[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: What is a lightweight language

Shriram Krishnamurthi <sk@cs.brown.edu> writes:

> Dan Sugalski wrote:
> > Don't dismiss or discount psychological hurdles. Bad, *bad* idea. They're 
> > by far the biggest problem--the technical issues almost always pale in 
> > comparison.
> I suspect I have spent at least as much time as you have thinking
> about the psychology -- both perceived and observed -- of students
> learning to program for the first time.  We also developed a
> comprehensive programming environment -- DrScheme -- designed
> specifically to addresses the care of beginners.
> So, yeah, I'd say we have the psychology angle held down okay.
> I think Joe Marshall's objection is strange, and hopefully isolated.

It may be strange, but it isn't isolated.  Go on to comp.lang.lisp and
suggest that first-class continuations be added to Common Lisp.  The
general perception is that first-class continuations are unnecessary,
inefficient, difficult to implement and to understand.  There is the
perception that because scheme supports first-class continuations,
that it induces programmers to use them with abandon.

It is astounding how these sorts of myths persist even among people
who ought to know better.  (Of course I don't mean to paint everyone
who reads comp.lang.lisp with the same brush.)

The fact of the matter is that first-class continuations *aren't* used
very often in `standard' code, and that CATCH/THROW, structured error
handling, and a thread package will cover virtually all practical of
first-class continuations.

For this reason I don't much care if a language has first-class
continuations or not.  Sure, it is a bonus, and I *always* put them in
to languages that I implement (REBOL 1.0 has first-class
continuations), and they make writing the error handler and debugger
far, far easier, but the end user doesn't care.

> Maybe at MIT, they specialize in scaring their students by throwing
> big words like "continuation" (more than two syllables) at them.

MIT students are expected to learn three syllable words by the time
they graduate, but it is fairly easy to fake it if you have a good
dic... diction... er... Webster's.