[open-bibliography] Comments on transformed BNB data

[William Waites]

* [2010-11-24 14:26:23 +0000] Ben O'Steen <bosteen at gmail.com> écrit:

] On Wed, 2010-11-24 at 13:26 +0000, Deliot, Corine wrote:
] > I'm sorry but I still don't understand. I've looked at
] > http://bnb.bibliographica.org/entry/GB8102507, which I believe is the
] > RDF/XML you were pointing me to. 

That one doesn't appear to have a dces:date. This one does:

http://bnb.bibliographica.org/entry/GB5011841.rdf

Most of them also have dc:issued which is converted to a proper date
(typed literal) where possible.  

] Will, can you shed any light on why the two namespaces have been
] combined?

They haven't. We just call the legacy namespace dces and the current
ones dc. Please don't get hung up on the prefixes used, they're
entirely arbitrary. (I didn't realise it could cause this much
consternation, I'll put it on the TODO list to tweak the prefixes used
for serialisation).

] Also, in examining the example I linked to above, I noticed it created
] an End of the interval of "9999", even though the original shows an
] undefined end point, in this case defined by "[2000?]-". I think this is
] a situation where omission is far better than the use of a value like
] 9999. The interval makes perfect sense upon removal of the hasEnd node
] IMO.

I certainly agree. This is in the data we received, right? I'm pretty
sure I haven't seen anything we do that invents arbitrary dates.

Cheers,
-w
-- 
William Waites
http://eris.okfn.org/ww/foaf#i
9C7E F636 52F6 1004 E40A  E565 98E3 BBF3 8320 7664