RE: XHTML/XML comment

Thanks Arjun, Frank and Kent,
I wasn't thinking about i18n, which is the point, isn't it? Very
instructive. I wasn't complaining so much as I was curious; I also always
write in lowercase and tend to quote my attributes already, so I'm not
nearly as inflamed as some folks apparently are, nor do I plan to implement
XHTML by Thursday.
Thanks again,
Chris Luebcke

> -----Original Message-----
> From: www-html-request@w3.org 
> [mailto:www-html-request@w3.org]On Behalf
> Of Arjun Ray
> Sent: Monday, January 31, 2000 10:15 AM
> To: www-html@w3.org
> Subject: RE: XHTML/XML comment
> 
> 
> 
> 
> On Mon, 31 Jan 2000, Christopher Luebcke wrote:
> 
> > my question is why XML (and thus XHTML) was created as
> > case-sensitive in the first place (especially if neither SGML or
> > HTML share this characteristic).
> 
> Short answer:  Unicode.
> 
> There are scripts which don't have any case distinction at all; even
> among those that do, case equivalence is not a given (i.e. one form
> may exist but not the other); and the Unicode standard recommends that
> case substitution where needed should be to lowercase.  Interestingly
> enough, this recommendation is in direct contrast to SGML, where case
> substitution is to uppercase.
> 
> (It's actually a mistake to characterize HTML as case-insensitive.
> The standard requires all such names to be folded to uppercase.) 
> 
> The XML spec, having taken a hard line on internationalization from
> the beginning, couldn't keep SGML's case substitution rules in the
> face of Unicode's superior recommendation, and since case substitution
> *is* problematic in general, decided to eliminate it all together.
> 
> Given that, the specific choice of lowercase wasn't necessarily
> arbitrary.  The weight of technical evidence points to that being the
> best choice.
> 
> One thing to note, though.  If the W3C's protestations about SGML and
> XML are worth anything, then you're not obligated to switch, let alone
> stampede, to XHTML.  SGML is a lot about the continued viability of
> document and data formats.  Use the "older" standards if you must:
> they haven't gone away.
> 
> 
> Arjun
>   
> 
> 
>

Received on Monday, 31 January 2000 13:53:03 UTC