Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!snorkelwacker.mit.edu!bloom-beacon!dont-send-mail-to-path-lines
From: mouse@lightning.mcrcim.mcgill.EDU
Newsgroups: comp.windows.x
Subject: Re:  XDrawRectangle w/ negative width/height?
Message-ID: <9102071335.AA10848@lightning.McRCIM.McGill.EDU>
Date: 7 Feb 91 13:35:57 GMT
Sender: daemon@athena.mit.edu (Mr Background)
Organization: The Internet
Lines: 48

> So, the question is:  can you specify a negative width or height for
> a rectangle with no unexpected side effects, and, if that is legal,
> are there any known bugs when doing this?

At the protocol level, XDrawRectangle turns into a PolyRectangle
request.  The width and height for the rectangles in a PolyRectangle
requests are CARD16, which is an unsigned type.  Therefore, you
definitely are not going to get what you seem to be expecting.

The Xlib document describes the width and height arguments to
XDrawRectangle as "unsigned int", which perhaps should tell you
something :-)

Just what you *will* get depends on your Xlib, your compiler, and your
machine: when you pass negative width and/or height values to
XDrawRectangle, what does it generate in the transmitted request?  I
would guess that the most common thing is for it to blindly take the
low-order 16 bits, which usually means that negative numbers turn into
very large numbers (eg, -100 becomes 65436).

As far as I can see, it's your responsibility to make sure that the
width and height are always positive by interchanging values when
necessary.  It doesn't even take much code:

MyDrawRectangle(disp,d,gc,x,y,w,h)
Display *disp;
Drawable d;
GC gc;
int x;
int y;
int w;
int h;
{
 if (w < 0)
  { x += w;
    w = - w;
  }
 if (h < 0)
  { y += h;
    h = - h;
  }
 XDrawRectangle(disp,d,gc,x,y,(unsigned int)w,(unsigned int)h);
}

					der Mouse

			old: mcgill-vision!mouse
			new: mouse@larry.mcrcim.mcgill.edu