recommendations for persistent URL/resolver services?

[Chris]

Recent topics seem to suggest that purl.org domain approval is very high latency, and is granted in an unknown proportion of cases. That fact, combined with my own experiences with purl.org -- in which it has been intermittently down -- makes me wonder whether I should look to other organizations for persistent URLs.

From what I can tell, the main options are:

* https://w3id.org/
* handle.net
* Registering a domain and setting up one's own server to do redirects.

Can anyone provide recommendations? At the moment the first seems to be the most attractive choice, since handle.net is as expensive as domain registration plus a low-end virtual machine hire, and I'd rather not do recurring admin myself.

Thanks!

[David Wood]

Hi Chris,

It’s no secret that I like w3id.org because I’m involved in it. I became involved after OCLC became, IMO, less of a good steward of purl.org.

Callimachus (callimachusproject.org) is another project I’m involved in that implements PURLs - and many more options for PURLs than either the PURLz project or w3id.org. A single “partial” PURL was recently combined with “proxy” PURL functionality on Callimachus to provide redirects for all of Wikipedia in a single PURL definition. That is URL curation on a rather grand (and fun) scale.

Please let me know what you end up doing. Others share your frustration.

Regards,
Dave
--

http://about.me/david_wood

--
You received this message because you are subscribed to the Google Groups "persistenturls" group.
To unsubscribe from this group and stop receiving emails from it, send an email to persistenturl...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Monica Omodei]

But for those of us who started using purls many years ago when purl.org was the recommended domain unless you were running your own purl resolver server, we are sort of stuck with it now as by their very nature you can't switch to another resolver for an ongoing identifier service

I too have been underwhelmed since OCLC took over support for the purl.org resolver service. Their admin interface is broken in parts, and they have taken decisions to cut remove options but the documentation doesn't reflect the change. It can be highly confusing.

Monica

[David Wood]

Hi Monica,

I agree - and feel your pain. I suppose it is possible to redirect your redirects so you can switch services, but that would be a mess!

One minor correction: OCLC has operated and supported purl.org since its inception. It is only in the last few years that they have devalued support for it.

Regards,
Dave
--

http://about.me/david_wood

[Monica Omodei]

Ahhhh.. Thanks for the correction. I knew something had changed :-)

Maybe we need to find a way OCLC can get recognition for having provided this service in the beginning and committing to its ongoing support. Supporting 'persistence' is not something you can back away from. I know they are a non-profit cooperative so perhaps the membership need to be lobbied to support allocating some funding for better support. 

Because we are program funded, our own domain could not be guaranteed as being persistent so we needed to host our purls in another domain.  w3id.org did not exist at the time unfortunately.

Monica

Project Manager

Australian National Data Service

[Stian Soiland-Reyes]

I've been wondering about making a bot that can generate a "backup" of
purl.org for w3id.org - perhaps having https://w3id.org/purl.org/ as an
alternative namespace.

Then if worst come to worst, http://purl.org/* can just redirect to
to https://w3id.org/purl.org/* - management of the PURLs would then
have to be managed under w3id.org - which I know has a slightly
higher threshold level than the purl.org UI.


This would mean trawling the purl.org via the API and autogenerating
a folder structure with .htaccess and perhaps a minimal README.md
about who ha(d) ownership. Would that in theory be possible, and..
would it be legal?

Do anyone know much about using the purl.org APIs? They seem
mainly centered on batch updating via an XML format.



Excerpts from Monica Omodei's message of 2015-10-16 23:21:10 +0100:

--
Stian Soiland-Reyes, eScience Lab
School of Computer Science
The University of Manchester
http://soiland-reyes.com/stian/work/ http://orcid.org/0000-0001-9842-9718

[Norman Gray]

Greetings.

