It was pure genius for Apple to adopt NextStep and Unix, and it probably saved the company. I’m all for language experimentation, but it would have been more much-needed genius had Apple instead recognized the emerging importance of Python, and bet the bank on something like appscript for Python, instead of growing their own language.

Perl predates Python, and was highly original in its time. Yet, as others have observed, its author is a linguist, Python’s author is a mathematician, and it shows. That AppleScript was designed to be a more “human” language for “the rest of us” (how many people have switched to Linux out of contempt for that phrase?) is its fatal flaw. Like Perl, it lacks a mathematical predictability for its syntax and semantics.

If I had to cite three programming languages that one should understand, master and maintain fluency in, I’d list C/C++/Objective C (to understand how a computer most efficiently works), Haskell (to understand how the human mind most efficiently works) and Python (to get things done, if you can afford to have your code screw up at runtime). One writes better Python code, coming off a stint of Haskell programming than coming off a stint of C programming. (Yes, I’ve coded in Scheme, Ruby, OCaml, … and I stand by this list.)

From this point of view, appscript is a crucial development, because it works with a language everyone should know, not wasting their time with a different language that isn’t worth their trouble.

i agree with your assessment of appscript — i’m very enthusiastic about it — but i’d like to offer a couple of counterpoints to the larger picture you paint; this is not to take away from appscript, but to add some appreciation for where it stands and how it can develop

i’m not sure what “niche” you refer to above, so i think it’s worth mentioning Frontier, which predated AppleScript, has a broader scope than AppleScript (all the interapp capabilities, much more built-in functionality, and later a complete TCP/IP suite & server); Frontier has had two periods, including the present, in which it was free (the core is now open-source at frontierkernel.sourceforge.net); i like Frontier’s no-nonsense syntax, object database and outliner editor; Python/appscript is the logical next step for Mac Frontier programmers, and while Python is of course much more sophisticated, i think it could still benefit from Frontier’s outliner IDE concept and approach to persistence

also while i agree that AppleScript as a language has major usability flaws, i sputtered a bit at “perhaps the worst” — it has a unique, dynamic and expressive (if subtle, contorted and ambiguous) OO design that is much more sophisticated than so called “OO” VB, for instance; one might also question how AppleScript might have developed if it really had ever reached a “wide audience”; AppleScript is only now reaching belated maturity, hunchback wizard that it is; still, Python + appscript will need more integration and support� it can supplant AppleScript for sheer convenience

finally, as you seem to acknowledge, anyone who might make headway with appscript in its current state will really need some respect for and comfort with AppleScript; for appscript to develop into the core Mac OS X interapp tool it could become, i think a synergy will be needed, such as direct ways to address Python/appscript objects from AppleScript