Alan Kay has agreed to do an AMA today

Hi Alan,

On the "worse is better" divide I've always considered you as someone standing near the "better" (MIT) approach, but with an understanding of the pragmatics inherent in the "worse is better" (New Jersey) approach too.

What is your actual position on the "worse is better" dichotomy?

Do you believe it is real, and if so, can there be a third alternative that combines elements from both sides?

And if not, are we always doomed (due to market forces, programming as "popular culture" etc) to have sub-par tools from what can be theoretically achieved?

I don't think "pop culture" approaches are the best way to do most things (though "every once in a while" something good does happen).

The real question is "does a hack reset 'normal'?" For most people it tends to, and this makes it very difficult for them to think about the actual issues.

A quote I made up some years ago is "Better and Perfect are the enemies of What-Is-Actually-Needed". The big sin so many people commit in computing is not really paying attention to "What-Is-Actually-Needed"! And not going below that.

I fear this is because "What-Is-Actually-Needed" is non-trivial to figure out. Related: "scratch your own itch", "bikeshedding", "yak shaving".

Exactly -- this is why people are tempted to choose an increment, and will say "at least it's a little better" -- but if the threshold isn't actually reached, then it is the opposite of a little better, it's an illusion.