[open-bibliography] BibSoup/BibServer collaboration model?

[Jim Pitman]

Naomi Lillie <naomi.lillie at okfn.org> wrote:

> Following on from the below, please see here for a forum for comparisons:
> http://wiki.okfn.org/Projects/jiscopenbib2/managementtools

Following are some reactions, some of which could  be incorporated in this wiki, but others
seem better discussed first on the list as they concern the structure and purpose of the wiki and related systematic
efforts by this WG rather than the proposed wiki content.

There should be a link to http://en.wikipedia.org/wiki/Comparison_of_reference_management_software

I think it should be emphasised more that BibServer/BibSoup is not competing in the
domain of reference management. Rather, BibServer/BibSoup facilitates publication, display and sharing
of bibliographic data collected by whatever means. It seems critical to me for early success in expanding
interest and support for BibJSON/BibServer/BibSoup that we emphasise and facilitate the interoperability aspect:
that BibJSON is a rich enough format that it can embed the structure of most data in other standard bibliographic formats,
and that it should be possible to create and maintain an open source community of coders with a shared interest in frictionless
sharing of bibliographic data maintained with whatever management tools.  We should aim to facilitate import and export of biblio
data to/from any of the systems listed in the wikipedia page, and develop a community of coders with an interest in supporting
this activity.  These ref man systems naturally compete with each other for users and developers, largely for commercial reasons.
We are offering a different product in this space, which aims to open up the data presently managed in these systems, and
to make it much more freely available and reusable. I think we should use pages like the wiki to to convey some of the excitement of 
this effort, to generate interest in the commuity, and get coders to spend their time writing converters and apps based on BibJSON/BibSoup 
rather than submitting entries to offensive competitions run by proprietary bib management systems which end up owning the code and exploiting
it for their own commercial purposes.

About "please keep bias to a minimum, we want facts not opinions!".
I dont think NPoV is the way to go.  We already have the Wikipedia page for the facts. I see little point in replicating the wikipedia table
unless we can make it more accessible or better organized or machine-readable. But just copying from one wiki to another wont achieve any of those things.

I would rather see a table of progress of integration between BibJSON/BibServer/BibSoup and each of the systems listed in the wikipedia table.
That seems to me what is special which this community can provide: guidance with code and exemplar datasets for how to get data in and out of these management
systems and to exchange it into the public domain with tools and services this community is supporting.
We would prefer the big biblio management systems to cooperate with us by providing BibJSON exports with an open licence, rather than for us to have to  maintain a large number of converters
and risk legal issues if we are too aggressive in claiming data for the public domain. I see the same issue in dealing with all of
Mendeley, Microsoft, Google Scholar, Zotero, BibSonomy, .... The risks and rewards vary from provider to provider.

How to kick start that community development process, to get some momentum going in supporting interoperability between BibJSON/BibServer/BibSoup and the dominant
reference management systems? I dont think we have answered that question yet, or adequately defined the community structure or political process by which 
it can be achieved. The political process could be some mix of grass-roots activism, connector code sprints, data liberation efforts, and high level contact between OKF appointed 
representatives and representatives of the dominant ref management systems. Individuals like Peter MR and myself have contacts with reps of some of these systems. But we do not
currently have the political authority to make representations or agreements on behalf of OKF or its WG. 
Some template for interaction like http://www.isitopendata.org/ would also seem desirable. But if the data owners claim it is not open, we need strategies to make it so.

I'd also be glad to see more of Peter's rants against pseudo-openness and Mendeley copyfraud under the OKF flag.
We had quite a spirited discussion on the list of issues around appropriation of Bib data from the public domain. 
Why not air some more of that out in the blogosphere?  This is a strategic issue for the openbiblio community.
I am not clear how  OKF can help resolve such issues through the WG structure. Perhaps it is best just for Peter to blast
away from his blog and for OKF WG to take a more moderate position. I am not sure.  

In any case, I hope this post may stimulate further discussion of how best this WG could assist on the community side in supporting
the great efforts Mark has made in developing openbiblio services.

--Jim

----------------------------------------------
Jim Pitman
Professor of Statistics and Mathematics
University of California
367 Evans Hall # 3860
Berkeley, CA 94720-3860

ph: 510-642-9970  fax: 510-642-7892
e-mail: pitman at stat.berkeley.edu
URL: http://www.stat.berkeley.edu/users/pitman