I just loaded up both your plot.ps and your plot.pdf. I can't see any of the issues you are talking about in the ps plot. The axes look the same thickness to me, and a look at the file will show that the postscript used to draw them is virtually identical. I don't see any variation in thickness of the line, or it's colour, but if I did I would expect that to be due to the antialiasing code trying to smooth out the line. If so then that's your viewer, not the underlying code. A quick look at the underlying postscript will show you that the curve is drawn consistently throughout its length, and the postscript to draw the two axes is identical apart from the location. (lines 156 and 202 of your file)
I do see the difference in the axis thickness in the PDF, but again given the identicality of the axes in the postscript I have to put that down to the ps/pdf converter, or the PDF viewer. My guess is that it's antialiasing again. Some viewers will antialias a horizontal or vertical line that is not exactly placed on a pixel boundary, making it seem thickner; some will shift the line so it is on a pixel boundary. Again it's your viewer's choice, not the postscript. Postscript can't know where the pixel boundaries will lie ahead of time.
David Clayworth Maplesoft GUI Developer
P.S. You can of course use postscript directly to draw x squared, but I wouldn't recommend it for anything more complicated. In fact it takes fifty lines of postscript to draw the axes, without the labels. And do you really want to be converting plot coordinates to postscript coordinates in your head, for fifty line segments, without a mistake?