I somewhat agree, but note that I qualified point 2 with “Aristotelian”. The alternative “prototypical” view (which is what most want to use inheritance for) is

I somewhat agree, but note that I qualified point 2 with “Aristotelian”. The alternative “prototypical” view (which is what most want to use inheritance for) is unsuited. Regarding the link, I remember Eric Lippert’s series of blog posts and his struggle to (initially) model a prototypical type problem using inheritance. I remember thinking at the time that he was limited in his object modeling experience to only the constructs provided by the C# language. There are no language constructs that specifically deal with rules relating to (optionally) collaborating objects, as he discovered.