As a collective response to email inquiries about where things stand
with PURLs, here is a summary of where I believe things stand.
Last November 19, I reported to this list that the purl.org server at
OCLC could no longer be accessed and that the PURLs identifying DCMI
Metadata Terms could no longer be maintained. In practice this means
that while the PURLs still resolve, redirects cannot be edited to point
to updated documentation and new terms cannot be coined [1]. (Attempts
to log in get the error message: "Sorry, the administrative interface is
experiencing technical difficulties. Login capability is currently
disabled.")
This problem affects not just DCMI Metadata Terms, but many other
Semantic Web vocabularies, such as Good Relations, as well as
vocabularies being developed under the DCMI umbrella for LRMI. The
problem casts a shadow over DCMI's current bid to update the
international standard ISO 15836, which includes PURLs. I have also
received email from people who have promoted PURLs in their communities
after OCLC encouraged their use at a DC-2014 session in Austin.
As of last November, discussion was taking place on two mailing lists
and the W3C Permanent Identifier Community Group [2] was developing the
idea for a Github-based service, w3id.org, to provide a "secure,
permanent URL re-direction service for Web applications", to be operated
by the Community Group (or rather, by organizations pledging support to
the Community Group).
There has been no further discussion on the Google Groups persistenturls
mailing list aside from an expression of anxiety from an organization
that has promoted the use of PURLs for identifying research grants and
scientific terms in their community [3].
Discussion has recently revived on the W3C public-perma-id list about
the mechanics of a Github-based service for managing redirects.
However, Norman Gray has asked the group for a reality check [4]:
> Can I just check what the goal is, here?
>
> Is it:
>
> (a) Create a w3id.org redirect for each/most purl.org redirect and
> abandon the purl.org ones in whole or in part?
>
> (b) Support purl.org as a domain -- ie, support all/most existing
> http://purl.org/.../blah redirects -- even if it's implemented by a
> similar mechanism to w3id.org?
>
> (c) Work towards a mechanism for creating new purl.org redirects by
> whatever means?
>
> I hope it's not (a) -- I feel it would look un-reassuring to abandon
> well-established persistent identifiers.
>
> Option (b) would surely be a minimum.
>
> I think there's a strong case for aspiring to (c). The purl.org
> 'brand' is a fairly well-known one, even if it's had its technical
> troubles in the last few years, and being able to support it to the
> extent of creating new purl.org redirects would be an impressive
> display of community commitment to the idea of persistent URIs.
He also writes [5]:
> My _impression_ was that OCLC was willing to pass on the purl.org
> domain, as well as the database contents, to a suitably constituted
> successor. I got that impression because I believe the widely assumed
> goal is at least (b), and that wouldn't be possible without a bequest of
> the domain.
This is really all I know about where the issue stands. We are of
course hoping for a good solution, though the continued uncertainty
means that DCMI, LRMI, and various other initiatives must continue to
put important decisions on hold.
Tom
[1] https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1511&L=DC-ARCHITECTURE&D=0&P=3711
[2] https://www.w3.org/community/perma-id/
[3] https://groups.google.com/forum/#!topic/persistenturls/Zpd4BHQxxIM
[4] https://lists.w3.org/Archives/Public/public-perma-id/2016Mar/0006.html
[5] https://lists.w3.org/Archives/Public/public-perma-id/2016Mar/0009.html
--
Tom Baker <[log in to unmask]>
|