Since the development about kat slowed down to zero, I was looking for another solution, and finally, some weeks ago kitten, in the meantime renamed to strigi, rised as a phoenix.
The development goes on quite fast, a first snapshot can be seen on kde-apps, a homepage is up and running, a html interface was uploaded in svn and kfile is used in future versions to index stuff also.
There are also others helping with the development of strigi and thinking about the future.
What I really like about the discussions of this new engine is that they often speak about commenting on files and tagging these – I really, really like the ide of having the possibility to add tags and keywords to almost every file available on my machine. It would help much with sorting and finding stuff, well, at least as long as I can search it. Although I do have beagle which helped me one or two times with finding stuff, it’s no way to sort and organize stuff.
And since I started working a bit at least with delicious I now know how big the advatanges are to have the abilitiy to add as many tags as I want to to specific subjects – I can find them later quite easy!
But besides the tagging, there started some discussion about the new search machine: here are some doubts that the extended attributes are the right way to store data (I think they are, btw.). As well as with the other blog post, the comments are worth a read and look at different aspects of the possible solutions.
They also show another solution I never heard of: Tracker is another indexing mechanism, comming from gnome/freedesktop. Sounds interesting, but I think at the moment the best example how it should like and work later on is still leaftag. Haven’t seen any development there lately.
Anyway, whatever solution is chosen, I just hope that it will be more or less compatible with Gnome – I want to have the option to just launch Gnome instead of KDE without the need of retagging everything.
But even more important is that the new solution will be also compatible with digikam, amarok, kontact, and so on – imagine a system where you can just read all KDE files and information and where these are provided just as they should!
Btw., talking about tagging and KDE 4 – I’ve recently read this blog post about session handling, and it sounded pretty interesting to me. Just the idea is as new as it is sensible, and I would like to see that in KDE4! There is also a small discussion in the comments whre I tried to embrace the idea and others add their thoughts. Nice to read.