[r6rs-discuss] JAR statement
andrew-scheme at areilly.bpc-users.org
Sat Feb 21 18:48:09 EST 2009
On Fri, Feb 20, 2009 at 08:25:39PM -0800, Arthur A. Gleckler wrote:
> > Oh, THEM... well, good then. They don't come
> > around here. There's nothing to worry about.
> Not so fast. My point was that Java is an example of what it's like
> to have case sensitivity, and talking about code in Java requires
> being specific about the case of every identifier.
This is a poor argument. Java code is that way not because the
language is case sensitive, but because early-on the guardians
of the language decreed that such use of case would be
"preferred Java Style". (And they learned that from C++. I
don't remember where C++ went off the rails. Early C++ code
wasn't infested with camel-case, as far as I remember. Hmm.
Smalltalk uses cammel case a lot. Maybe that's whre it came
C is also case sensitive, and the archetype for C is Kernighan
and Ritchie's style, which almost never has that kind of
distinction. There, the only style-inspired use of case is
all-caps for #define macros (which is not universally applied,
of course.) Similarly, Eiffel is a case sensitive language
with lots of use of types in definitions, and its preferred
style doesn't seem to lead to much in the way of grotesqueries
(lower-case everywhere except for types, which are upper case.)
> Marc's excellent proposal of a special syntax for case-sensitive
> symbols solves the problem of interfacing with case-sensitive
> languages without requiring that we give up the benefits of case
> insensitivity for the vast majority of symbols that don't require
> distinguishing case.
There aren't any benefits of case insensitivity. Simplicity is
a far more evident benefit of case sensitivity, IMO: A symbol
is just what it is, no other combinations of characters match
More information about the r6rs-discuss