It’s just data

FFKAR

Timothy Appnelwhat I humorously refer to as the "format formerly known as RSS" or FFKARI'm in.  Question: why must rdf:about and link be redundant?

The answer as to whether or not items' rdf:about is redundant lies in the way that RDF is constructed. The value of an rdf:about attribute is that it is always recognized by an RDF parser as being a URI-ref--viz. a parser should always say in its output that "this is a URI-ref". It's a basic datatype of RDF. OTOH, the value of the link element (the range of the property) is a literal. This is basically an untyped string: for all the RDF parser knows it could be an integer, a URI, a name, or something else.

Whether this makes any practical difference or not is another matter. I'm sure that most aggregators, adapting their existing RSS 0.9x parsing techniques, derive the item's URI from the link element. Some advanced RDF tools can even turn a literal-datatype into a URI-datatype, but you wouldn't expect most to. Are there any RDF tools that currently (or will ever) rely on the Item being non-anonymous? Answers on a postcard...

And as for the argument that the rdf:about attribute lets you refer to the same item in two different feeds, note that you can achieve the same effect by declaring rss:link to be a daml:UnambiguousProperty, and then use the much ballyhooed "smooshing" technique to merge items.

Posted by Sean B. Palmer at

How is that acronym pronounced, I wonder?

8^(

Posted by Patrick Logan at

And in a completely self-congratulatory note I would like to point out that Aggie RC4 reads both Timothy's and Sam's new feeds just fine.

/advertisement :)

Posted by joe at

The short answer is: backwards compatibility, so 0.9-format readers could read 1.0.

Note that it's not "must" but "should". In practice, when they differ, rdf:about is the permanant URI for the item (like the newly introduced <guid>) and <link> is the fetchable URL.

Posted by Ken MacLeod at

Ken, take a close look at my ffkar feed... the rdf:about "points to" the the information on the entry in a format that would be understandable to a rdf tool, and the link "points to" the information on the entry in a format that would be understandable to a browser. In both cases, the urls are intended to be permalinks.

Posted by Sam Ruby at

Reading the RDF specs, in sections 3.3 and 3.5 might be helpful to understand what's going on with about, aboutEach and the like.

http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/

Posted by anonymous

at

I got slightly corrected in #rss-dev.

rdf:about doesn't imply "pointing to" anything, it's a simple (globally unique) identifier.

rss:link, then, is the URL pointer to format understandable to a browser.

Normally they can be the same thing because the URI of the form understandable to a browser makes a fine globally unique identifier too.

If they differ, it's usually the rss:link that changes over time and rdf:about that is permanent.

To point to additional RDF, RDF libraries will automatically retrieve rdf:seeAlso resources, and (as an option?) may try anything listed as an rdf:resource, but usually not try to resolve anything in an rdf:about. I haven't played with those parts in any RDF libraries, so hopefully someone who has can better answer.

Posted by Ken MacLeod at

RSS revolt

People appear to be sick of the syndication format wars.  Some are protesting, some are creating new formats with new names, others are simply boycotting.... [more]

Trackback from dive into mark

at

if you are working on bing search engine and you are get the more information on bing search engine. some person are working and get the true result. the result are true and right if you are earn the more knowledge on this website to first go to the website.

Posted by bing rewards points at

Whenever different things are organized to work in a better way, there are some other things that can make all things possible especially when there are new skills to be adopted.

Posted by PhD Dissertation Writing Services at

Add your comment