[open-bibliography] German National Bibliography as LOD under CC0

[Peter Murray-Rust]

On Thu, Jan 26, 2012 at 6:04 PM, Jim Pitman <pitman at stat.berkeley.edu>wrote:

>
> Peter Murray-Rust <pm286 at cam.ac.uk> wrote:
>
> > Bibserver can, in principle, allow annotations. All bibliographies offer
> > great scope for this.
>
> True. More simply, BibServer should allow any Biblio maintainer to
> support  selection of sub collections by interested users. This should be
> the
> first step I think. Maintenance of the user base and permissions will be
> the
> main burden on the BibServer maintainer. It is not clear to me whether
> national bib maintainers would be able to support this burden. I rather
> doubt it, unless there is a technical FTE designated to deal with such
> things.
>

I don't think this is necessary. Let's say that OKF (say) clone the British
National Bibliography today and put it on a Bibserver. It's
read-everywhere/everybody. It's (let's say) writable by registered people.
These people migh add everything from location of real books, to copyright
info (Public Domain Calculator), to pointers to Wikipedia. This is an
OKF-bibserver, not the BL's, and so long as that's clear then it doesn't
harm the BL and may well enhance it.

Clearly we have a versioning task here, because the BL will continue to
output releases. We don't have to feed the info back to the BL. But there
might be other sites which value our annotations (e.g. copyright
calculations) and they clone a subset from OKF. Similarly we might swap
information with Wikpedia.

We start to get an ecosystem of bibliography. The key thing is that we can
navigate it. IDs and versions are critical. But they should be manageable.
So every so often consenting partners in this might do merge and update (cf
Mercurial or Git). Each node in the network has its own community rules.
(Wikipedia allows anyone to play but erases vandalism - OKF might open
write-permission to anyone in a given project, etc.)

I think that smaller, subject-specific biblios, like for malaria, or
> probability, would be more easily capable of managing users, due to the
> much smaller scale. But then these installations have to be able to handle
> ingest of data from  multiple sources including e.g. national biblios.
>

It all depends on what people want. The fundamental truths are:
* no bibliography is perfect (or ever can be)
* no bibliography is up-to-date
* no bibliography is comprehensive (or can be)
* different people want different things.

Let's say I want a bibliography of PGWodehouse (I want to make a giant
graph of all relations of all characters). This is a final task. I clone
the PGW subsection of the bibliography and that's all I need. I annotate
each entry with Gussy Fink-Nottle's appearances, and Offy Prosser and the
Empress of Blandings. The BL aren't interested or worried by this. But
maybe a PGW-society might get very interested and clone and distribute this
resource among its members.

Malaria will need updating. As often as we can manage, but not more. That's
harder. But it's still a unidirectional process. Deletions and updates form
a problem. Some will require manual intervention at first.

But it's far better than not having anything at all

P.




-- 
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-bibliography/attachments/20120126/43260667/attachment-0001.html>