[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Overuse of strings

This page is part of the web mail archives of SRFI 83 from before July 7th, 2015. The new archives for SRFI 83 contain all messages, not just those from before July 7th, 2015.



I just felt obligated to put in my $.02 here and join in opposing the
string syntax for <lib-path>. It is decidedly un-schemish.

Strings are for I/O: they are character sequences that contain data that
is transmitted to humans, the operating system or to other processes,
They should not be used as identifiers (symbols and unique
heap-allocated objects are for that), nor as containers of structured
data (list structures are for that).

>From the SRFI:

(library "hello" "scheme://r6rs"
    (display "Hello World")
    (newline))

Here "hello" is used as an identifier: we are not the least bit
interested in the fact that it begins with 'h' and ends with 'o', we are
only interested in whether someone imports a library with exactly the
same name or not.

Likewise, "scheme://r6rs" is used to represent structured data: it is an
URI with the scheme "scheme" and the authority "r6rs". Again, the
textual contents of each of these components aren't interesting, merely
the facts that both of these are identifiers with a special meaning.

So I suggest

"hello"         -> hello
"scheme://r6rs" -> (scheme r6rs)

Only users of lesser programming languages are forced to stick with
strings to represent identifiers and structured data. We have symbols
and s-exps. Let's use them.


Lauri