Hi Dan, I was being sarcastic in saying "will always be up-to-date". In fact I think it is highly unlikely that metadata will be kept up-to-date because of the investment required. However, if it is not up-to-date, how valuable is it? In HealthInsite we link to a lot of consumer health fact sheets on different partner sites. These are often reviewed and updated. We have to regularly remind our partners to put clear dates on their fact sheets and to make sure these match their metadata dates - the selection of the date is a manual process; it cannot be automated because a computer will not know if a change to a page is something trivial or a genuine update. With any manual process, there is a risk of error or omission. We have made a considerable investment in developing a metadata workflow that reduces errors. Other organisations around the world might not be willing to make this level of investment. I agree with you that there were problems with extending the DC dot model - your example is very messy. However we do not need these sort of extensions. Finally, we have no idea if anyone else is using our metadata or the metadata that is embedded in the pages of our partner sites. No one has written to enquire about using it, or to complain that it does not suit their needs. Cheers, Prue [log in to unmask] Dan Brickley <danbri@DAN To BRI.ORG> [log in to unmask] Sent by: cc General DCMI Subject discussion Re: DC and legacy implementations list [SEC=UNCLASSIFIED] <DC-GENERAL @JISCMAIL.A C.UK> 03/12/2008 11:38 AM Please respond to Dan Brickley <danbri@DAN BRI.ORG> Prue Deacon wrote: > There seems to be an assumption > that, when the technical coding framework is in place, then everyone will > create/maintain accurate, consistent metadata content and (hey presto!) the > Semantic Web will connect everybody to everything and will always be > up-to-date. I'm sorry you've got this impression re up-to-date, and I'd like to track it down to source. Have people told you this, or written this, ... or you're inferring it from other behavour and claims? Why on earth would it always be up to date? Does anyone think this? (Regarding "connect everybody to everything", we already are all connected already, but enough of that for now...). My sense of Semantic Web (and my motivation for the project) comes from pretty much the opposite perspective. The fact that metadata is so expensive to create and difficult to maintain is what motivates me to try to make sure it can be as widely linked and re-used as possible. A big aspect of the quality story is making sure the metadata is actually used and human visible, which is why wiring it up to HTML formats (the GRDDL and RDFa technologies) and to query systems (hence SPARQL) are both important. RDF does come with some conceptual baggage for sure, but well we never managed to define a non-RDF DC cleanly (including an extension model --- dc.publisher.address-2.postcode etc ) that people were happy with. I hope the boring/ugly bits of RDF (definition of 'literal' etc) can be swept under the carpet through better docs, without leaving too many lumps... cheers, Dan -- http://danbri.org/ ______________________________________________________________________ "Important: This transmission is intended only for the use of the addressee and may contain confidential or legally privileged information. If you are not the intended recipient, you are notified that any use or dissemination of this communication is strictly prohibited. If you receive this transmission in error please notify the author immediately and delete all copies of this transmission."