[Esip-preserve] Stewardship Best Practices - Identifiers

[Curt Tilmes]

First, I'll confess I haven't gone through everything you sent in
great detail yet -- I'll try to do so at some point.  Anyway, my
apologies if my points are addressed and I missed them with my cursory
read.

Second, I just want to point out that I agree with Ruth, there are
several different issues (the various use cases) we are tackling at
once.  We may (likely will) end up with different answers for
different problems.  A problem that prevents the proposed answer for
problem A from addressing the needs of problem B are known and ok.  We
accept them.  We acknowledge that one answer probably can't be found
that will address all the problems (though we're open if someone wants
to suggest one!).  We are really trying to pick a set of solutions
that address the needs as a whole, even if those solution involve
multiple simultaneous approaches.


On 10/07/2010 02:35 PM, Bruce Barkstrom wrote:
> [summarizing: you re-arranged the bits and got a different hash]

Of course.  You need a defined canonicalization algorithm prior to the
hash, something comparable to the UNF approach
(http://thedata.org/book/unf-version-5), though we'll both agree that
that canonicalization is insufficient for our needs.

It is at least conceivable that we could adapt such a canonicalization
for, for example, HDF, that would result in the same hash for
equivalent data.  It sounds like something like that was presented at
the HDF meeting ("the HDF mapping approach")?

> c.  As an interim approach to identifiers, I think it would be
> straightforward to use the current naming conventions and hook them
> up to some of the identifier schemas identified in the paper.  DOI's
> that resolve to file or collection names are probably as good as
> anything else if we're looking for simple identification.
>
> However, I do not think this kind of approach should be sold as a
> method of concocting permanent, unique identifiers.

We agree.

I think our proposed use of DOI is as a global, unique, resolvable
identifier for (in EOS-speak) "ESDT+Collection".

They are not sufficient (in general) to identify precise granules, and
we aren't proposing to use DOIs in that way.

> Hope this is helpful.  We've got a lot of listening to each other to do.

Yes.  I'll try to read through all you sent in more detail...

Curt