[Open-access] NISO Recommendations on Open Access Metadata

Tue Jan 7 10:29:17 UTC 2014

The thing that could save this would be a registry of licence URLs, so
that a CC By document ALWAYS uses (say)
        http://creativecommons.org/licenses/by/
(even though that resolves to a "not found page") rather than
        https://creativecommons.org/licenses/by/
        http://creativecommons.org/licenses/by/1.0/
        http://creativecommons.org/licenses/by/3.0/us/
        http://creativecommons.org/licenses/by/4.0/
        http://creativecommons.org/licenses/by/4.0/legalcode
Or any of the other many other credible candidate URLs

Software needs to be able to do something very simple:
        if (!strcmp(licenceURL, "http://creativecommons.org/licenses/by/") ...
otherwise people are surely going to implement swathes of error-prone
and mutually incompatible heuristics.

-- Mike.


On 7 January 2014 10:22, Cameron Neylon <cn at cameronneylon.net> wrote:
>
>>> The first was that any such element could easily be inconsistent with the actual license. So if the license_ref link is to CC BY but someone encodes non-derivative in metadata which do you believe? And inconsistency is already a problem as Daniel repeatedly tweaks us about (with good cause!). So avoiding the risk of adding to that is important.
>>
>> I acknowledge this point. However ...
>>
>> Without this, the specification is essentially useless. It's not
>> machine-readable metadata in any but the thinnest, most legalistic
>> sense. The specification boils down to "here's a URL to a licence
>> text, get your lawyer to download it and look it over". I can't do
>> anything with that.
>
> No I disagree. I accept its less useful for non-standard and proprietary licenses but having gone over this I don't think it is either legally, or possibly even theoretically, possible to build a general machine readable standard for rights - and the core problem is that those non-standard licences are just less useful anyway, because they're non-standard. In practice the way to do this, as you say is for a single organisation to create a widely used standard that everyone understands. CC have done that - the URL to a CC license is as good as machine readable. And to the extent it isn't then there's CC-REL.
>
>> Yes. A huge job that will either never happen, or be ready to use in
>> ten years, by which point this specification will have been left in
>> the dust by whatever pragmatic solutions wins out when the standard
>> turns out not to meet the actual need.
>
> And that specification will be CC licenses. For which the recommendation will support better standardised transmission and the adoption by Crossref means we've got a centralised place for collecting those license statements. The place to standardise is the licenses themselves, not the transmission mechanism. I'm more or less convinced that nothing else can work - but am happy to hear arguments to the contrary.
>
>> Really? "Not a step back" is the level of our aspirations now?
>
> When the risk was allowing certain parties to say "But NISO says that 'free-to-read' means Open Access so how can you argue with that?". Yep, and avoiding that alone was worth the six months that I sunk into it. That we got agreement from a diverse set of stakeholders that they should a) put their licenses at a defined place on the web and b) provide the URL to those terms with core metadata was a bonus. That we've separated metadata elements on reading rights from re-use rights is also good.
>
> You have to defend the flanks as well as lead the advance.
>
> Cheers
>
> Cameron
>
>
> _______________________________________________
> open-access mailing list
> open-access at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/open-access
> Unsubscribe: https://lists.okfn.org/mailman/options/open-access