Email list hosting service & mailing list manager

perhaps I've missed something ... John Clements (20 Jan 2000 22:21 UTC)
Re: perhaps I've missed something ... Lars Thomas Hansen (20 Jan 2000 22:38 UTC)
Re: perhaps I've missed something ... Shriram Krishnamurthi (20 Jan 2000 22:52 UTC)
Re: perhaps I've missed something ... Lars Thomas Hansen (20 Jan 2000 23:02 UTC)
Re: perhaps I've missed something ... John Clements (20 Jan 2000 22:58 UTC)
Re: perhaps I've missed something ... Lars Thomas Hansen (20 Jan 2000 23:05 UTC)
Re: perhaps I've missed something ... John Clements (20 Jan 2000 23:12 UTC)
Re: perhaps I've missed something ... sperber@xxxxxx (21 Jan 2000 07:38 UTC)
Re: perhaps I've missed something ... Lars Thomas Hansen (20 Jan 2000 22:44 UTC)
Re: perhaps I've missed something ... John Clements (20 Jan 2000 23:09 UTC)
Re: perhaps I've missed something ... Per Bothner (20 Jan 2000 23:01 UTC)
Re: perhaps I've missed something ... Matthias Felleisen (20 Jan 2000 23:18 UTC)
Re: perhaps I've missed something ... Per Bothner (20 Jan 2000 23:55 UTC)
Re: perhaps I've missed something ... Matthias Felleisen (21 Jan 2000 01:04 UTC)
Re: perhaps I've missed something ... Per Bothner (21 Jan 2000 01:49 UTC)
Re: perhaps I've missed something ... Matthias Felleisen (21 Jan 2000 02:40 UTC)
Re: perhaps I've missed something ... thi (21 Jan 2000 09:58 UTC)
Re: perhaps I've missed something ... Per Bothner (21 Jan 2000 18:36 UTC)
Re: perhaps I've missed something ... sperber@xxxxxx (22 Jan 2000 10:32 UTC)
Re: perhaps I've missed something ... Per Bothner (23 Jan 2000 20:02 UTC)
Re: perhaps I've missed something ... Shriram Krishnamurthi (23 Jan 2000 20:50 UTC)
Re: perhaps I've missed something ... Per Bothner (23 Jan 2000 21:25 UTC)
Re: perhaps I've missed something ... sperber@xxxxxx (24 Jan 2000 07:30 UTC)
Re: perhaps I've missed something ... Michael Livshin (24 Jan 2000 16:55 UTC)
Re: perhaps I've missed something ... sperber@xxxxxx (25 Jan 2000 07:43 UTC)
Re: perhaps I've missed something ... Michael Livshin (25 Jan 2000 11:02 UTC)
Re: perhaps I've missed something ... sperber@xxxxxx (25 Jan 2000 11:31 UTC)
Re: perhaps I've missed something ... Matthias Felleisen (25 Jan 2000 13:47 UTC)
Re: perhaps I've missed something ... sperber@xxxxxx (24 Jan 2000 07:29 UTC)
Re: perhaps I've missed something ... John Clements (20 Jan 2000 23:59 UTC)
Re: perhaps I've missed something ... Per Bothner (21 Jan 2000 00:18 UTC)
Re: perhaps I've missed something ... Shriram Krishnamurthi (21 Jan 2000 00:03 UTC)
Re: perhaps I've missed something ... Per Bothner (21 Jan 2000 00:37 UTC)
Re: perhaps I've missed something ... Shriram Krishnamurthi (21 Jan 2000 08:39 UTC)

Re: perhaps I've missed something ... Per Bothner 21 Jan 2000 00:18 UTC

John Clements <xxxxxx@cs.rice.edu> writes:

> You believe that an "economy of concepts" may be achieved by merging
> set!  and set-car!.  I believe that this is fundamentally misleading.

To clarify:  I don't think merging set! and set-car! *per se*
leads to "economy of concepts".  However, it may be a more economical
way to *organize* the concepts.

Alternatively, one can argue it requires one big new concept
"location" (or lvalue if you prefer) (which is implicit in the
semantics anyway); once you have that concept, it unifies or
simplifies a lot of getter/setter pairs, which leads to fewer concepts
*total*.

I don't think it makes much difference either way formally.
The question is which style is more natural in practical use
I can see this is largely a matter of taste and personal preference.
And a matter of different cultures.

> If this were the case, then set! could presumably operate on any
> environment, and not simply the current one.

I was trying to make the point that the two forms of set! are
related, not give the semantics of anything.

> I claim that rather than a first-class entity, the environment is
> more or less an implementation detail; the clearest model to use in
> teaching programming is that of algebra, and application by
> substitution.

Except the top-level environment, which R5RS at least suggests (but
does not require) be a first-class entity.
--
	--Per Bothner
xxxxxx@bothner.com   http://www.bothner.com/~per/