RE: [xri] Very rough algorithm outline

["Scott Cantor" <cantor.2@osu.edu>]

Drummond Reed wrote on 2009-06-17:
> I will put it on the agenda for tomorrow's telecon. Again, I urge everyone
> concerned with this issue to attend, including Scott if he can make it.

I'm at a conference that I drove 7 hours (and begged funding) to attend, so
I would say the likihood I'll be getting on the phone to argue about this
is....low.

I do not really believe that people are looking at this problem from a
suitably constrained perspective. Arguing about whether you can implement
XML Signature (or maybe I should say c14n) in Ruby is beside the point. The
question is whether you can write code to serialize XML with some strict
namespace logic.

The goal is to be compatible with the full excl c14n spec, not to require
implementing it. I don't feel certain that this distinction has been getting
across.

With respect to interop, if you write some code and want to be sure it's
correct, there are plenty of test vectors and Java/C++ implementations to
use to compare against. I would completely agree that absent that step, the
effort is likely wasted.
 
-- Scott