[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message] RE: The perils of P18S (was Re: Why RDF is hard )
>> RDF/XML may be butt-ugly and confusing, but IMHO the model is quite >the >> opposite. > >I have no problem with RDF data model although I don't see anything >special about it. Quoting Joshua Allen it makes it "...possible for anyone in the world to publish assertions about things in such a way that everyone else in the world can locate them, use them, and in turn make assertions about them." I don't know of any other standard with that power. >> I have yet to see an XML application that couldn't be done without >XML, so >> that argument isn't particularly strong. The point is that in the same >way >> a lot of applications are considerably easier to implement using XML, >> quite a few applications are considerably easier with RDF, and such >> practical applications are now popping up all over the place. > >You can't compare XML with RDF. I was only trying to demonstrate that your argument was effectively impossible to refute, but didn't actually prove very much : "I have yet to see a single successful application of [technology X] that couldn't be done without [technology X]." If you had said "I have yet to see a single successful application of RDF that couldn't be done better, or more easily without RDF", then I would have had more to chew on. XML was generally accepted with minimum >controversy except from SGML folks. RDF, on the other hand, is >extremely controversial. There are quite a few RDBMS people who would also disagree on the first point. But again, that something causes controversy really says little about its merits. There are concerns, fine - that's what xml-dev and other such forums are for. I don't think there's any need for someone to walk in front of horseless markup carrying a red flag though. Open Directory Project and RSS 1.0 use RDF, >but they didn't really need it. In both cases RDF adds considerable advantages, at very little cost. I admit that if you judge the applications in isolation and with frozen requirements then there isn't much benefit. For a one-off, standalone directory there is little to be gained from using RDF (though not having to build a model entirely from scratch is a time saver). The use of RDF means that dmoz can be easily extended in all directions. The RDF nature of RSS 1.0 means that it is also extensible without reinventing the wheel for every new feature. A slightly different, less vertical view is required to see the greatest advantages. For example, RSS 1.0 can *with no modification* use categories from the Open Directory. An blogger-like application that took advantage of this could for example place its content directly into the Open Directory, and going the other way, when displaying the blog entries it could pull out 'what's related' items from the Open Directory. Ok, there's work involve in setting up the processing to do this, but a large, potentially difficult part of the work (the model) has already been done. What's more the same processing system could be used with other sources of data, for instance the images and documents produced by Adobe's tools (which now incorporate RDF). I don't know what practical RDF >applications you are talking about. If they are indeed 'popping up all >over the place' now, what were RDF folks doing all this time? The WG group have been very busy improving the specs and developers have been producing things along the lines of the list I've pasted again below. Cheers, Danny. Data integration : B2B web service mediation Catalogue integration Database integration example Financial Portals Gene Ontology Mozilla eScience Data Grids Data-dependent agents : Financial Assistant ITTalks Jema Shopping assistants Virtual Travel Agent Knowledge management : Community formation Community portals Helpdesk support OntoShare - community of practice support PatMan PlanetOnto Sun GKE ePerson Semantic indexing and semantic portals : Community Arkive Community portals Context aware links Curriculum Online Distributed topic portals HP Portal ITTalks Museum portals MusicBrainz PlanetOnto Score Semantic tagging TAP semantic search Personal information management : Bibliography workbench Community bookmarking Event tracking Genealogy assistant Haystack Ideas workbench Jema Mozilla ePerson Metadata for annotating and enriching : Annotea Assumption tracker Bibliography workbench Community Arkive Community bookmarking Distributed topic portals EARL Gene Ontology MIT/HP SIMILE project Metadata for description, discovery and selection : B2B trading market-places DCMI registry Edutella HP Portal MIT/HP SIMILE project MUSE Recommendation Networks Scholnet SeLeNe Semantic tagging Sun GKE Web service description and discovery Metadata for media and content : Adobe XMP Arkive internal MIT/HP SIMILE project Knowledge formation : Assumption tracker Bibliography workbench ClaiMaker/Scholonto Community Arkive Community formation DMOZ - Directory Mozilla - open directory Ideas workbench SWAP - semantic web and peer-to-peer Catalogue and Thesaurus management : Catalogue Management Catalogue integration DMOZ - Directory Mozilla - open directory Thesaurus management Syndication : Event tracking Rich Site Summary/RDF Site summary Syndication [1] http://www.w3.org/2001/sw/Europe/reports/open_demonstrators/hp-applications- survey.html#Catalogue%20integration see also http://www.w3.org/2001/sw/Europe/reports/open_demonstrators/hp-applications- selection.html
|
PURCHASE STYLUS STUDIO ONLINE TODAY!Purchasing Stylus Studio from our online shop is Easy, Secure and Value Priced! Download The World's Best XML IDE!Accelerate XML development with our award-winning XML IDE - Download a free trial today! Subscribe in XML format
|