[open-bibliography] Library support and REST
pm286 at cam.ac.uk
Tue Oct 26 17:12:45 BST 2010
On Tue, Oct 26, 2010 at 2:26 PM, Karen Coyle <kcoyle at kcoyle.net> wrote:
> Quoting Peter Murray-Rust <pm286 at cam.ac.uk>:
>> The only reason I mentioned OL was that we in OKF cannot take data from OL
>> and pronounce it OKD-Open.
> What this means is that your definition of OKD-Open requires you to only
> take in data directly from the rights owner.
That's correct. The other critical thing is that the data gets stamped as
open. Then the Open licence travels down the chain.
> In a data cloud environment, I'm not sure that this is sustainable.
> Provenance is going to break down very quickly, especially when we deal in
> "statements" rather than "records". This is, of course, why there is a W3C
> group looking at provenance.
I am less hung up about the provenance explosion than some people. In
bioscience there is no explicit licence on the genome, structome, etc but
it's all used as if it were Open and everyone knows this. It works. The
community norm is that biodata on the data sites is effectively open. The
only thing that they worry about is DOS.
> I don't know how we solve this dilemma, but to me this puts OKD-Open in a
> "pre-cloud" position -- which is not a bad thing. Essentially a set of
> self-declared open data that people can select from. In the strict
> definition of that, very little of the data in OL comes from such a
> pre-cloud repository with rights attached, and therefore rights are mostly
> undeclared. It's more of a "free the data" approach, as opposed to "this
> data is free/open."
When we work with early adopter publishers - such as BMC - then the approach
is to label the data as Open. I agree that someone may take these pixels
from a frog image and overlay them on this gel and annotate them with that
graph. The original authors will not chase the community through the chain
and claim owenership of the frog's nose. If the data are used to make a
scientific claim and they are seminal to the paper then they will need to be
the risks of being sued for re-using a scientists work without cascading
attribution are minimal compared to the risk of being killed by
beta-naphthylamine, retroviruses, etc.
People use my data. If it matters to their argument they quote me. If it's
in a myriad of places they don't and I don't expect it. If I use an Open
bibliography I'll probbaly quote it generally - e.g. references were
obtained from Wikipedia... etc.
> Karen Coyle
> kcoyle at kcoyle.net http://kcoyle.net
> ph: 1-510-540-7596
> m: 1-510-435-8234
> skype: kcoylenet
> open-bibliography mailing list
> open-bibliography at lists.okfn.org
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the open-bibliography