.org/mailman/listinfo/python-win32
