On Mon, Mar 04, 2002 at 01:59:45PM -0500, Tod Harter wrote:
> It is MUCH easier to write low level code that deals with fixed width chars, 
> yes. This is fundamentally why i18n is such a big deal is that its a LOT 
> harder to reliably write algorithms that have to deal with all sorts of 
> different ways of representing data! One standard fundamental data type, ie 
> short, int, long, whatever you want to call it makes things MUCH easier. 
> Trust me, I've been writing software for 20+ years, I know from personal 
> experience! 

Contrast that with the experience that it's MUCH easier to write
string-handling code without anything as weird as a "char" type (think
perl, think combining characters in unicode).

And then consider that you get the low level code that enables
"text-oriented" programming for free with every recent perl, gtk and a
bunch of other libraries.

The char type in programming languages should die, in its meaning as a
piece of a text string.


On a different note...

The main subject of this discussion, the inability to know which
character set is used in from submissions, is ironically "solved" by W3C
using UTF-8.

See http://www.w3.org/International/O-URL-and-ident.html and
http://www.w3.org/International/O-URL-code.html

At least the more recent W3C recommendations that use URIs make it a
point to mention the use of UTF-8 explicitly.

-- 
Bart.

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to