Case-sensitivity: why -- or why not? (was Re: Damnation!)

[Blake Winton]

On 21 May 2000 05:45:16 GMT, Will Rose wrote:
>Mike Fletcher <mfletch at tpresence.com> wrote:
>I think the problem is not case-sensitive vs. case-insensitive, it's
>case-sensitive/case-preserving vs. case-insensitive/case-preserving.
>I have no problem reading COBOL or early BASIC in monocase (apart
>from the obvious problems of the languages).  It's when one token
>can have several different forms that the problems start.

Well, there we go.  Perhaps we can have something which is neither
case-sensitive, nor case-insensitive, based on the following principles.
1) Your variables can be named anything you want.
2) Any variable which differs from a preceeding one only by case
   generates an error.

It should be _very_ easy to tell which variable generated the error, and
how to fix it, but in order to "encourage" newer programmers to think
more methodically, I feel that the compiler shouldn't make the change
for them.  Perhaps the IDE, in beginner mode, would be nicer.

I suppose that's similar to case-sensitive, but I feel it differs in two
main points.  Firstly, the error messages would be much nicer to users.
Secondly, people wouldn't get confused by stuff like "abc=aBc", since
the second form would be disallowed.

>It would be nice if the language enforced reasonable default rules,

Something like Java's guidelines?

>I'd suggest enforcing the use of upper or lower case only, rather than
>mixed case.  That seems the least bad option; not much more of a
>nuisance than the current whitespace rules.

Speaking as someone who had to use Modula-3, Ugh!
(Modula 3's largest mis-feature, for me at least, was that all keywords
 had to be typed in ALL-UPPERCASE.  My little finger got really sore
 after four months of programming.  :P)

Move beyond the C64!
Bill gave us the shift key for a reason, let's use it!  ;)

Later,
Blake.
-- 
7:25am up 31 days, 11:36, 4 users, load average: 1.19, 1.06, 1.02