Print

Print


Hello,


as in the responses to my mail my motivation was interpreted quite
differently I just want to make clear what I would expect from Bibframe
and Library of Congress regarding data licensing.
 
Taking a look at licensing of library data now and in the past, I think
that both Jörg and Ed make a valid point. Generally, libraries in the
past haven't thought much about licensing their data nor was it common
practice to have contracts with every partner library and other
institutions you exchanded data with. (At least, I can tell that from
German practice and I know that contracts that (did) exist weren't
written by lawyers and thus are sometimes in large parts ineffective.)
The cooperative practice of data exchange between libraries was more
grounded on mutual agreement and goodwill but on written contracts or
licenses. This practice worked for quite a while until the web came
along and sharing of data more and more moved from a bilateral acts to
being a public act as it happens (or at least should happen) on the web.
That was the time when the uncertainty arose that Ed speaks about, and
now programmers and others who are interested in the data often don't
know under which conditions, if at all, they can re-use data.


We see an analogous development in other areas. For example, writers and
artists didn't have to think about licensing their content before there
was a way to easily make their stuff available to everyone with internet
access. In a print-world I don't have to think very much about
licensing, I sign the publishers contract and let him make my work
available. For more than ten years now, publishers on the web
increasingly have started explicitly stating what people can do with
their stuff. Creative Commons (and again Aaron Swartz who helped build
the technical basis for CC) certainly was and is an important enabler
for licensing things on the web as CC licenses make this act so easy. I
believe it is crucial that libraries and other cultural heritage
institutions provide a good example and make clear what people can do
with the stuff provided by libraries, be it digitized material,
bibliographic data or other things.


That said, my first request towards Bibframe was already articulated by
Ed:


1."[I]t is extremely important that Bibframe provide the mechanics for
explicitly stating the license associated with the bibliographic data
being expressed".


This is the minimum Bibframe should provide and it can be done on the
data model level. But I understand the Bibframe initiative not solely
being about a data model but about building a framework for the future
infrastructure of exchanging and sharing library data. A shared data
model is an important part here as methods for data exchange and
synchronization are which Bibframe will also work on. (BTW, I hope
Bibframe will not try to re-invent the wheel but re-use existing
solutions like these being developed by NISO and OAI in the ResourceSync
initiative [1].)  Additionally, I would say that in building a framework
for the future infrastructure of data exchange in and out the library
world, there is one thing at least as important as a data model and
exchange methods: that is _open_ licensing. Open licensing doesn't only
mean that license terms are made explicit when publishing things on the
web but using licensing terms that conform with the open definition.[2]


Thus, I have two other request towards Bibframe/LoC regarding open
licensing:


2. Library of Congress should - as one of the first and most important
implementers of Bibframe - provide an example of open licensing by
choosing CC0 or - maybe even better for most bibliographic data - a
Public Domain Mark [3]. (Although Loc data might already be in the
public domain it would be good for many potential re-users - especially
in an web-wide, international context - to have an explicit statement
about the data's public domain status. See John Wilbanks post on that
topic at [4].)


3. Furthermore, Bibframe should promote the other institutions. Endorsing the "Principles on Open Bibliographic
Data" [5] might be an easy start for doing this.


I hope this long mail makes the motivation behind my initial query
clearer. I am looking forward to hearing which parts of the licensing
aspect will be adressed by Bibframe.


All the best
Adrian


[1] http://www.niso.org/workrooms/resourcesync/ 


[2] http://opendefinition.org/okd/

[3] http://creativecommons.org/publicdomain/mark/1.0/


[4] http://del-fi.org/post/19787775081/us-government-and-cc0 


[5] http://openbiblio.net/principles/ 




>>> Ed Summers  02.02.13 11.07 Uhr >>>
This is an excellent (and timely) point, thanks for raising it Adrian. I
would just like to observe two things:

a) Linked Data *can* be used in closed contexts, where Web technology is
useful for data interchange within an enterprise, or within the context
of
a commercial agreement. I'm not saying it should, but just that Linked
Data
itself is neutral about the license. We use Linked Data principles
internally at LC for some back office applications, which nobody outside
LC
sees, or needs to see.

b) There is a big difference between encouraging the Library of Congress
(as a public institution) to share its bibliographic data using open
licenses, and having the Bibframe specification itself (which could
theoretically be used by a commercial entity like EBSCO, Elsevier, etc)
require that data be openly licensed. Personally (insert standard
disclaimer here) I think the former makes a whole lot of sense, but as
Kevin said, it is likely to be outside of the scope of Bibframe as a
data
format. This does beg the question of what venue there is to discuss
this
issue though. Perhaps there is a perception that this discussion list is
the place to discuss the licensing of content at the Library of
Congress?

