Re: [office-comment] Section 1.5 - implementations SHOULD use the namespace prefixes given.

[Gavin Thomas Nicol <gtn@rbii.com>]

On Oct 25, 2004, at 11:24 AM, Uche Ogbuji wrote:
> I appreciate this implementation difficulty (I assume Qt DOM is purely
> 1.0, then), but I really think it's a bad idea to abuse the namespace
> REC by mandating prefixes.

I'd rather not abuse the *XML* REC by requiring their use!! You are now 
in a catch-22 position: either you *require* all OO format readers to 
support namespaces *and* XML, or you require them to just require XML. 
In the interest of interoperability, I know which one I would choose.

> A suggestion of prefix is OK, but not a mandate.

Why not? You're either going to have to mandate a prefix, or mandate 
use of namespaces which are *not* part of the XML specification. Why is 
is allowable to mandate additional complexity (and hence greater chance 
for bugs) in an application (i.e. widely-scoped architectural and 
implementation changes), when it is not allowed to mandate use of a 
prefix (which ultimately would only change the string constants used 
for comparison)? This seems to me to be yet another fallacious 
justification for their use.