I think a more interesting classification would be which languages are favored by people who eat corn on the cob typewriter style, and which are favored by people who eat it spiral style.<p>I'm not joking. Corn on the cob eating style among mathematicians has been found to be a strong indicator of whether that mathematician favors algebra or favors analysis: <a href="http://bentilly.blogspot.com/2010/08/analysis-vs-algebra-predicts-eating.html" rel="nofollow">http://bentilly.blogspot.com/2010/08/analysis-vs-algebra-pre...</a><p>Someone should start making sure that at catered events where hackers congregate there is plenty of corn on the cob, and gather this data.
Accepting Yegge's framework for purposes of this response:<p>Haskell is beyond conservative, but it does take great pains to try to address the reasons that people might want a more liberal language. However, it's often really painful to get that code to typecheck compared with the effort to get the same code correct in python or ruby. Sometimes Haskell is the worst of both worlds: it requires mind-bending reasoning to write "liberal"-like code, when the "conservative" guarantees you get in return aren't for very interesting correctness properties.<p>I'm starting to think my ideal language would be gradually-typed (something like Groovy): use static typing by default, but fallback to dynamic typing for the pieces that require superhuman effort to type correctly (unless you believe that the correctness guarantees from static typing pay off for the code you're writing--e.g., code in rarely exercised error handlers).
I really don't see why there is a need for everyone and their dog to respond to Yegge's rambling blog post. Is language [X] conservative or liberal is now littering the HN front page.