Sherman Monroe wrote:
Hugh,
On Fri, May 15, 2009 at 5:18 PM, Hugh Glaser <[email protected]
<mailto:[email protected]>> wrote:
I have difficulty in interpreting what you say - we are looking at
the same "linked data browser" page?:
http://ec2.monrai.com:8890/facets/
All I can see is a page that says: "Lookup things named "" " So
the only possibility of linked data browsing is to put in
something with a name.
So I put something topical in (from this list).
Since I see
...
<Station rdf:about="http://ontologi.es/rail/stations/gb/MAN";>
<foaf:name xml:lang="en">Manchester Piccadilly</foaf:name>
...
I try "Manchester Piccadilly" and get some response, but nothing
leads me to
http://ontologi.es/rail/stations/gb/MAN.rdf or any of the data I
find there.
Similarly topical from this list, I find no way of browsing
http://foaf.qdos.com/lastfm/people/keithalexander#me or getting to
that page by searching for "Keith Alexander".
It appears you're right in this instance. However, here is some info
we have on Manchester Piccadilly (Stations)
<http://www.razorbase.com/?qid=10000101242426667642>. And here are
records we have for People named Keith Alexander
<http://www.razorbase.com/?qid=10000091242426317534>. We don't claim
to have visability over the entire Linked Data cloud, but to say we
are a browser of the Linked Data cloud is, I believe, as accurate as
to say my car is a car although it hasn't been every place a car can
go :) Google admits it may never be able to index a very large portion
of the web <http://en.wikipedia.org/wiki/Dark_internet>, but that
doesn't make them any less of a web search engine.
I've had conversations with Kingsley about this distinction, and based
on his insight re. warehousing all this data, I can identity at least
three LOD clouds (or at least three segments of it):
- warehouses (produced from published RDF dump archives)
- dynamic (sponged)
- crawled where you sponge/crawl progressively across the likes of
PTSW, Sindice etc.. that receive pings and crawl data
Even a 4th where people publish linked data, expose an sparql endpoint
but don't release archives. Better methods of advertising their SPARQL
endpoints and RDF servers once they come online would greatly help us
expand coverage.
-sherman
Sherman,
Hugh's key point is that you are not behaving like ODE, Zitgist,
Marbles, Tabulator, etc..
Also, this simply means: use the Sponger that comes with your Virtuoso
instance :-)
Kingsley
I also see no way of doing "3. Raw URI Lookup" of this URI.
Yes, this is a feature that is pending in the razorbase UI, but will
be added.
I assert that if this is a "linked data browser", then since
http://ontologi.es/rail/stations/gb/MAN.rdf conforms (I think) to
"How to Publish Linked Data on the Web", then I should be able to
browse it using the browser.
If I can't, then it is not a "linked data browser".
Hmm, Internet Explorer doesn't implement some portions of the
Javascript and CSS standards, but no one argues that IE is a W3C
compliant browser and CSS/JS client. But I do understand your point,
and it's something worth keeping in mind.
Of course, this browser is very capable of browsing your excellent
single site, but that is not the same thing, I am sorry to say.
Thanks so much for all your insight, feedback is always welcomed :)
--
Thanks,
-sherman
I pray that you may prosper in all things and be healthy, even as your
soul prospers
(3 John 1:2)
--
Regards,
Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen
President & CEO
OpenLink Software Web: http://www.openlinksw.com