All that being said, and returning to Adrian's point, I do think it is
extremely important that Bibframe provide the mechanics for explicitly
stating the license associated with the bibliographic data being
expressed.
I'm thinking of something like how the rel="license" Microformat works
in
HTML [1] or how the Creative Commons encourages licenses to be expressed
in
HTML [2]. This mechanism could be used to say exactly what "open" means,
and could also be used in more closed contexts.

Contrary to Jörg's point, I think that the MARC world we currently live
in
is marked by uncertainty: more often than not we don't really know if
the
data we have in hand is open or closed, or if parts of it are open, and
parts of it are closed. And I would argue we don't have agreement about
what "open" or "closed" actually mean. OCLC has recently been showing
the
way forward in their use of ODC-BY [3] to license data sets. I think it
is
extremely important that Bibframe data provide a mechanism for
explicitly
stating the license associated with the data, and that we strongly
encourage its use it as a community, even in situations where the
license
is not deemed "open".

Just so it's not forgotten I opened an issue ticket for it over on
Github
[4].

//Ed

[1] http://microformats.org/wiki/rel-license
[2] http://wiki.creativecommons.org/RDFa
[3] http://www.oclc.org/us/en/news/releases/2012/201248.htm
[4] https://github.com/lcnetdev/marc2bibframe/issues/6


On Sat, Feb 2, 2013 at 3:02 AM, Amanda Xu  wrote:

