Filtering and identifiers and logging.

[Rod Widdowson]

I've dug into the internal name handling in V2 some more and I am pretty
sure the idea is to isolate the "filter" namespace from the "spring"
namespace. Further it isolates names used in one <AttributePolicyGroup> from
others; finally I *think* that ensures that two objects of different types
can have the same (user visible) name.


'/AttributeFilterPolicyGroup:PolicyExample3/PolicyRequirementRule:MyAndRule'

Is /[1]:[2]/[3]:[4]

Where
[1] is the type of the top level element
[2] is the id of the top level element
[3] is the type of the element in question
[3] is the Id of the element in question or a random ID

> Hmm, I don't really recall all that

I think that it never was exposed because the filter IDs were never exposed.

> so I don't know what the purpose was.
> I guess he didn't just use IDREF and link things by IDs directly?

No, they were uniquified (as per above)


> I think the stack traces are a lot more obnoxious, and honestly it's for
> occasional debugging, not routine logging anyway.

Fine,  I'll leave as it.  If people complain we can always fix that.