Thanks, but that's not a full solution---how does the application (the Wiki, the newsfeed---whatever) know whether to process and interpret tags? If it were really plain text, the "this is <em>really</em> cool" could be a plain text example of how to use to use markup, rather than actual use of markup.

Garret

Spartanicus wrote:
Garret Wilson <garret@globalmentor.com> wrote:

  
A common use nowadays, especially with wikis and newsfeeds, is to store 
XHTML fragments (such as "this is <em>really</em> cool") to be later 
integrated into a larger XHTML document.

What content type should we use for XHTML fragments?
    

As long as the application that processes the fragments handles it
properly it shouldn't matter, as these fragments should never be
accessible in other ways.

I'd use text/plain