Assigning ARKs for resources in dark archives

[Mark Jordan]

Hi everyone,

Is anyone assigning ARKs for digital content that is not intended to be accessible on the web? My specific use case is that I want to provide locally-managed ARKs for archival information packages stored in a dark archive, with no direct access via conventional http. These resources would likely be stored on heterogeneous file servers (e.g. with Windows share paths, Linux filesystem paths, etc. as the ARK target). Targets would be updated as files get moved to different locations resulting from storage migrations, etc. The ARKs for these resources would not be published, but would be for internal (to my organization) use only.

Has anyone done this?

Thank you,

Mark

[Donny Winston]

I have not done that specifically, but I am assigning with the intent for web-accessibility to vary over time for a given ARK. For example, a data distribution may be “dark”, it may be “embargoed” before being made accessible after say 2 years, or it may simply be on cold storage and thus a request for its target will e.g. get a Retry-After header in the response.

As for not even publishing the ARKs themselves (like returning a 404 to external requesters), I have not done that.

--

You received this message because you are subscribed to the Google Groups ARKs group. To post to this group, send email to arks-...@googlegroups.com. To unsubscribe from this group, send email to arks-forum+...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/arks-forum?hl=en

---

You received this message because you are subscribed to the Google Groups "ARKs" group.

To unsubscribe from this group and stop receiving emails from it, send an email to arks-forum+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/arks-forum/db8884ce15204230a5f0f0f95c48aa65%40sfu.ca.

--

Donny Winston, PhD (he/him/his) | Polyneme LLC

https://donnywinston.com | https://polyneme.xyz

If I've emailed you, I'd love to speak with you.

Schedule a meeting (15min+): https://meet.polyneme.xyz

[Tallman, Nathan]

Hi Mark,

 

At Penn State, we are planning to mint ARKs for dark preservation objects that will not be accessible to the general public. We will save the preservation ARK in object metadata; similarly the preservation repository metadata will include the access ARK.  We may still include a target to resolve them to a preservation repository where only authenticated users will be able to access them, but it’s not really necessary as the ARK itself will be maintained as an identifier for the object and can easily be searched. The ARK spec definitely supports non-web use cases and they can even be used for physical objects.

 

I believe the University of Houston does something similar by minting ARKs for dark preservation objects but only including the identifier (not a resolvable ARK) in their public metadata.

 

Thanks,

Nathan

 

-- 

Nathan Tallman (he/him)

Schedule a Meeting

Chat with me on Teams

--

[Phillips, Mark]

Hey there Mark, 

At the University of North Texas we also mint ARKs for digital items that are going into our digital preservation repository that generally will never be used outside of that system (a system we call Coda).  We have a mapping from the public access ARK identifiers to these internal ARK identifiers. 

We do not publicly list the Coda ARKs in our public interface and generally they are looked up when the access ark string is queried in the preservation repository which results in one or more preservation identifiers associated with that access identifier. 

Mark

---

From: arks-...@googlegroups.com <arks-...@googlegroups.com> on behalf of Mark Jordan <mjo...@sfu.ca>
Sent: Sunday, February 20, 2022 12:01 PM
To: arks-...@googlegroups.com <arks-...@googlegroups.com>
Subject: [EXT] [arks] Assigning ARKs for resources in dark archives

 

[Mark Jordan]

Hi Mark, Nathan, and Donny,

Thanks very much for your responses, they are very useful.

One additional question: if the resource an ARK identifies does not have any public-facing representation on the web (which would be the case with my dark-archival packages), what would the ARK resolve to in the unlikely case one made its way to the open web? In other words, in the absence of a redirection target, is there an expected/standard response (e.g. a 403 Forbidden)? I assume that if metadata was requested with '?', that metadata would be returned, but is there an expected behavior in this case if metadata was not requested? 

Mark

---

From: arks-...@googlegroups.com <arks-...@googlegroups.com> on behalf of Phillips, Mark <Mark.P...@unt.edu>
Sent: Monday, February 21, 2022 10:59 AM
To: arks-...@googlegroups.com
Subject: Re: [EXT] [arks] Assigning ARKs for resources in dark archives

 

To view this discussion on the web visit https://groups.google.com/d/msgid/arks-forum/SA1PR01MB7342DE8FD2E87D37FBAF696AEF3A9%40SA1PR01MB7342.prod.exchangelabs.com.

[Donny Winston]

I would expect a 403 response in that case.

I imagine the ARK spec may not take a position on that, as it's specific to the HTTP protocol, and the spec seems to like to emphasize that ARK resolution should not be tightly coupled to HTTP.

To view this discussion on the web visit https://groups.google.com/d/msgid/arks-forum/6d488d45e4c44290918888010b67ea18%40sfu.ca.

[John Kunze]

One additional question: if the resource an ARK identifies does not have any public-facing representation on the web (which would be the case with my dark-archival packages), what would the ARK resolve to in the unlikely case one made its way to the open web? In other words, in the absence of a redirection target, is there an expected/standard response (e.g. a 403 Forbidden)? I assume that if metadata was requested with '?', that metadata would be returned, but is there an expected behavior in this case if metadata was not requested? 

I would refer to the "Generic ARK Access Service" section of the specification, excerpted below from the current (v34) spec. There's latitude to return "a sensible object proxy" when returning the object isn't possible (for whatever reason it isn't possible). 

There is nothing to prevent you from returning a statement that, for example, the object exists but isn't generally available because <and give your reasons>, and here's a brief description: <your formatted metadata>. Of course, you're not required to return a description for the given ARK, depending on your obligations to end users and to object owners. 

----------------------------

5.1.  Generic ARK Access Service (access, location)

   Returns (a copy of) the object or a redirect to the same, although a
   sensible object proxy may be substituted.  Examples of sensible
   substitutes include,

   *  a table of contents instead of a large complex document,

   *  a home page instead of an entire web site hierarchy,

   *  a rights clearance challenge before accessing protected data,

   *  directions for access to an offline object (e.g., a book),

   *  a description of an intangible object (a disease, an event), or

   *  an applet acting as "player" for a large multimedia object.

   May also return a discriminated list of alternate object locators.
   If access is denied, returns an explanation of the object's current
   (perhaps permanent) inaccessibility.

To view this discussion on the web visit https://groups.google.com/d/msgid/arks-forum/d6f2e834-09a6-426d-aee7-7a7900d6a9c3%40www.fastmail.com.

[Mark Jordan]

Thanks John - small world, just as your message hit my inbox I was reading v29 of the spec (linked from https://arks.org/specs/). I was unaware there was a v34 already.

Mark

---

From: arks-...@googlegroups.com <arks-...@googlegroups.com> on behalf of John Kunze <j...@ucop.edu>
Sent: Tuesday, February 22, 2022 3:44 PM
To: ARKs

To view this discussion on the web visit https://groups.google.com/d/msgid/arks-forum/CAOmcTKjQ1uNLJrLTxJqv%2BXmCZKVDsCT9ytpxZeP6Go1cuhnfWg%40mail.gmail.com.