- From: Steve Harris <steve.harris@garlik.com>
- Date: Wed, 2 Nov 2011 14:29:17 +0000
- To: Ivan Herman <ivan@w3.org>
- Cc: W3C RDF WG <public-rdf-wg@w3.org>
OK, but I'm not comfortable with the idea of some URI being used as an alias for some other URI. Having stores and tools silently map from one to the other is going to be confusing for users, and having the stores and tools treat them as equivalent is also going to be strange. - Steve On 2011-11-02, at 10:42, Ivan Herman wrote: > Steve, > > I do not think we should get into the discussion on how microdata and HTML5 should handle this. Let the SWIG task force figure this thing out. > > The issue for this WG is more general. If you edit any RDF file, if you write some programs for managing RDF data, you may have to write down the full URI-s for the various RDF(S) terms. It may only be me, but I never ever remember the URIs, so the first thing I do for, say, a turtle file is to find another turtle file and cut and paste. Having a handy and easy-to-remember URI for, say, 'type' would be helpful. > > We may of course decide that this is not a real problem and dismiss the issue as a historical artifact... > > Ivan > > > > > On Nov 2, 2011, at 11:16 , Steve Harris wrote: > >> Does it have to be a URI? What about using "a" as in Turtle, or "type"? Makes it more obvious that it's an alias, not a URI. >> >> I think using http://something.example/type and silently transforming it to http://www.w3.org/1999/02/22-rdf-syntax-ns#type will be confusing for users. Especially if real-world systems use both in practice. >> >> It could also be defined as an entity in HTML5, e.g. >> >> <!ENTITY type "http://www.w3.org/1999/02/22-rdf-syntax-ns#type"> >> >> then it's just >> >> <link itemprop="&type;" href="http://type.example.org" /> >> >> - Steve >> >> On 2 Nov 2011, at 09:43, Ivan Herman wrote: >> >>> I thought that this working group needs some more discussions, because we have already solved all our issues:-) >>> >>> More seriously: there is an issue on which this WG, maybe, could/should have an opinion. There is a long discussion going on in the Data in HTML SWIG Task Force[1], while looking at the microdata->RDF mapping. The issue is around the ugly URI that one has to use for rdf:type... >>> >>> [[[ >>> Background: >>> >>> - at present, an item in microdata can only have one type, that is syntactically accepted via their @itemtype attribute >>> - people have expressed the need to have several types for a single entity >>> - the only way to do it now is to explicitly 'encode' a type in microdata, something like >>> >>> <link itemprop="http://www.w3.org/1999/02/22-rdf-syntax-ns#type" href="http://type.example.org" /> >>> >>> - there is no way in microdata to abbreviate a URI with a prefix:extension syntax, ie, to be able to write rdf:type. In other terms, users _must_ write down the full URI. Let us face it, nobody can remember that stuff (I know I can't:-) >>> - the solution that is emerging is to define a separate 'type' attribute in, say, schema.org, ie, that people could say >>> >>> <link itemprop="http://schema.org/type" href="URI-to-Type" /> >>> >>> because people might remember that more easily (which is true). Then... I presume microdata->RDF mappings would have to _know_ that particular attribute and turn it into rdf:type. This is really ugly, but that is the direction things seem to go. >>> (- of course, the proper solution would be for microdata to allow several types in one statement (like RDFa allows). But the microdata community is reluctant to do that.) >>> ]]] >>> >>> This discussion does raise a more general issue, however: does it make sense for us to think in terms of URI aliases for the RDF(S) terms? Something like http://www.w3.org/rdf/type, or something similar. It would make the human authoring of RDF easier (even if all RDF syntaxes allow for some sort of a prefix definition, which greatly alleviates the problem). >>> >>> I realize this may be a huge can of worms, in terms of deployed applications, specifications, etc. In other words, probably the only way to do that would be to define some sort of a canonical alias? redirection? etc, of humanly readable URI-s to the current ones. And even that may be hairy. But it may be worth some discussions... >>> >>> Thanks >>> >>> Ivan >>> >>> >>> [1] http://www.w3.org/wiki/Html-data-tf >>> >>> >>> ---- >>> Ivan Herman, W3C Semantic Web Activity Lead >>> Home: http://www.w3.org/People/Ivan/ >>> mobile: +31-641044153 >>> PGP Key: http://www.ivan-herman.net/pgpkey.html >>> FOAF: http://www.ivan-herman.net/foaf.rdf >>> >>> >>> >>> >>> >> > > > ---- > Ivan Herman, W3C Semantic Web Activity Lead > Home: http://www.w3.org/People/Ivan/ > mobile: +31-641044153 > PGP Key: http://www.ivan-herman.net/pgpkey.html > FOAF: http://www.ivan-herman.net/foaf.rdf > > > > > -- Steve Harris, CTO, Garlik Limited 1-3 Halford Road, Richmond, TW10 6AW, UK +44 20 8439 8203 http://www.garlik.com/ Registered in England and Wales 535 7233 VAT # 849 0517 11 Registered office: Thames House, Portsmouth Road, Esher, Surrey, KT10 9AD
Received on Wednesday, 2 November 2011 14:29:55 UTC