# Twine and Linked Data on the Semantic Web: A Quick Comment
Just read “Twine and Linked Data on the Semantic Web“, a blog post by Nova Spivack.
He says that Twine has a SPARQL endpoint, it’s just not open yet. He also says that it is able to consume Linked Data.
This is all good, however, even though it’s stated in the Twine Semantic Tour, application/rdf+xml fetching (using cURL) doesn’t work from command line:
curl -H "Accept: application/rdf+xml" http://www.twine.com/item/1sj51123-2g5
<html>
...
Now this comes up because of Unauthorised access, which is a real pain!
If you have logged in to Twine Beta then you can access the RDF from browser by going to: http://www.twine.com/item/1sj51123-2g5?rdf=
But I don’t really want to do that…. so my suggestion:
Make it entirely Linked Data compliant:
1. Use URIs as names for things
2. Use HTTP URIs so that people can look up those names.
3. When someone looks up a URI, provide useful information.
4. Include links to other URIs. so that they can discover more things.
(Quote from “Linked Data - Design Issues” by Tim Berners-Lee)
With Twine, if it were downloadable the database backend could then be powered by ODBC… allowing it to connect to any database management system including: MySQL, PostgreSQL or even the mighty OpenLink Virtuoso Universal Server.
More information is available about the awesome open-source platform-independent implementation of ODBC and X/Open specifications called iODBC.
Nova, remember: be open, use existing ontologies, link out, link in, take advantage of HTTP, and finally be meaningful!
This is the way to gain full benefit of the Linked Data web, and gain respect from users and developers of the Linked Data web. Interconnectedness is key.
(p.s. Nova… have you got a personal URI, if so, does it output FOAF based RDF?)
[update]
I forgot to say, the profile pages on Twine don’t even give you FOAF based RDF… which is really annoying.
- My profile on Twine (doesn’t return FOAF)
- My profile on Myopenlink (does return FOAF)
[/update]
[another update]
Tom Heath gives a bit more detail about some of the bad bits of Twine (in regards to the Linked Data issues) in his blog post titled “Twine and Linked Data“.
[/another update]
March 31st, 2008 at 11:44 pm
Hi Daniel. Thanks for your feedback; it echos that of many others as well. We do in fact use a combination of both our own and existing ontologies. Our platform is completely ontology agnostic, save for RDF, RDFS, OWL, and a couple of our own that are needed for bootstrapping. Currently all resources in twine are dereferencable with known, HTTP-based URIs that point to their corresponding web page. All of the fundamentals of linked data are there, even if not yet fully accessible in every way. It comes down to a question of time and resources: we have an enormous backlog that includes not only data interoperability tasks, but many more basic app usability issues as well - all of which have to be prioritized alongside each other.
One thing should not be in question: we are *very* concerned with being an open platform, not a silo. But it takes time to do things. We’ll get there. Until then, we very much appreciate the thoughtful feedback. We’re committed to listening to what the community wants.
Jim