Entities and Char sets

[Julian Bond <julian_bond@voidstar.com>]

There's a debate happening in at least two other XML forums about the
problems of entities and char sets. As usual there are at least 5
players in this game.

- Content Author. They don't care about the transport. They just want to
communicate. And if that means using the Mac high order "'", they'll
just do it. They frequently use & signs and occasionally they use a <
sign. They may know enough to convert to &lt; if it's aimed at HTML but
not always.
- XML Toolkit author for senders, or the RSS creator. All they want is
for the content to get through. They may or may not have read the specs.
More likely is that they just copied someone else's code or a bit of
sample data.
- XML Toolkit author for parsers, or the RSS reader. How little can I
install and still read all the RSS out there and in whatever RSS
flavour. Oy! These people double encode entities. Vey! Those people use
the Netscape DTD and it's disappeared. Karamba! What's that entity and
where did *that* come from? Pants! Unescaped & in the <title>.
- Content Reader. What was the author trying to tell me? Caf&eacute;
What does that mean? "We were talking about the issues with the [snip]
element". where [snip] = <title> and just never appeared.
- XML Bigot, sorry Pedant, sorry Guru. You WILL NOT accept bad XML.
Well, yes, but I wanted to read the content.

With RSS, trying to get the right DTD, language, encoding is not that
bad, because 9 times out of 10, the reader is tolerant and the content
ends up in a tolerant browser. But when you're trying to get the content
down an XML-RPC pipe and the toolkits don't interoperate because they
barf on whitespace between tags or tell you the name-value is missing
when actually it's just empty, it all becomes a huge PITA.

So what is the *correct* way of dealing with reserved and high order
characters in RSS content, such that the maximum number of clients can
read it and get the content to the reader? 

-- 
Julian Bond    email: julian_bond@voidstar.com
CV/Resume:         http://www.voidstar.com/cv/
WebLog:               http://www.voidstar.com/
M: +44 (0)77 5907 2173  T: +44 (0)192 0412 433
ICQ:33679568 tag:So many words, so little time