Offer guidance on an analog to external datastreams

[barmintor]

Using Content-type: message/external-body; access-type=URL; for referenced data requires the client to have unmediated access to anything described at the URL. This is generally not the case for content historically described as 'external', for which a repository acts as a proxy.

[awoods]

Agreed, the external-body approach here is more like the historical 'redirect' datastream type.

[ajs6f]

+1 to Fedora <4 "redirect" being the right comparison.

[ruebot]

@barmintor trying to understand on what needs to be done to resolve this issue, do we need to update the language in:

• 1.3 http://fcrepo.github.io/fcrepo-specification/#httpPOST
• 1.4.1 http://fcrepo.github.io/fcrepo-specification/#httpPUTLDPNR

...or is there more to it?

[awoods]

@barmintor ^^

[barmintor]

It's not clear to me what analog we can or should offer in the spec. I created this ticket in response to a call on which it was suggested that external-body was a way to migrate external datastreams, which it is not.

[barmintor]

I think I also mentioned that it was generally an impediment to migration at Columbia, but that's not a spec issue.

[awoods]

It does not appear that there is anything to be done towards resolution of this issue. @barmintor, if you agree, would you be willing to close it?

[kefo]

I think this can be closed. The issue here - clarification that the message/external-body resource is a "redirect" - will be resolved once #40 is resolved. It's a duplicate.

[awoods]

@barmintor : I believe this issue has been resolved with: #121
If you agree, would you please close this issue?