[okfn-discuss] open shakespeare: update

[Rufus Pollock]

Jo Walsh wrote:
> On Sun, Jun 25, 2006 at 07:19:48PM +0100, Rufus Pollock wrote:
> 
>>Jo Walsh wrote:
>>
>>>very nice! questions:
>>>- http://gutenkarte.org - anywhere near some of your dreams? sadly
>>> there are 'issues' in re the source + metacarta releasing it; but
>>> the parser/geotagger API is open, as is the client code. 
>>
>>Looks very cool. This would tie in really well with the microfacts idea 
>>we've talked about. Of course an open API does *not* = open knowledge 
>>(as you point out when mentioning the source issues).
> 
> 
> Gutenkarte/GeoParser is an odd case for this. What's behind the API is
> a combination of data and process, so this is not just an 'Open API'
> for something that, given the original public domain body of
> knowledge, it would be possible to reconstruct without an effort that
> went further than 'arrangement'. 

Not quite sure what you mean here -- do you mean that they have done 
work to organize and tag the data (i assume the geoparser is a 
combination of geoname thesaurus + natural language parser)

> GeoNames for example offer a similar geotagging service to MC; they
> distribute the data from which it is built under a CC-BY license at
> http://www.geonames.org/export/ . Some of the webservices are easily

Another item to add to ckan :)

> 'extrapolatable', others are not. Commercial services reuse the data,
> but via a webservice interface, geonames doesn't get attribution. And
> arguably they are redistributing a lot of data that comes from public
> domain (US gov) sources; what exactly is it they're "re-licensing"?

In Europe they could be relicensing under the database right: this is a 
right specifically design for the collection/organization of 
data/information which you have not produced. In the US despite Feist 
common-law copyright still protects many collections of information 
(advertising listings etc)

> On a completely different tack. Recently I've been goaded into reading 
> a lot of puritan sermons by a friend who i was *startled* to discover
> subscribes to a literalist, calvinist-creationist flavour of
> christianity. I have no background in this stuff whatsoever, no
> context for a framework of understanding even how a mind can hold
> these beliefs; so i am slowly sloughing through the King James too. 
> 
> There is this incredible tradition of exegesis, cross-reference and so on. 
> There are some rather neat search-and-reference tools online, but none 
> of them i have found have machine-readable API interfaces - stuff like
> http://www.blueletterbible.org/freeoffer.html that none-the-less does
> some good phrase matching and something like the 'beads on a string' /
> 'atoms of facts' approach which you talk about, Rufus. 

This is exactly what we should try to do with open shakespeare (once 
we've done shakespeare it shouldn't be too hard to repeat for other texts).

> I want automation for the context of the sermons - something that
> should be straightforward to do in a lot of ways, as references come
> with chapter and verse and are very often in quotes. Once I've got a
> reference of some degree of confidence via an 'Open API' of whatever
> degree of openness (say a reparser of a search interface like the
> above, that spews out RSS/RDF, or a service built on that which can
> take a body of text, cross-reference and spit it back out with links
> and in wiki markup format); I republish the annotated data - is it now 
> open and free? 

In my view you always need to make source available. A web-api is always 
a service sitting on top of some data/code etc. An open service (i.e. 
api) does not equal open data or open code.

With regard to the openness of the republished data: yes it is open if 
you licence it is as such (and yes you probably do get a new right be it 
copyright/database right for annotating and structuring the original 
information).

> Open exegesis... WikiTalmud? Perhaps I'm being silly, but let me
> become a fool that i may be wise, and so on. 

I don't think this is necessarily silly at all. Exegesis/commentary on 
texts is perfectly suited to an open approach and for it to work well in 
a decentralized collaborative manner requires a good 'api' to the 
underlying text.

I've been collecting links for annotation tools here:

   http://www.okfn.org/wiki/SoftwareTools

and using del.icio.us:

   http://del.icio.us/tag/openknowledge+tools+annotation/

> Another thing that strikes me is how the core of this is a persistent,
> predictable URI scheme. Sigh, years ago when I first ran into the
> INSPIRE directive and started banging my head against the EU body of
> legislation i registered ecdirectives.info to do this, provide predictable 
> URIs for Directives in such a way as they'd be semi-automatably
> cross-referenceable - mostly to help myself unpick the 'This paragraph
> is without prejudice to Section 2.2 of 2003/05/23' reference nest. 
> As often with my software I never got to the point of making it usable
> by others. It had CGI::Wiki as a backend, and that's obsolete now.
> ISTR Gavin Bell et al had a wiki-annotatable concordant version of the
> proposed European Constitution going, but the Constitution was pretty
> obviously tanking by the time they actually launched it... that was
> mediawiki based. 

This is another good idea for an open knowledge project. The obvious 
people to talk to would be the publicwhip/theyworkforyou guys since I 
know they are working on an open db of UK law (persistent uris and 
everything I would imagine). Francis would you care to comment?

While you are at it why not post a proposal for a project in the 
incubator section of the wiki:

http://www.okfn.org/wiki/#head-47fd6e3d8441963f1270d74a880a7d96ba1ed46f

You could even post a link to your old (broken) code :)

Regards,

Rufus