[okfn-discuss] 4 Ideas for Defending the Open Data Commons by Simon Chignard

[Michael Bauer]

Peter,

> This is critical. I've had this myself - people want to dump my data and I
> want them to dump it but it can be technically hard work.I have a million
> web pages for http://wwmm.ch.cam.ac.uk/crystaleye.
> * they are listed as Open Data.
> * anyone can scrape the web pages. However there is no trivial iterator and
> no list of pages
> * we (jim)  built an RSS tool that downloaded chunk-by-chunk to make it
> easier to  get better packaged data. But not everyonce can use RSS easily.

If the data is the webpages - a simple index will do. With something like
proper structuring (or even RDFa or microformats) the webpages simply
become the API. This depends on your collecting mechanisms and can be built
cheaply when planning the project. 

> If we put very high constraints on ease of download then people will be
> unable to afford to make their data Open.
> 
> This is particularly true for data which is (a) dynamic (b) distributed.
> The bioscience data is both of these. It's effectively open as anyone can
> get chunks ate any time, but probably *couldn't* even be completely
> compliant.

I do see the point here. (And also Rufus' point). One thing that I do think
needs to be given is the bulk option (so one can build an API oneself -
this API can even offer cheaper rates ;) 

In my opinion these considerations (and the availability as bulk) has not
been discussed in the article. I just stumbled accross this from a user
perspective (and thought - wait a minute, I wouldn't use an API with these
terms). IMO this point would need more discussion in the article and
charging fro API access is presented as a great idea (which it mostly isn't
IMO).

Michael
-- 
Data Wrangler with the Open Knowledge Foundation (OKFN.org)
GPG/PGP key: http://tentacleriot.eu/mihi.asc
Twitter: @mihi_tr Skype: mihi_tr