Impressive. But can't spare several GB for file index. What about xapian? Is also kind of file indexer and search engine. And it is already provided in -current. For me I can use is it to look for some term word in multiple pdf files. However it is in most cases author/title or inline search. I tried to build xapian myself on 14.2 but I failed. More or less these things I should have in my head rather than on the computer. I tried myself little work-around - convert pdf to text and then use simple text search - whatever - cat | grep, awk - little bash. That's kind of sort.

Edit: but of course this is all about choices: to have a choice to use and to have a choice to not use.

My baloo index is roughly 7% of my total data in ${HOME}:

1 members found this post helpful.

Why are you guys talking about file indexing? And why do you use it?
I've never used file indexing, and bash/coreutils do provide som rather hefty tools to find files if need be.

Never needed anything like that though. I try t keep my files in some kind of order.

2 members found this post helpful.

Finding files is easy using locate in a console. Finding file content is not so easy in a console. Baloo shines in indexing the contents of your files.

It's a killer app for sure. It seems that indexing stopped at about 60000 of 230000 files in my ~/.

From syslog:
KDE Baloo File Indexer has reached the inotify folder watch limit. File changes will be ignored.

A lot of the files are duplicates in different folders, so maybe it's smart enough to skip duplicates.
I'll move my kernel trees someplace else just to see what happens

Inotify folder watch limit can be set higher apparently.
fs.inotify.max_user_watches=524288 in sysctl.conf should be enough.

1 members found this post helpful.

Traditionally I have enabled only basic indexing in baloo, but I may try the full shebang again. Being able to hit alt+f2 and search for a file directory name is powerful. Searching content is never something I thought I would need but I'm starting the see the appeal.

For what it's worth, with basic indexing enabled my baloo index has never gone above a nominal amount of disk space usage.

I have no opinion either way on baloo because I don't use KDE, but when searching for content I usually just:
Works quick enough for me anyways.

Edit,
However this does not work well with archives.

Ahh, good point. I can understand that. But what are some practical uses for that for normal private individuals? Can anyone mention any common examples? I just can't understand how file indexing can be such a big thing..

I have to admit that my comments about liking, the stability and resource use of KDE(5) is for someone who has never used any file indexing functions.

Sure, that works fine for plaintext files.

What were talking about here is indexing software that will scan the content of the likes of PDF files, Libreoffice documents, etc.

Baloo handles all of that and more, and it is very small (the sources were under 1MB last time I checked which was a while ago).

Its main problem right now is error handling and recovery or lack thereof. This has been a fairly longstanding bug:

https://bugs.kde.org/show_bug.cgi?id=368557

Recoll (see my earlier post in this thread) uses xapian under the hood.

Sounds like pretty impressive capabilities.

Is a few hours all it normally takes? I've left mine on all day and it's still going. My home isn't that big, either; only 144 GB used. At the moment it seems to be usable despite the indexer running, but that hasn't been the case the whole time. The index is currently 5.3 GB. I'm just wondering why this seems to be having so much trouble on my system. I would like to use it, and the rest of my Plasma 5 experience has been very nice so far.

True, and that is pretty much all that I need. I do know a few people that would benefit from actually using baloo, but they are just as stubborn as I am and just don't use it.

While I do have a ton of PDFs and LO files, I generally keep them pretty well sorted and on a USB drive, so I really don't have a need for an indexer. Mostly I'm searching through SlackBuilds in my local mirror so grep is what works best for me.

OK I think I get it now. Can't believe I didn't see it sooner. I apparently assumed we were all talking about recent installs but it must be that some of you have been running -current for quite awhile, perhaps even as your main which it certainly can do. Baloo takes time proportionate to the number of files in $HOME and my fresh install had very few when I upgraded from KDE4 to Plasma5 so that's why my "crippled" time was so short and sweet while others are hit for many hours. The sheer magnitude of difference between what was for me a few minutes compared to up to a full day for at least one member initially just knocked me senseless... that and the fact that I didn't relate to putting $HOME to such extensive purpose. My tendency on single owner systems is to integrate.

I'm sure this is obvious to those who use -current and Plasma 5 regularly but I will choke down my embarrassment at my oversight to point out that when Slackware 15 arrives, even if you decide to check out how Baloo might benefit you personally, any fresh installs will be EZPZ and since everything from then on is incremental, that too will never be a problem. Of course as we now know for certain it is also possible to disable all of the "nasties" so far mentioned by members here and it will be effortless to get a DE with Plasma 5 that will outperform Xfce in just about every way possible and look the way you prefer while doing it. This is likely good news for Xfce users since surely Xfce devs check out the competition and are aware of just how far KDE's sea change paradigm that began with a horrible start in KDE 4x has finally been beautifully refined. Xfce will need to up their game.

1 members found this post helpful.

I don't know about most people, but I keep /home on a separate partition and don't format it when doing a fresh install. The initial baloo stuff will not be "EZPZ" for people like that.