Comment on ‘‘Linear least‐squares fits with errors in both coordinates,’’ by B. C. Reed [Am. J. Phys. 57, 642–646 (1989)]