Re: [Lynx-dev] non-ascii characters in URLs

[Keith Bowes]

Je 2012-Jan-03 je 15:01, David Dorward skribis:
> 
> <!ENTITY % URI "CDATA"
>     -- a Uniform Resource Identifier,
>        see [URI]
>     -->
> 
> Where CDATA is defined as:
> 
> > CDATA is a sequence of characters from the document character set and may 
> > include character entities. 
> 
> With a note:
> 
> > For some HTML 4 attributes with CDATA attribute values, the specification 
> > imposes further constraints on the set of legal values for the attribute 
> > that may not be expressed by the DTD.
> 
> This is definitely one of the latter, so what the validator says in this 
> instance is not very relevant. 
> 

Does the spec limit the characters allowed in %URI;?  I suppose though
is to update the obsolete standards (URI names, HTTP, etc.) to address
two big issues:
1. URIs nowadays are often IRIs.
2. UTF-8 is the de facto standard encoding nowadays.  To impose a
   default of ISO-8859-1 just seems odd in our modern world.

If HTTP were to be updated, I'd recommend to also use a more modern date
format used by e.g. i.a. Atom.