[Esip-preserve] On Earth Science Data File Uniqueness

[Curt Tilmes]

On 02/09/11 17:14, Bruce Barkstrom wrote:
> Curt's approach is a bit pessimistic.

Well, it's not really my approach -- It's the approach recommended by
the identifiers assessment activity and described in the conclusions
of the paper.  I'm just advocating for it.

What aspect of that approach is pessimistic?

> One approach would be to recommend that anyone creating a UUID
> should put it in a tabulation of created objects and perhaps make
> that tabulation available to some socially responsible registration
> authority.  This approach wouldn't require real-time registration,
> but would at least let someone know that an object with the ID had
> been created.

I think that would be great, but we've never had a lot of luck with
central registrations.  (Perhaps ECHO?)

BTW, The ITU operates such a registry:
http://www.itu.int/ITU-T/asn1/uuid.html
They have a somewhat complex procedure for the registration.

My problem now is that during bursts of reprocessing, my own database
can barely keep up with the rate at which I create objects, much less
could someone else's central server do so for our whole community.

I'm not certain the use case of "I have an object with an identifier,
I have no idea where I got it, who is responsible for it, or what it
is, but I need to know those things" is a common one.

Even if we managed to come up with resources to create such a service,
what exactly would it do?  How would you keep it up to date?

<brainstorming>

Perhaps we could record a bare minimum, mapping each UUID => DOI.

You could always use the DOI service as the primary resolver to find
the curator for the granule at some time.  Then just ask that curator
(with some well-defined service) about the granule.

Have to think about that more...

I still think we must proceed with our work without the assumption
that such a central service exists or will ever exist.

Curt