Re: WOFF and extended metadata

On Mon, 2010-06-21 at 21:03 +0100, Laurence Penney wrote:

> For the sake of a guaranteed, predictably simple document structure.

Since there's already hierarchy, dealing with hierarchy isn't the issue.
Encoding strings with < and > does not simplify - it merely moves
the complexity from people who write specifications and software down
to document authors who are often much less experienced. 

> One danger: What if the included unencoded XML uses elements already  
> used in the metadata? Some clients would barf on the unexpected  
> attributes and content of those elements.
Then they need to be fixed, if such clients actually exist.

What about clients that crash if there is a "w" in a value name? You
don't ban "w", you fix the clients.

A plausible compromise might be to restrict general values to XHTML
markup (which includes Ruby and right-to-left bidi overrides I think)
and to leave vendor or application-specific extensions to another place.

But if a user agent doesn't know how to render it, it needn't.

Anyway, my goal in this thread is just to make sure the needs of
internationalization and accessibility are met.  Crazy stuff like
encoding markup, or like having values that can't contain bidi or ruby
markup, or that can't easily be processed by text readers, will, as
Chris has said, come back to bite us later.

Part of my role at W3C is to try to make sure that people "play nice"
with XML...

Best,

Liam

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org www.advogato.org

Received on Tuesday, 22 June 2010 22:44:54 UTC