Re: draft-nottingham-atom-format-01

[Bill de hÓra <bill@xxxxxxxxxx>]

Tim Bray wrote:

> On Dec 13, 2003, at 7:52 AM, Bill de hÓra wrote:
>
> > Please mandate that Atom markup shall not use the default namespace 
> > mechanism. IME default namespaces are a crock - they don't play 
> > nicely, if at all with regular XML.
>
>
> No.  See http://www.ietf.org/rfc/rfc3470.txt (beneficial reading for all 
> on this list IMHO) but in particular section 4.3.  There is no way to 
> check this restriction in standard conforming software, so it's out of 
> bounds. -Tim

I'm aware of that RFC, it doesn't change the assertion re default 
namespaces, it doesn't address mixing unamespaced and default 
namespaced markup. I've yet to see an instance where an XML carrier 
format using default namespaces was a good idea. I also believe that 
software can detect whether ns prefixes are being used. Nonetheless, 
would you compromise on a 'should' directive?

Bill de hÓra