Re: Some questions about Ada.

[bobduff@world.std.com (Robert A Duff)]

In article <dewar.831213757@schonberg>, Robert Dewar <dewar@cs.nyu.edu> wrote:
>I see it! I think it is terrible Ada style to be inconsistent in casing
>keywords or identifiers. In fact I think it is nice if the compiler has
>an option to prevent such sloppiness (when we compile in internal GNAT
>mode, consistent casing is enforced).

i agree.  it's not a matter of really misundERSTanding things, assuming
you know the rules.  it's just that doing things in an unconventional or
Sloppy way makes the code harder to reAD.

By the way, does the gnat switch disallow declaring two overloaded
things called, say, NonEval and NoneVal?  I assume that it merely
requires that each reference to one or the other is capitalized like its
declaration.

>However, I think that having case sensitivity in languages is a very
>bad idea. For one thing it makes it  very difficult to talk verbally about
>programs, and if people really start defining identifiers like
>Time_of_Day and Time_OF_Day to mean different things, then it damages
>readability badly.

I think it's already nearly impossible to express programs by talking,
so I don't find that argument interesting.  Whenever I'm sitting in a
design session, and somebody says a piece of a program out loud, it is
nearly always necessary to write it on the blackboard for people to
understand.

By the way, mathematicians use case sensitivity -- e.g., "Let X be the
set of all x such that x has so-and-so property."  How do mathemeticians
do that over the phone?  I suspect the answer is, they don't bother
trying.

- Bob