On September 27, 2003 04:52 pm, Alexandre Julliard wrote: > IMO the right place for that info is in a regression test that > actually checks the message order. It should be pretty trivial to > write a window procedure that takes an expected message list and > checks that each successive message it receives corresponds to the > next entry in that list.
Good idea. How do we simulate the mouse, for these cases: Moving the mouse Moving by dragging the title bar (after WM_NCHITTEST and WM_SETCURSOR) (outline move) Sizing by dragging the thick borders (after WM_NCHITTEST and WM_SETCURSOR) (outline move) Clicking on inactive button ANd how do we differentiate between messages sent and posted? Is this difference important to test? -- Dimi.