followup on yesterday's discussion of the namespace proposal

[Julian Reschke <julian.reschke@greenbytes.de>]

Hi,

in yesterday's telco, we heard concerns about potential overlap between 
the proposed new globallyUniqueName attribute, and existing attributes 
like "id" and "package".

1) It appears that there is no "package" attribute; it may have been 
discussed earlier, but I can't find it in Draft 61a.

2) Regarding the potential overlap with "id"... That attribute is 
currently defined as:

	"This attribute contains a system-assigned ID which
	uniquely identifies this property."

So this would not work, as the whole point of the globallyUniqueName is 
that it can be used to have the same identifier across multiple 
repositories and protocols; so, by definition, it can't be 
system-assigned -- maybe we can just relax or rephrase that requirement, 
though.

Also,

3) for consistency with other object types like documents, the new 
attribute should probably be present on all object types (for instance, 
in JCR, node types are defined as expanded names as well).

Feedback appreciated,

Julian

-- 
<green/>bytes GmbH, Hafenweg 16, D-48155 Münster, Germany
Amtsgericht Münster: HRB5782