[open-linguistics] Inclusion of 'non-open' resources in the LLOD cloud diagram

Tue Sep 8 10:44:09 UTC 2015

On 7 September 2015 at 15:47, Víctor Rodríguez Doncel <vrodriguez at fi.upm.es>
wrote:

> Dear John,
>
> Thanks for posting this relevant question.
>
> As a collection I would love to see it as inclusive as possible.
> Please do remind that the O in LLOD can also be interpreted as "Open
> Standards" rather than "Open Data".
>

The problem with that is that the 'Open' is clearly in reference to the
Linking Open Data project, which explicitly exists to link Open Data (
http://www.w3.org/wiki/SweoIG/TaskForces/CommunityProjects/LinkingOpenData#Project_Description).
LOD (and Open Data, Open Culture, Open Source) has sufficient mindshare
that attempting to replace the known value of 'Open' with another would
seem deliberately deceptive.

Why not just drop 'Open'? Does keeping it serve some other purpose that
outweighs its potential to mislead?

> Which is the rationale behind excluding valuable Linked Data resources?
>

>
Also, having non-free resources empowers the value of free-resources ---the
> more resources in the cloud the higher the value of each of them
> separatedly. A partially commercial LLOD matches the web in general --which
> is also partially commercial.
>

'Commercial' and 'not Open' do not mean the same thing, though. Much of the
(genuinely) Open Data _is_ commercial: commercially produced, commercially
consumed... far beyond a single commercial activity of making proprietary
data available for sale.

Commercial _consumers_ are the primary beneficiaries of LOD, as it greatly
simplifies matters for them. Even if the question were only of licensing,
the prospect of having to check each resource to obtain a licence is often
enough (or, rather, *was*, pre-LOD) to dissuade many potential commercial
consumers from using *any* linked data. But it's not simply a question of
licensing: much linguistic data is collected on the strict condition that
it never be commercialised, so there isn't even the possibility of
obtaining a licence.

It can also be against the interests of commercial producers to have their
data included as nominally open data, as in jurisdictions with strong
advertising requirements they could run the risk of finding themselves
unable to demand the purchase of licences.

Also, I don't think it's helpful to group together CC-BY-SA and CC-BY-NC.
CC-BY-SA has restrictions, sure, but only if you are preparing derivative
works or redistributing - not when using (neither activity is considered
"use" under copyright law in Berne signatory countries) - whereas CC-BY-NC
forbids commercial use[1].

[1] The text of the licence restricts this only to activities that are
generally considered the domain of copyright law. IANAL, but I don't
imagine it would take a particularly good lawyer to successfully make the
case that making CC-BY-NC data available for query in a publicly accessible
database counts as either 'public display' or 'public performance'.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-linguistics/attachments/20150908/16429c65/attachment-0003.html>