[Nepomuk] How to config virtuoso-t?

[Vishesh Handa]

On Thursday 09 Jan 2014 06:34:53 Michael Jerger wrote:
> Am Mittwoch, 8. Januar 2014, 13:08:30 schrieb Vishesh Handa:
> > The next version of KMail is not going to be using Nepomuk. We're moving
> > away from virtuoso and the whole "semantic web".
> > 
> > So you should have a better KMail experience with 4.13
> 
> Sounds good for KMail - bad for the semantic web idea.
> Would your new Mail-/Contact-/Calendar-Storage have an semantic interface? I
> like the idea, to have RDF acces to all my stuff (which does not mean, that
> all data has to be stored redundant in virtuoso also)?
> 

Unfortunately no. We're dropping RDF completely. While RDF has great potential 
it has a greater learning curve and optimize RDF storage is hard. Specially 
since the Nepomuk project was also responsible for fast indexing and searching 
in KDE.

We still believe in the overall ideas of the Semantic Web of having higher 
level resources which define things, but we're not going to be using RDF as the 
main storage mechanism.

Maybe someday someone will write a filter layer which shows the data as RDF.

> > > enterprise service I've similar problems - so do you have made any brain
> > > work for this topic allready? May I be interested in your ideas and
> > > experiance?
> > 
> > Most of the Nepomuk developers had grand ideas about sharing the
> > Nepomuk data. In fact I once prototyped a distributed nepomuk service
> > which would register itself on the local network and allow sparql queries
> > to be run on all the systems in a local network.
> > 
> > That was terribly impractical because of the issue of privacy. We then
> > thought about implementing some sort of an ACL for the rdf data, but the
> > more we tried to plan it out, the more complex it got. Eventually, nothing
> > got done.
> 
> Okay - thanks for your straight answer. I've seen some whitepapers about
> "data based authorization" in semantic graphs and I agree totaly - it seemd
> to be very complex. Maybe this complexity is apropriate for the enterrise
> context. I will give it a try.
> 

I'd be interested to know how this turns out. Virtuoso currently does offer 
some basic ACL like features with respect to RDF graphs. You can set 
permissions on who is allowed to access and modify certain graphs.

In Nepomuk, each application's data is in its own graph. This did 
*theoretically* allow us to stop different applications from messing with other 
application's data.

> 
> Michael

-- 
Vishesh Handa