Sorry to weigh in late on this. Here's the way it works in all 3 current
drafts:
1) Per our conclusion in early January, the Syntax spec defines the $
namespace and says that it can be populated by a) XRI TC specs, b) other
OASIS specs, or c) (using cross-references) other standards bodies. The
Syntax spec also refers to: a) the Res spec for $ names for resolution, and
b) the Metadata spec for $ names for XRI metadata.
2) The Metadata spec refers to the Syntax definition of the $ space, says
that it (the Metadata spec) defines $ names for XRI metadata, and says that
the Res spec defines all $ names for resolution under the top-level "$res".
3) The Res spec defines "$res" and all identifiers under it.
I believe this delegation model is pretty clean. The Res spec can own "$res"
forever. The Metadata spec can own $l, $d, $v, and $-. Other OASIS specs can
share the $space as needed by spec'ing their own $ names and then going
through the OASIS review process (which should be relatively easy to
manage). And other standards bodies can establish and manage their own $
spaces use cross-references.
=Drummond
-----Original Message-----
From: Peter C Davis [mailto:]
Sent: Wednesday, March 09, 2005 1:56 PM
To:
Subject: Re: [xri] Groups - XRI Metadata 2.0 Working Draft 04 (Word)
(wd-xri-metadata-04.doc) uploaded
On Wednesday 09 March 2005 04:26 pm, Wachob, Gabe wrote:
> Peter-
> I think that would be a little confusing, actually. I think in
> metadata we are explicit that $ namespaces can be defined in other
> specifications. We do *not* (I believe) reference *from* the metadata
> spec to the resolution spec -- that was done to prevent a circular
> dependency between metadata and rez. However, I'm fine with such
> circular references and maybe that'd clear it up.
>
> If we don't want to change this, I'd be happy with the status
> quo as well.
>
well, the MD spec does normatively (i think) ref Resolution already. so
we've
already made the dependancy. i'm ok w/it as is ... this DOES mean tho that
we need the OASIS registry folks involved if other specs (outside XRI) can
define these (we had language which had reserved the '$' namespace for
XRI-TC
use only. i do not recall seeing that in the MD spec (it used to be in the
XRIv1.0 spec tho).
--- peterd
---------------------------------------------------------------------
To unsubscribe, e-mail:
For additional commands, e-mail:
