> On Saturday 06 September 2008, Sebastian Trüg wrote:
>   
>> On Saturday 06 September 2008 10:20:42 Sven Schwarz wrote:
>>     
>>> I would say, your proposal really makes a good and concrete starting
>>> point!
>>>       
>> thanks. I feared you would not like us trimming down the ontology since
>> here "real-life-I wan't that stuff working now" developers meet "I want
>> this stuff to be Überclean" researchers. ;)
>>     
Ooops, maybe, I outed myself as a non-conform or even strange researcher 
here ;-)
Actually, as I also have a (strong!) hacker part in my soul, and I hated 
it to struggled with such pure-research driven stuff a lot myself. I 
also wanted to have running prototypes that show how things can work. 
But most of the time, I had to solve problems that are not interesting 
at all... One of the results is, that I will never ever include a 
(prototypical) RDF _repository_ into any software again if I am not 
forced to do so... ;-)

But actually, some of the concepts make sense and turned out to be 
really useful. For example, having an "ontology" (according to Ludger 
van Elst I should better call it "shared conceptualization") of the 
observed user operations (NOPs) really makes sense in the long run... 
Some day (say after the next world war) there will be a real good 
inference engine and whatnot that works just perfectly (and efficiently 
;-)) on RDF data... These days will be really great... OK, until then, 
we just take the NOP ontology as a quite meaningful schema that is a 
shared conceptualization among people that want to make use of 
automatically observed user operations... That's not too bad, I think! :-)

As I said two weeks ago (within a talk about my future life-planning): I 
like discussing about theoretical research and bleeding edge algorithms. 
But, I don't love them just for the sake of being so weird. If there 
isn't any use for the theory I will stop talking about it as soon as my 
pint of beer is empty ;-)
So, let's take the interesting (parts of) research results and put 
something useful and running together with it. If that works, let's go a 
step further.


In the same sense, I also absolutely agree on Aaron's words:

Aaron J. Seigo wrote:
> as one of those real-life-i-want-that-stuff-working-now developers, i'd just 
> like to point out that this is also just a starting point for my project 
> (plasma). 
>
> there were bits of the ontology that we looked at and decided were 
> interesting 
> but beyond where we could realistically land with the first steps. so i 
> expect 
> the ontology to mature closer to the original one over time in the actual 
> implementations; how close will be something only time can tell at this point 
> i think =)
Yesss...

'Nice weekend!
Sven
_______________________________________________
nepomuk-kde mailing list
[email protected]
http://lists.semanticdesktop.org/mailman/listinfo/nepomuk-kde

Reply via email to