> > I would be very happy if software, deliverables, and access to
resources
> that are needed to build and operate systems that can process Bibframe
> >data are open sourced and liberally licensed. I personally see a new
> market for a new type of system vendors, with other competences and
> >different viewpoints, influenced by the open-ness of the Semantic Web
> community.
>
> The key issue is to enable library users to choose the appropriate
copy
> whether it is under restricted license or "open sourced and liberally
> licensed."  The appropriate copy is defined as the copy representing
not
> only the most authoritative, authenticated> lock for a researcher's task completion.
>
>
> *Amanda Xu    *
> *[log in to unmask] (email)*
> **
>
>
>   ------------------------------
> *From:* Olivier Speciel 
> *To:* [log in to unmask]
> *Sent:* Friday, February 1, 2013 11:14 PM
> *Subject:* Re: [BIBFRAME] Bibframe & Licensing
>
> Yes.
> Tks.
>
>
> Olivier Spéciel.
>
> Le 2013-02-01 à 12:53, "Jörg Prante"  a écrit :
>
> > Hello,
> >
> > I think what Adrian refers to is the point that many ILS vendors
tend to
> policies that lock bibliographic data away in silos. If libraries want
to
> export their data from the product, they often need to buy additional
> licenses. That is naturally an issue to be carried on with the vendors
how
> they create licenses for their customers, according to their business
plans.
> >
> > It's our common concern that Bibframe may be used just as a "product
> feature label" for continuing the same old policy to restrict
libraries to
> get their bibliographic data processed, for example, to build union
> catalogs.
> >
> > By adding open access APIs and smart linked data mechanims to
library
> systems, we hope to overcome the issue, so libraries can fully embrace
the
> promise the Semantic Web and Linked Open Data is making - free,
> unrestricted linked access to academic resources worldwide, anytime,
from
> anywhere, by any person. In memoriam Aaron Swartz.
> >
> > In a MARC world, we could have agreed that licensing bibliographic
data
> packages is an orthogonal topic. MARC records were closed entities,
> sequentially produced in files, and transmitted in packages over a
computer
> network, by a controlled workflow between known parties. But in a
> Bibframe'd world, the whole Bibframe concept simply won't work if
external
> references in Bibframe streams would link to catalog entities that are
> behind access-restricted systems, just because libraries are not rich
> enough to pay their vendors for all the extra licenses.
> >
> > So the quest is, how the building blocks around Bibframe are
provided -
> if entities and resources are officially labeled with the term
"Bibframe",
> who is allowed to re-use und re-distribute them without permission.
> >
> > I would be very happy if software, deliverables, and access to
resources
> that are needed to build and operate systems that can process Bibframe
data
> are open sourced and liberally licensed. I personally see a new market
for
> a new type of system vendors, with other competences and different
> viewpoints, influenced by the open-ness of the Semantic Web community.
> >
> > Best regards,
> >
> > Jörg
> >
> >
> > Am 01.02.13 17:37, schrieb Kevin Ford:
> >> Dear Adrian,
> >>
> >> IN re-reading your email, I may have conflated two distinct issues.
> One being the licensing around the BIBFRAME model and the other, which
I
> didn't address but which you mention, that has to do with the
licensing of
> bibliographic data.
> >>
> >> The latter, at this time, is beyond the scope of BIBFRAME, in just
the
> same way that licensing issues surrounding data described in MARC does
not
> bear directly on the usage licensing of the MARC format itself.
> >>
> >> Warmly,
> >> Kevin
> >>
> >> On 02/01/2013 10:31 AM, Kevin Ford wrote:
> >>> Dear Adrian,
> >>>
> >>> This actually came up once before on this list [1], the summation
of
> >>> which was, from Sally, BIBFRAME "will be [made available for use]
> >>> liberally like MARC."
> >>>
> >>> That said, we need to work on clearly articulating license issues.
 Now
> >>> that we've reached a few milestones, we have some time to look
into
> >>> these but the initial comment on this from November remains
accurate:
> it
> >>> will be made with little or no restrictions.
> >>>
> >>> Suffice it to say, uptake is important and liberal licensing will
> foster
> >>> that.
> >>>
> >>> Yours,
> >>>
> >>> Kevin
> >>>
> >>> [1]
> >>>
>
http://listserv.loc.gov/cgi-bin/wa?A2=ind1211&L=BIBFRAME&P=R5716&I=-3&X=482FBE44475D7F8C44
> >>>
> >>>
> >>> On 02/01/2013 04:27 AM, Adrian Pohl wrot> >>>> representing bibliographic data and moves to the discussion of
future
> >>>> data exchange and protocols, I would like to point to another key
> >>>> aspect of a future bibliographic framework. It's the licensing
aspect
> >>>> that already was discussed in 2007 after the publication of LoC's
> >>>> report "On the Future of Bibliographic Control" [1] that prepared
the
> >>>> ground for Bibframe. At this time, some people were "concerned
that
> >>>> the report lacks any discussion of a key component for any future
of
> >>>> bibliographic data: open licensing and access" and published a
> >>>> response to the Library of Congress [2] (co-drafted by the Open
> >>>> Knowledge Foundation and Aaron Swartz, see [3]). As far as I
know, LoC
> >>>> never reacted to this response. Since 2007, many libraries and
related
> >>>> institutions have embraced open licensing. Thus, integrating a
> >>>> licensing policy into the bibliographic framework shouldn't be a
big
> >>>> deal by now.
> >>>>
> >>>> I am curious whether it is planned to address the licensing
aspect in
> >>>> the development of Bibframe and when this will happen. I would be
> >>>> happy if Bibframe made clear how the web-wide free flow of
> >>>> bibliographic data will be LEGALLY ensured and won't be hampered
by
> >>>> intellectual property right and licensing conditions.
> >>>>
> >>>> All the best
> >>>> Adrian Pohl
> >>>>
> >>>>
> >>>> [1] http://www.loc.gov/bibliographic-future/
> >>>>
> >>>> [2] http://wiki.okfn.org/FutureOfBibliographicControl
> >>>>
> >>>> [3]
> >>>>
>
http://blog.okfn.org/2013/01/14/goodbye-aaron-swartz-and-long-live-your-legacy/
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> Adrian Pohl
> >>>> - Linked Open Data -
> >>>> hbz - Hochschulbibliothekszentrum des Landes NRW
> >>>> Tel: (+49)(0)221 - 400 75 235
> >>>> http://www.hbz-nrw.de
> >>>>
>
>
>