- From: Paul Tyson <phtyson@sbcglobal.net>
- Date: Thu, 23 Aug 2012 22:05:01 -0500
- To: Kingsley Idehen <kidehen@openlinksw.com>
- Cc: public-ldp-wg@w3.org
On Thu, 2012-08-23 at 13:02 -0400, Kingsley Idehen wrote: > All, > > Of late, I've been writing a series of posts [1][2][3] that demonstrate > why Turtle is such an important syntax for crafting RDF documents. > Basically, its solves the biggest challenges to Linked Data > comprehension which include the following: > > 1. triple visibility > 2. triple comprehension. > > Historically, RDF/XML was an utter dud re. the above. History will > record this as one of the biggest snafus of an era. By that I mean, > putting such a misguided syntax at the front door of something so > important. Sounds like a lot of people have some sort of hateful grudge against RDF/XML. I won't dispute its shortcomings, nor defend W3C's tardy acceptance of other RDF exchange syntaxes. But, as a bridge to RDF from hierarchically-structured data (e.g. XML), it really is quite elegant and useful--a point that I wish the RDF/XML-bashers would recognize. As for newcomers mistaking RDF/XML for just another XML schema and coming away with a greater dislike for both XML and RDF, that's an unfortunate outcome due (at least partly) to poor publicity and advocacy. I suppose people had trouble with XSLT and XML Schema for the same reasons. Any semantically-laden use of XML will cause this problem, because XML is usually taught just as syntax and most people do not expect to have to think very hard about the meaning. What this comes down to for LDP systems is that while RDF/XML need not be front and center, it would be good if LDP systems accepted RDF/XML input so that those who need to work with large corpora of structured documents would have an easier path. Regards, --Paul
Received on Friday, 24 August 2012 03:05:45 UTC