Re: Modern languages are case sensitive?

[Florian Weimer <fw@deneb.enyo.de>]

dewar@gnat.com (Robert Dewar) writes:

> a) the following program is quite awful, and should
>    obviously not be permitted in any language:
>
>      Main : integer;
>      main : integer;
>      ...

In C#, you can have

        int main;
        int main;
        int int;
        intother;

Of course, this is only the visual presentation. ;-) The actual
representation uses alternative representations of ASCII characters
(LATIN SMALL LETTER DOTLESS I followed by COMBINING DOT ABOVE) and a
ZERO WIDTH SPACE.

Clearly, Unicode is not suitable for identifiers.  There are five
different ways to represent a symbol which looks like a capital H!

> By the way, if I was writing a C compiler, I would include
> a warning if two identifiers differed only in case, so that
> the same desirable semantics is achieved.

Some case-sensitive OO languages (for example, Python) have quite a
useful idiom: capitalized identifiers refer to classes, all-lowercase
ones to temporary objects of the corresponding class.

> And I can't help but get a chuckle out of someone thinking
> of Unix as a "modern operating system". Indeed if you
> do look at modern operating systems (e.g. Windows XP, or
> OS/2), they are case insensitive.

AFAIK NTFS *is* case sensitive.  It's the Win32 subsystem which
presents a case-insensitive view of the file system.  Using other
subsystems, you can create names which only differ in case.

(Consumer versions of Windows were not even case-preserving, which I
consider a bug.)