... how does Versiontracker compute the overall rating as 1 star, after three votes (1-star twice and 5-star once?) ....? Surely it should use the mathematical average (mean) of the ratings (1+1+5 divided by 3 raters = 2.3, or 2 stars - in an ideal world, it would disable double voting and the true mathematical average of the two votes would be 1+5 divided by 2 raters = 3 stars) not either the mode or median (both of which would be 1 star, as currently shown)

It's a mathematical aberration that's unfair to the guy porting this vintage prog to OSX.
So my vote of four stars (one notch above Neutral on the ratings page) is to restore some fairness ... let's hope the rating system reflects that!

;o)