🚀 go-pugleaf

In article ... tswingle@oucsace.cs.ohiou.edu (Tom Swingle) writes:

   1.  Do you prefer programming in a language which is case-sensitive or case-
   insensitive (or does it even make a difference to you), and why?

It depends.

If the language designer has got the names of the built-in keywords and
identifiers right, case-sensitive is my preference. 

By ``right'', I mean lower or mixed case; not UPPER CASE; IT'S TOO HARD TO READ
AND I'M DEAF ENOUGH ALREADY.

   2.  How significant an issue is case-sensitivity in your preference toward 
   one language over another?

It wouldn't make me pick a sematically or syntactically inferior language, just
because it was case-sensitive; it might well induce me to write a suitable
filter. 

   3.  Since this article is posted to the C and Pascal newsgroups, here is a
   question for the respective users of each language:  To the C programmers,
   do you feel that you would like C better if it were case-insensitive; and 
   to the Pascal programmers, do you feel that you would like it better with 
   case sensitivity?  

I've played both games, and with different flavours of Pascal cases. As a C
programmer, I'd hate it if it were case-insensitive; of course, that might be
becuase of all the code that would break. (I certainly willfully exploit
case-sensitivity when writing C.)

I'd certainly prefer Pascal to be case sensitive (isn't it? I can't recall), so
long as keywords and built-in names weren't made upper case.

   4.  If you were designing a language, would you make it case-sensitive, 
   case- insensitive, or maybe even allow for an option to go either way. 

All the languages I have designed (ur ... that's three, I think), and most of
the ones I have been co-designer of, have been case-sensitive. I'd not have an
option to switch.

The only arguments I have heard against case-sensitivity have been:

(a) You have to remember the case of the letters, not just the spelling of the
name. 

Response: so long as conventional use of case in names prevails, this should
not be a problem. Certainly if names such as ``cooRdinaTeS'' were used, it
would be HoRrIblE. But the usual use of case distinction falls into a rather
smaller set of patterns, and these patterns often correspond to differences in
use -- for example, leading Capitals for Type names.

In any case, similar objections apply to the common use of abbrevs in progs --
one persons meat being another pers poison; is it pnt for pint or point? Env or
envir for environment? rm, rem, or remove for delete? And have you had (as I
have had) problems with British vs American spelling -- as in color, flavor,
-ize where I'd have (and my reflexes provide) colour, flavour, -ise?

(b) Names differing only in case can be confused.

Yes, they can. But it doesn't happen very often. And names differing in a
single character can often be confused, too, but we don't make a linguistic
brouhaha about it. I'll happily use ``File'' for the type, and ``file'' for a
variable of that type, and not expect myself, or other readers, to be confused.

(c) But it fails the telephone test!

How often do *you* read programs over the telephone? When the case is
important, don't you add case markers -- ``capital String'', ``big FILE'', or
voice change? The telephone test is a hint in the right direction, but it's not
perfect; heavens, what about ``for'', ``four'', ``4'', ``fore''? Am I not
permitted to call a variable ``#'' -- sorry, ``hash''? How do *you* pronounce
``+'', or ``==''? [OK, these aren't strictly fair; after all, we can use
syntactic clues to work out what was meant. But code is primarily visual, not
vocal; let's respect the strengths of the different media.]




--

Regards,    | ``"I can't suit myself," said Weinbaum, a little petulantly.
Kers.       | "I work for the Government".'' - Blish, "The Quincunx of Time".