That came out more combative and impatient than it should have. LBO isn’t the right place for such disagreements, anyway. I won’t be writing more on this thread. Especially since I am over quota!
—ravi
On Oct 26, 2011, at 9:26 PM, // ravi wrote:
> On Oct 26, 2011, at 6:36 PM, Tayssir John Gabbour wrote:
>> On Wed, Oct 26, 2011 at 5:54 PM, // ravi <ravi at platosbeard.org> wrote:
>>> McCarthy made great contributions to Computer Science, but I am glad that his (and Minsky’s) brand of AI, as well as Lisp, failed by and large (though these days there is a significant renewal of functional languages).
>>
>> Lisp succeeded after being on life-support, much like Apple. Take
>> Clojure, for instance; or Hacker News, a major tech industry hub which
>> is written in a lisp.
>
>
> I do note the recent trendiness of FP, Clojure, Scala, Haskell, etc. But by “failed” I mean that measured by millions of lines of code, or by importance of project (ranked along the lines of: kernel, compilers, protocol stacks, web browsers and servers, etc), and finally as pedagogical devices, Lisp and FP are either non-existent or share the space with other languages.
>
> This is not intended to imply that I hate Lisp/FP, think it’s a bad language, or any such thing. I just fear the possible obscurantism and priesthood they could introduce (as a science) in the craft of programming. Imperative languages are successful because they are intuitive. Users don’t need to talk about the lambda calculus (poor old Church, one of my idols) or monads or immutability, i.e., they do not need an advanced degree in CS, to do imperative programming. They can literally hit the ground running. And write good code. I recall watching undergrads, especially non-CS majors, wilting upon encountering Lisp (or having it forced on them), after their initial excitement with what they could do in a few lines of C (Perl wasn’t popular back then).
>
> I don’t buy Crockford on Javascript being “really” a functional language (just accommodating the whole lot of curly brace lovers, to use Steve Yegge’s term for us proles of programming), any more than I buy his idea that playing tricks with closures somehow elegantly provides private variables. He is very good - technically and as an author, I will give him that.
>
> Nor do I deny that ideas such as immutability or first-class functions haven’t been useful when borrowed into imperative languages.
>
>
>> But I certainly don't mean to get into a flamewar about who wields the
>> bigger instrument — this hypermasculine industry is quick to label
>> ideas as winners and losers, based on what happened in the market.
>
>
> i.e., I am hyper masculine and trolling for a flame war on dick length, because I used the word “failed"? Unix, TCP/IP, the web - these are not “markets”. These are reality.
>
> In many ways, this is similar to the content of the "OWS vs Panel” threads.
>
> —ravi
>
>