Trying to use #Nepomuk and #Akonadi in #KDE 4.8 is still an exercise in pain and frustration. So far, I have hit KDE bug 289932 (https://bugs.kde.org/show_bug.cgi?id=289932), KDE bug 259355 (https://bugs.kde.org/show_bug.cgi?id=259355) and something only reported in Gentoo so far (bug 401155, https://bugs.gentoo.org/show_bug.cgi?id=401155).

Despite how good Nepomuk may be for some people (like for #Conquirere, #Bangarang, or whatever), I really don't feel like messing with it for #Tellico. What a mess.

The API keeps getting adjusted (look for DEPRECATED in the Nepomuk classes). Performance problems keep popping up (read the KUbuntu-users list for January). Permissions and security for the Nepomuk database are still not clear to me. The ontologies are all over the place, overlapping and contradicting (nao vs. nie, etc. Maybe that's a "feature"...).

I'm not a complainer. At least, I don't think so. But I've tried multiple times to either sketch out an ontology for Tellico or figure out how to start integrating Nepomuk. And every time so far, I give up within a month or so. The ontology thing really drives me nuts. It feels like the problem that Karen Coyle blogged about recently (http://kcoyle.blogspot.com/2012/01/bibliographic-framework-rdf-and-linked.html).

I think what it comes down to is that I'm just not smart enough for this. The guys who are developing Conquirere and Bangarang and KMail have something I don't. Having an simple embedded copy of Sqlite in Tellico feels really attractive right now.
Shared publiclyView activity