[Nepomuk] Plasma Media Center - semantic desktop advices

[Sebastian Trüg]

And please use the query API instead of coding SPARQL queries whenever
possible. :)

As for changing the folders to be indexed: there are plans to provide
configuration classes. So far it is really about changing config files.
In the long run that should not be needed anymore. I am talking about
the future here where everything is indexed. :P

Cheers,
Sebastian

On 06/05/2010 08:57 PM, Alessandro Diaferia wrote:
> 
> 
> 2010/6/5 Vishesh Handa <handa.vish at gmail.com <mailto:handa.vish at gmail.com>>
> 
>     Hey Alesssandro
> 
>     Creating such a backend shouldn't be too difficult. The database
>     doesn't need to be created, you could just use Nepomuk's database.
>     Currently, Nepomuk uses Soprano for storing RDF data, which in turn
>     uses Virtuoso. There are 2 things you will need to do -
>     1. Make sure all the data which you want accessible via PMC is
>     indexed. You can use Nepomuk's KCM to do that, or you can directly
>     edit the nepomukstrigirc file.
> 
> 
> The user is supposed to give such information. Is there an API for
> adding paths to the indexing or shall i just programmatically edit
> nepomukstrigirc?
>  
> 
>     2. Construct Sparql queries to segregate data based on its type.
>     (Music, Video, Pictures, etc)
> 
>     The Nepomuk File Ontology is exactly what you're looking for. It has
>     classes like nfo:Audo, nfo:Image, nfo:Video, or if you want to go
>     for a more generic option - nfo:Media. Example -
>     "select ?url where { ?r nie:url ?url. ?r a nfo:Audio. } LIMIT 10"
>     This should give you 10 Audio Files.
> 
> 
> Ok that's great. Fortunately I had to learn SPARQL for my bachelor's
> thesis project so I'm rather comfortable with it; just need to have a
> look at those ontologies :)
> 
> 
>     I think Nepomuk is perfect for what you want to do, cause it already
>     segregates the data, and you won't need your own database.
>     Additionally, you could even search by Artist, Actor, Director, and
>     other properties. Plus, if those properties aren't present in the
>     File, the Web Metadata extractor should get those values ( GSoC
>     project 2010 )
> 
>  
> That's rather a nice project indeed then :)
> 
> 
>     Another thing that could be done is that you could display
>     information about different Artists, Actors, Movie Directors, Imdb
>     ratings, etc and you will have to do next to nothing to get that
>     info. Nepomuk should handle all the messy work. You just need to
>     perform the queries. ( Wait for KDE 4.6 for this to get implemented )
> 
>  
>  That's our real purpose of course :)
> 
> 
>     *Far fetched Idea :* If I manage to implement Metadata sharing
>     properly, you could theoretically search through other machine's
>     Nepomuk database. So, I guess that would add a social feature to the
>     PMC.
> 
> 
> That would really be amazing :)
>  
> 
> 
>     - Vishesh Handa
> 
>     On Sat, Jun 5, 2010 at 3:32 PM, Alessandro Diaferia
>     <alediaferia at gmail.com <mailto:alediaferia at gmail.com>> wrote:
> 
>         Hello list,
>         you probably have heard about Plasma Media Center (PMC).
> 
>         I would need some advice from regarding media type files handling.
>         Little briefing: PMC allows different browsing backends to be
>         inserted in order
>         to allow different kind of file browsing, This leads to giving
>         the chance to the user to both
>         browse among his local files or among e.g. videos from youtube,
>         pictures from picasa and so on...
> 
>         Regarding the local file browsing we need to do something like:
>         the user selects the folder PMC should search for media files, a
>         database is created with such files
>         and later the application will always show such files
>         categorized and what not. In addition to this
>         we'd like to separate pictures, videos and audio files. Later
>         one backend for each type (music, picture and video)
>         will make use of such databases in order to allow the user
>         browse through them.
> 
>         Now your help: what's the best way for doing this from the
>         Nepomuk point of view?
>         I hope my explanation was clear enough. Of course just ask for
>         any clarification.
> 
>         Thanks in advance for your help.
>         Regards
>         -- 
>         Alessandro Diaferia
>         KDE Developer
>         KDE e.V. member
> 
> 
>         _______________________________________________
>         Nepomuk mailing list
>         Nepomuk at kde.org <mailto:Nepomuk at kde.org>
>         https://mail.kde.org/mailman/listinfo/nepomuk
> 
> 
> 
> 
> Thank you very much for your clear explanation on what it has to be done!
> 
> Cheers
> -- 
> Alessandro Diaferia
> KDE Developer
> KDE e.V. member
> 
> 
> 
> _______________________________________________
> Nepomuk mailing list
> Nepomuk at kde.org
> https://mail.kde.org/mailman/listinfo/nepomuk