do...until wisdom needed...

[Douglas Alan]

"Alex Martelli" <aleaxit at yahoo.com> writes:

> > I know one thing for sure: I know more about programming languages and
> > programming language design than you do or ever will with your
> > attitude.  Regarding listening to Alex, I'd listen to him if he
> > weren't also a prick.

> Possibly.  On the other hand, I seem to be slightly more refined
> than you at Usenet flamewars, judging for example from the styles
> used to insult -- rude, direct name-calling on one side, subtler
> indirect venom on the other.

I didn't come here for a flame war.  I came here for an intelligent
discussion, but if you're going to be an asshole trolling for a flame
war for no good reason, then I am going to call you on it.

> > I'll tell you who I've listened to quite carefully: Guido, who has
> > a lot of good ideas, and the designers of Lisp (eg, Dave Moon,
> > John McArthy, Guy Steele, all of whom I've met

> You may have been so busy _listening_ that you never cared
> enough to *read* anything by or about the second one of these
> guys 

I read the Lisp 1.5 manual, among other things.

> -- pretty hard, otherwise, to understand how such an outstandingly
> brilliant individual as you could possibly have so laughingly failed
> in such a simple spelling task.

I have dyslexia.  Are you going to pick on gimps as an encore?  Btw, I
did look up the spelling, and the source I found had it mispelled.  It
didn't look right to me, but that's what life is like when you're
dyslexic.  Sometimes the word "the" doesn't look right.

> Spelling flames are tacky,

Yes they are, as are you, in general.

> In particular, I wonder if you have read his 1980 paper he still
> holds up (as of Jan 2001) as representing his current thinking
> on LISP.  His comments there on POP2 succintly express his belief
> that 'algol-ish' syntax (Python's is similar) is inappropriate
> for a language where as in LISP "anyone can make his own fancy
> macro recognizer and expander".

No, I haven't read that paper, but I do know that Lisp 1.5 had an
"algol-ish" syntax, so he didn't necessarily always think this way.

> Also the note that even in LISP "certain macro expansion systems"
> lead to "over complicated systems when generalized" (although they
> "always work very nicely on simple examples") is interestingly
> topical.  Fancy seeing you quote the guy (albeit in a mis-spelled
> way) and asserting you have "listened carefully" while apparently
> ignoring his words.  One might wonder if the LISTENING has resulted
> in any amount at all of UNDERSTANDING.

Are you trying to say that McCarthy is opposed to procedural macros?
Or is he only warning that every coin has two sides?  I note the word
"certain" in your quote, rather than "all".  Neither "listening", nor
"understanding", by the way, mean "agreeing whole hog".  I refer you
to any dictionary.  And I would be extremely surprised if Dave Moon
turned out to be opposed to procedural macros.

This is what I do understand from what I've heard, from what I've
read, and from my own thoughts: Every feature you add to a language
has its pitfalls and so should be considered carefully.  Conversely,
many times if you take something away from a language in the name of
simplicity, you merely transfer the complexity onto programs written
in the language.  If a language is too simple then some problems need
to be solved over and over again by the users of the language, and
it's sometimes best to solve the problem once and for all and put the
solution into the language.

Counter to what you seem to think, your ideas, Mr. Martelli, are not
the word of god.  Many different points of view can be reasonably held
by different people -- on language design in general, and on Python,
in specific.

> Of course, the designers of _Dylan_ (who can hardly be accused of
> being LISP-ignorant:-) disagree with this stance -- they built a
> powerful language with algol-ish syntax and macros (as we as truly
> powerful underlying semantics, multi-methods first and foremost) and
> with it they've been trying to take over the world for quite a bit
> longer than Python has.  Reflecting on the similarities and
> differences, and the different amount of success that has so far
> (the first 12 years or so...) smiled on both endeavours, might
> perhaps (to a person of refined perception) prove more fruitful than
> _just_ reading about theories.

Languages succeed and fail for all sorts of reaons other than on their
strict merits as a language.

> It appears to me that the crucial design-goal that
> Python and Dylan did not share was *SIMPLICITY*.

I beg to differ.  Dylan is a relatively simple language as far as
languages go.  Also, I am currently working with a language called
Proto that is far simpler than Python, and yet it has both
multimethods and procedural macros.

> > personally), who are far smartter than you, me, or Alex will ever be.

> There is (fortunately, I think) no single axis along which
> "smartness" is measured (cfr. Gould, "The Mismeasure of Man", IMHO
> still his masterpiece, for a very well written and deep examination
> of pseudo-scientific attempts to establish such an axis, all the way
> up to that fraud named "the IQ").

You preach to the choir.  I'm not aware of Howard Gardner including an
"obnoxious wise-ass" category to his list of multiple intelligences,
and I can't find it in the index to Minsky's *Society of Mind*, but if
such a category were to be recognized, you would certainly be near the
top.

> An interesting, if idle, pastime, is to wonder whether the variety
> of human endeavours can possibly be so vast that at some of them
> even _YOU_ might be called 'smart'... sure, the mind boggles, and
> yet...

Oh, my, Mr. Martelli, you are just so clever.  I wonder if I might *ever*
aspire to such incredible wit and cunning.

|>oug