On Jan 29, 2010, at 11:07 AM, Jo Walsh wrote:
dear all,
On 29/01/2010 16:18, Mr. Puneet Kishor wrote:
There is so much open and often linked data available on the web,
but
there has been a lack of an integration service tying together all
that linked
data into a more coherent experience.
This is great, but how is it different from http://ckan.net?
Having a service like this is extremely necessary, but having too
many
services like this, especially ones that overlap, might lead to
confusion. Clarity of purpose vis a vis existing efforts might be
very
helpful.
Well, that's how the semweb, or Web of Data, or Linked Data Web, is
likely to look in the short term. Lots and lots of aggregation
services collecting different subsets of the universe of triples. Re-
aggregating other sites, moving eyeballs around.
You can see a reflection of this in social network cross-posting and
aggregation services, for example FriendFeed. This was bought by
Facebook, I imagine because it was potentially or actually diverting
traffic away from FB.
An online service like FriendFeed collects your communication
streams from different sites together. A client like Posterous or
even Tweetdeck lets you send messages through to multiple sites. It
matters less and less where the information is made, where it is
stored, as long as it can be discovered.
Yes this leads to a lot of redundancy. We have got used to the
centralised, one-ring-to-rule model for types of online services,
redundancy can be confusing - one never knows exactly where to look,
and everything is slightly deja vu.
Years ago I worked on a prototype online service for a kind of self-
organising, distributed art festival in London. It pulled in RSS
listings from different event sites, correlated that with spatial
RDF from the OpenGuides wiki, FOAF maps of participants...
The event curators loved the *idea*. When they were presented with
the actual prototype, the main response "there are seven versions of
this event here. Which is the real one?"
Puneet, I think we just have to get used to overlap.
To use an over-used metaphor, I guess it will be Darwinian. I'd rather
have many services than none, but eventually, the stronger one or two
might survive. I am not pitching for CKAN over the other, and I am
certainly not sure if CKAN will be around 15 years from now. I hope
so, but no one can see the future.
From the point of the "user," I want to have one magic box in which I
enter my query, and then have that query be answered from wherever. Is
that magic box called "Google"? Is it my command line on my Terminal?
Is it a set of APIs that allow me to get to search and get to the data
programmatically?
From the point of the "contributor," I want to enter my data in one
place and have it be visible to all.
I am a big fan of CPAN (for those who might not be aware of it -- the
Comprehensive Perl Archive Network). It is a network of mirrors with
considerable redundancy so that no one node is canonical or critical.
CPAN is usually the model for all other C[A-Z]ANs, in the concept at
least, but not in architecture. It would take a considerable amount of
traction, history and commitment to have CKAN become like CPAN, but it
may be a worthwhile goal.
--
Puneet Kishor http://www.punkish.org
Carbon Model http://carbonmodel.org
Charter Member, Open Source Geospatial Foundation http://www.osgeo.org
Science Commons Fellow, http://sciencecommons.org/about/whoweare/kishor
Nelson Institute, UW-Madison http://www.nelson.wisc.edu
-----------------------------------------------------------------------
Assertions are politics; backing up assertions with evidence is science
=======================================================================
_______________________________________________
okfn-discuss mailing list
[email protected]
http://lists.okfn.org/mailman/listinfo/okfn-discuss
