[567] in Coldmud discussion meeting

root meeting help first first in chain previous in chain previous next next in chain last in chain last

Re: Assignment Operator.

daemon@ATHENA.MIT.EDU (Tue Nov 8 19:47:20 1994 )

To: coldstuff@MIT.EDU
Date: Wed, 9 Nov 1994 05:40:50 -0500 (EST)
From: colinm@colinm.ozemail.com.au (Colin McCormack)
Reply-To: colinm@colinm.ozemail.com.au

I've thought about this, and have a serious compromise suggestion:

Make ':=' a synonym for '='.

Then we can measure the relative productivity of people who use
':=' over those who use '=' and finally stick a cork in whining
of the form: `9/10 clueless programmers want '.
I have no objection (as I've stressed time and again) to people
voting with their feet (or even typing with them.)

[Satire mode on, from this point.]

We could also make '=:' a synonym for both, then the keyboard-challenged
(or differently coordinated) could be protected from their own dyslexia.

In fact, I think there's a strong argument for an entirely new programmer
class, which accepts a much more general syntax as programmatic input:
valid programs should have the same domain as String currently has.
Unfortunately, the functionality of the language might have to be 
constrained, a little.

I look forward to the day when eval_cmd is as follows:

@program $wannabe.eval_cmd
  arg expr;

  .tell("Yes, " + .name() + ", I think that \"" + expr + "\" is right!");

In the long run, we could incorporate stylistic tips, in what I like to
call a but-clause, which might be implemented as a search for naked `='
and consequent output:  .tell("But perhaps you should use `:=' instead
of `=' for increased clarity.")

I honestly believe this facility would solve many of the problems.