2

Hi folks,
Some years ago I decided to switch from Windows to Kubuntu. And I am still incredibly happy with KDE respectively KF5.
But originally I changed because of my inadequacy to keep my files together and so I stumbled upon the promise of a Semantic Desktop.
I know, there are several ways to solve my problem, even without computers. Butt...
The architecture of NepomuK was exactly what I dreamed of, the vision was activly maintained, and somehow I felt
at home within those topologies of components building the foundation of a semantic PIM.
I am talking about Akonadi, Nepomuk and Virtuoso, and it feels beeing nostalgic about Sega Saturn or the Hedgehog.

To get to the point:
What are my chances to:
> reanimate Nepomuk
> use Virtuoso as Akonadi Backend
> try to establish ODBC-Backends in other KF5-Applications
? ( I have no idea how to do this, but I think I have to! )
Okay, okay:
> Nepomuk will be an optional alternativ to Baloo
> use semantic data on your desktop, if you like or need it
> make the maximum depth of semantic reasoning configurable individually
> make the vocabulary of semantic configurable individually
> make the maximum age of indexed files configurable individually
> warn the user to tidy up their damn data, before complaining about 100% cpu blabla, sorry
> use baloo on your mediacenter, and
> use baloo on your fairphone, and search your personal data mine sets
> everyone is happy, nice and balanced.
And finally:
> port Freecad from coin3D to Qt3d (vulkan-backend)
> port Kad, Kraft, Kmymoney to Qt5 and integrate those closer
> with Kontact
Is a deadline of 2025 realistic, if I scratched c++ on the surface until now?

cheers
dom
ps: I forgot to mention drag & drop semantic between URIs (including xpointer-URIs) desktop-wide

Akonadi was not very well-received. It still exists, but is smaller.
What practical issues do you have with Baloo over Nepomuk?

Answering as one of the devs bitten in the past by the Nepomuk project, which might make me less than optimistic about your idea. But I'll try to be as objective as I can.

> reanimate Nepomuk

If you think of what are your chances to port it to KF5, I expect it not to be that problematic. If you are asking of the chances for nepomuk-related patches to be accepted by other projects, that will go much harder.

The idea behind Nepomuk (and semantic web/desktop) was really great, but the problem is that it is a "paper tiger" - it works on paper. When implemented, RDF generally fails. For its slowness, lack of structure (which, strangely enough is one of its selling points) etc. Virtuoso is one of the fastest RDF databases, and yet it is painfully slow (I'd say slow by design of RDF) even without deductive reasoning (inferencing that RDF should make possible).

> use Virtuoso as Akonadi Backend

I guess using the relational (SQL) part of Virtuoso would be easy since akonadi already has alternative backends. Connecting that data to the semantic stuff might not be trivial.

> try to establish ODBC-Backends in other KF5-Applications

This might be nice - to have a single database behind all KF5 applications. But this would require a lot of work (and, no need for ODBC, there are other ways to achieve the same).

> Is a deadline of 2025 realistic, if I scratched c++ on the surface until now?

This is something noone could predict now