KDE Frameworks 5.60 Bringing More Baloo Optimizations

[StefanBruens]

I was tired of Baloo way of indexing... It passed by a lot of things that should be indexed to me, tried everything even doing a revision in .config/baloofilerc, PDFs and so ignored here and there, about 3 million files to only index 570024.

Wheres the bugreport? Crying in some forum won't help.

[Vistaus]

wolfyrion It definitely grinds quite a lot on a fresh install, so if you have a large $HOME and transfer it often into a new setup I can understand your point perfectly well. But if you are working with lots of valuable company documents (> 100,000 files) generated by different people over more than a decade, then content indexing is a must-have.

We have locate for file locating and Catfish for content locating. Catfish works very well on Plasma and doesn't create huge indexes and is not a CPU hog

[Happy Heyoka]

I have Baloo enabled, and I use it and am happy to have it.
As a longer-term KDE user what annoys me is that for pretty much the same resource utilisation, Baloo produces (in certain edge cases) worse results than Nepomuk - when indeed, the opposite was meant to be true.
Yes, I know the whole semantic search thing is confused by the file system metadata thing, but in general, we still have some distance to go to reach the "dream" of semantic desktop search (and as someone with TB of docs online, I would love that to work).
At some point, I should volunteer some time to work on the indexing - the resource consumption should be far more tunable - including a mode that says "just use everything I have and get it finished already" for things like a new install.

[RomuloP]

Wheres the bugreport? Crying in some forum won't help.

It is nowhere because I tape fixed my problem, was not crying in some forum, was commenting on the state of it, I think people deserve to know what to expect, anyway nothing against baloo or the devs, I do understand KDE devs do not have all the time they need to fix everything, some of us do not have time to feed bug-trackers, some of us look on forums to get some glimpse about the tool to save time too. In future when I have time to report and follow the bug-tracker I will do it. It was just easier to build a tree walker script and get it working in a min.