Extensible property namespace work deemed insufficient; considering reversion.

Just added this comment to issue #1971[1] (aka "certain property names cause
non-wf XML responses", aka "the thing that spun cmpilato off on this
extensible property namespace business"):

{{{
Oh, man. Maybe this whole venture was ill-conceived. I just remembered the
other restrictions on XML Names, which include "cannot start with a digit"
and "cannot start with [Xx][Mm][Ll]". With this extensible property
namespace business in place, we'd still generate bogus XML for the following:

   myprop:2smart4u => <2smart4u xmlns=".../ext/myprop">
   myprop:xmlname => <xml xmlns=".../ext/myprop">

I feel like I've come back around to Ben Reser's 2004 reasoning: the
additional complexity I've added doesn't seem it worth it for the merely
partial fix it offers.

Unless I get encouragement to do otherwise, I will revert all this
extensible property namespace business from the codebase.
}}

Ben Reser, you tried to warn me -- I won't claim otherwise.

-- C-Mike

[1] http://subversion.tigris.org/issues/show_bug.cgi?id=1971

-- 
C. Michael Pilato <cmpilato_at_collab.net>
CollabNet   <>   www.collab.net   <>   Enterprise Cloud Development