Greetings, I haven't been following the original thread, so I'm responding just to Jürgen's point here.
On 2013 Apr 18, at 12:21, Jürgen Jakobitsch SWC wrote: > i do not really understand where this "the developer can't sparql, so > let's provide something similar (easier)" - idea comes from. > > did anyone provide me with a wrapper for the english language? nope, had > to learn it. I do and don't agree with Jürgen. Agree: SPARQL isn't a huge challenge; once you've got the idea of RDF, it's straightforward to grok SPARQL. Disagree: every little technical barrier cuts down the number of people who will investigate a technology (or a solution, or an application, or...) by some factor, and the lower that barrier is, the smaller the loss-fraction will be. But Agree: a supposed difficulty of learning SPARQL is not the SW's problem. There are so _many_ technical barriers to getting going with SW tools that it's the _number_ of barriers that dominates the overall loss-fraction, rather than the size of any barrier in particular. I cannot believe that making SPARQL easier to use will make a difference here (though I'd be interested to be proved wrong). In trying to evangelise for the SW in a highly techie but non-CS community, and in teaching the highlights in a non-techie community, I have come to the conclusion that it is damn hard to get into -- much more so than other technologies -- enough that people essentially _won't_ get into it unless they need to (because they've been instructed to learn about it) or unless they have a prior intellectual interest in such a way of thinking about the world. I don't think there's a short explanation of this. But relevant features of the SW world are: * There are _lots_ of components that have to be working before you can start playing around. This was partly addressed a few years ago with 'SemWeb in a box' (was it Danny Ayers who was involved with this?), but that's not sufficient. * The people involved are primarily CS academics. The specific problem with that is that this community is broadly (for various reasons) too good at handling complicated and variously buggy systems; is interested in, or amused at, handling such systems; and so is willing to cheerfully tolerate a much more ragged experience than most technologists. * CS academics, II: the 'interesting' stuff about the SW is the reasoning (and I'm not a CS academic, but I think that's interesting too, so I'm not disagreeing with this per se), but that means that a lot of the SW noise is about an aspect that most people find both perplexing and far from obviously useful. * Trivially, but _not_ ignorably, names: 'semantics', 'ontology', and reasoning (before we start talking of 'supervenience') make people think it's much harder and more arcane than it is. The major success of the LOD movement in terms of 'mindshare' is that it's evaded some of these problems, in the sense that it's fairly easy to explain and consume; the remaining problem is that providing it on the server side still requires negotiating these bear-traps. All the best, Norman -- Norman Gray : http://nxg.me.uk SUPA School of Physics and Astronomy, University of Glasgow, UK