On 16 Oct 2015, at 23:25, David Wood wrote:

> One minor correction: OCLC has operated and supported purl.org

> <http://purl.org/> since its inception. It is only in the last few

> years that they have devalued support for it.

purl.org was the first time I'd heard of OCLC, back in 199-something
(I'm not a librarian, and this was when I was starting to get interested
in persistent URLs), and I thought 'what a bold, public-spirited and
interesting idea -- what a splendid organisation'.

The not-quite-death that they've now inflicted on purl.org -- exactly
contrary to the point of PURLs and to the intentions of the non-branded
purl.org domain -- does them absolutely no credit at all. And to
whatever extent they represent the library profession, they're an
embarrassment to it; I cannot understand how librarians could be quite
this bad at preserving access to information.

Even killing the service completely would be an improvement.

All the best,

Norman


--
Norman Gray : https://nxg.me.uk

[David Wood]

Hi Stian,

This is a good idea. The PURLz software APIs are documented at purlz.org - at least in source code. OCLC is running PURLz version 1.6.1 (I think).

The problem is that w3id.org doesn’t provide a user interface, which is a requirement for many users. It is easy enough for technical folk to make a .htaccess file for Apache and deal with git forks, but that is going to leave out a lot of users.

Maybe we just need another public PURL service? It would be nice if it were run by multiple organizations, like w3id.org, but have a UI. What do people think about that?

Regards,
Dave
--
http://about.me/david_wood

[David Wood]

Hi Norman,

I feel your pain :/

Regards,
Dave
--
http://about.me/david_wood

[Norman Gray]

David, hello.

On 28 Oct 2015, at 19:18, David Wood wrote:

> Maybe we just need another public PURL service? It would be nice if it
> were run by multiple organizations, like w3id.org, but have a UI. What
> do people think about that?

Another possibility might be to put up a web-based interface that
generates a patch for the w3id.org stuff, which is sent to either a
human or robot agent to apply to the master version. A first version
could be pretty low-tech; a later version could potentially generate a
github pull request.

w3id.org looks to me like a Good Plan, and the fact that it's (so
obviously) based on git is its only major wart. I'm currently using
w3id.org in a tentative way, as a purl.org replacement.

[Stian Soiland-Reyes]

Yeah, one thing that makes w3id.org work reliably is that it is based
on a bog standard Apache HTTP server with not very complicated
.htaccess rewrites. That means there aren't any SQL or SOLR backend
services that could fall over, and it can scale horizontally to as
many nodes are needed, as the git repository can just be checked out.

(In fact the Travis CI tests that verifies that the w3id.org URLs
still works does exactly this on localhost - see
https://github.com/perma-id/w3id.org#link-checking )


Also w3id.org is run by a consortium rather than a single actor, which
helps spread the risk long-term:

> There are a growing group of organizations that have pledged responsibility to ensure the operation of this website. These organizations are: Digital Bazaar, 3 Round Stones, OpenLink Software, Applied Testing and Technology, Openspring, and Bosatsu Consulting. They are responsible for all administrative tasks associated with operating the service. The social contract between these organizations gives each of them full access to all information required to maintain and operate the website. The agreement is setup such that a number of these companies could fail, lose interest, or become unavailable for long periods of time without negatively affecting the operation of the site.


And worst come to worst, anyone can git clone the w3id repository at
their own Apache server and change their DNS and SSL certificates to
have their local instance. The equivalent for purl.org seems
currently to be a bit tricky, but perhaps OCLC could provide data
dumps?



Recently we have been seeing more requests for w3id.org PURLs on the
public-perma-id mailing list
https://lists.w3.org/Archives/Public/public-perma-id/ - so I agree not
requiring .htaccess editing skills is good.

I've also seen that the dialogue that comes out of the email system
can help with making better permanent identifiers, particularly for
things like ontologies.


However I agree on a simple UI for making a request, perhaps some
form with Javascript would work? It might be harder to do the
fork/pull-request thing, but the page can simply say "Email this to
the mailing list", this still allowing for discussion that might be
needed for users who are not quite sure what redirect they need.