Bill, So lets say that the Windows cursor is at 100, 100 (screen coords), the upper left-hand corner of the Cygwin/XFree86 client area is at 25, 25 (screen coords) and a warp request comes in to move to the X cursor to 200, 200 (X windows coords). I am thinking of two things that you could be saying:
1) The cursor stays in the same place because the request was for a translation of 100, 100, which is the same location that the cursor is presently at. (In other words, I am receiving an offset to move the mouse by and I'm instead treating those values as the final destination coordinates). 2) Or, I receive a request to move the cursor to a final location of 200, 200 and I do so, which fails to take into account that the X cursor root is at 25, 25 in Windows screen coodinates. Which one is it? I think it sounds like #2, in which case the (x,y) that I get in winWarpCursor are the final destination values, but I need to query for the upper-left hand corner of the Cygwin/XFree86 client area before calling SetCursorPos so that I can offset the final destination. We'll get it right eventually. Notice that no one has really request the WarpCursor functionality before, that's why it doesn't work yet. :) Harold "Trost, Bill" <[EMAIL PROTECTED]> said: > From: Harold Hunt [mailto:[EMAIL PROTECTED]] > > I'd like you to try the new Test 54 server that I just put > together and announced to the list. Please let me know if it fixes your > problem. > > That's better, but still not quite right -- it looks as though the pointer > is being visually warped to a window location, but logically warped to a > screen location. That is, the pointer now jumps by the X server window's > distance from the screen's <0, 0> location when I move the mouse. >
