Re: [xml-dev] Ten Years Later - XML 1.0 Fifth Edition?

["Andrew Welch" <andrew.j.welch@gmail.com>]

On 18/02/2008, Elliotte Harold <elharo@metalab.unc.edu> wrote:
> Pete Cordell wrote:
> > Original Message From: "Elliotte Harold"
> >
> >> Perhaps there's an Amharic or Cambodian keyboard out there somewhere,
> >> but I really doubt any keyboard contain any characters added in
> >> Unicode 4.0 and beyond.
> >
> > I've seen East Asians entering some oriental font (I couldn't tell
> > which) using a standard US QWERTY keyboard.  So just because there isn't
> > a lump of plastic out there with the right characters on it doesn't mean
> > the characters can't be entered.
> >
>
> I wasn't saying that, but it was suggested (by someone else) that having
> the characters available on a keyboard was a good criterion for
> establishing which characters we must support.

I was just pointing out that it would be hard to process XML if I
couldn't easily type the name characters - if an element name
contained a symbol how would I go about entering that symbol?

Equally, presented with some XML with characters in names that aren't
immediately obvious could be a real pain, for example <foo-bar> is
that an endash, an emdash or a hyphen?

The same encoding-style detective work could be called for: <foo?bar>
is that really a question mark?  Or does my font not contain the
glyph?  Or perhaps the encoding used didn't have a mapping...

Currently if an XPath looks right but is selecting nothing you check
the default namespace, I'm wondering if I'll also have to check the
underlying bytes to work out the actual characters too.

Perhaps XML names could be considered the programming langauge of XML
- just as everyone programs in the same Java or HTML, everyone could
be perfectly happy to program in the same XML...


cheers
-- 
Andrew Welch
http://andrewjwelch.com
Kernow: http://kernowforsaxon.sf.net/