So Nepomuk Strigi seems to work there inconspicuously in 13.1 doing its indexing.

How do I take advantage of it? Is there a program which will help me find things using this?

You might wanna read http://aseigo.blogspot.com/2010/05/i...muk-right.html

Actually they don't seem to work. I played with them and found that strigi/nepomuk was not able to find some files even by file name. All this thing is a real crap compared to Recoll. And the resource usage is insane.

Unfortunately IMHO this is the present state of the entire KDE4 project. Half working resource hungry peace of junk. I dumped it for good and I'll never look back.

Good luck in taking advantage of it.

Inconspicuously???? For me, its about as inconspicuous as the elephant in the room, with a room slightly smaller than the elephant.

I've only tried once with kfind, which I guess uses this load of brown stuff as its back-end (is this true, or is it just one of my deranged imaginings?). In finding files by filename, it seems at least an order of magnitude worse than (s)locate.

There is a possibility that it may be able to do something else, too, but what that is or how well it will do is, currently, questionable.

@ivandi
I can see your point, but I don't even think that this -the state of the software, itself- is the biggest problem. The developers have some idea of where they are trying to get to, but they haven't sold this vision to the users, and they haven't provided useful documentation of the wobbly jelly that is KDE these days. I wish I had realised what an extended period of pain this was going to be, and persuaded people to fork kde 3.x, to produce a (very slightly) enhanced version of that instead.

Thanks for the link sahko.

I have abandoned KDE, because they first abandoned me as a user, with KDE 4. So I no longer stay up to date with it and am not interested in what revision finally becomes stable (though I note that many say it is good in 13.1).

However, I have always been curious as to what the heck necromonger (aka nepomuk) is actually useful for. So when I saw your link above, I thought that at long last I might have an answer.

I read the page - twice - and still cannot tell you what it is useful for! Other than ephemeral references to "social-semantic desktop" and some "ability to integrate technologies horizontally" newspeak, I remain clueless!

Mr. Seigo repeats, "The more we use Nepomuk for the things it was truly designed to make possible, the more we'll see the benefits beyond it being a file indexer", but we never get any unambiguous definitions of what those things might be!

HAHA! But the gem in the article, by Mr. Nepomuk himself is this, "For me, Nepouk's ability to index my files is a nice feature. It's also one I currently have turned off due to personal preference."

He doesn't use it himself, so why the heck should I? Maybe he doesn't know what it does either...

And the OP's question remains unanswered...

There is a graphical front end to strigi, strigiclient. It will do searches and give results, but I suspect it will disappoint those looking for a polished front end. It is disconcerting to start typing a search term, see results, then continue typing, results disappear, then a full search term is entered and results reappear. Also some of the buttons do not appear to work correctly.

I think thats whole point of this blog post. "Nepomuk doesnt do anything useful...yet."
I have the feeling this is the reason kdepim wont be shipping with KDE SC 4.5 but rather 4.5.1. So that nepomuk is intergrated more into it.

I also share the sentiment that "i dont need no stickin nepomuk" and would rather it were completely optional, but as long it doesnt get in my way and doesnt hog my system i am willing to bear it.

File indexing is an useful thing. I use it to index my PDFs and even my xchat logs, pidgin logs and pan's cached messages. It makes it very easy to find information in tens or hundreds of pdf files. Think also of indexing all the docs in linux-faq and linux-howtos packges, man and info pages etc.

The problem in KDE4 is not the idea, it is the implementation. Nepomuk and strigi are heavy,slow and full of bugs. The same goes for akonady, great idea but poor implementation. Why should I run a mysql server on a netbook. It's crazy. Plasma is another poorly implemented good idea. I like the look but I don't like a desktop that can be crashed by some buggy widget.

For me KDE4 project has become a toy for a group of self-absorbed programmers. And its place is in /testing. Not in the stable slackware tree.

I don't use Nepomuk or Strigi myself, but if I'm not mistaken, strigi searches can be done through krunner (alt-f2). Nepomuk searches can definitely be done in the top bar in dolphin. There are probably other programs that can also access indexed data.

No. Kfind is little more than a front-end to ... wait for it ... find. It dates back to at least the KDE 3 series (and probably earlier) and doesn't access indexed data.

Exactly. It is so simple to turn off that it is hardly worth complaining about. In fact, complaining about it on LQ requires much more effort than simply disabling strigi and nepomuk.

that (alt-f2) was useful (up to a point); it took about 30 seconds to open, which was a bit slow, the actual search was reasonably quick and of better quality than kfind ... and then the window disappeared and the crash handler popped up:

neat; I don't think I'll be back soon! maybe I'll try Dolphin though, I haven't yet tried that.

In any case, although the kde devs always stated that kde 4.0 was a developer-only release, you might have thought that it would be getting a bit closer to usability for ordinary users, by now. And there ought to be some sensible, readable, documentation that gives you 'expert tips' on configuring and using stuff like this.

@sahko
...but even though it doesn't do anything useful it is, by default, imposed on all users; surely there was an option to make it off by default, until it does do something useful?

And Mr Seigo is pulling another interesting (but stupid) trick; he keeps going for the line that 'everyone is saying that it is only a file indexer, and it is so much more'

Well, firstly, exactly no one (except Mr Seigo) is saying that it is only a file indexer; 'only a file indexer' is totally useless, you need something to work with that index, or you have just wasted your time building an index that you'll never use.

And in saying that it is so much more, what he actually means is: 'We have plans for it to be so much more, but we haven't explained what this will do for the end user, or why they will want it, and anyway it doesn't do it yet and anyway we haven't committed to any particular timescale for it to do something useful...or even when it will be sufficiently bug-free that you could rely on it.'

So the answers to my question are:

strigiclient
krunner
dolphin

Another thing that I do not understand: I installed 13.1 with mysql disabled. I have subsequently enabled it ( because I do use it) but KDE should not know my mysql passwords. So how could this stuff be using mysql?

Alan

I don't have anything indexed yet, but now I'm thinking I'll give it a shot to see how it performs. I can't comment on the crashes, but let me try and I'll get back to you.

I think this is an issue to take up with your packager (ie Slackware). What I do is remove the nepomuk autostart files in /usr/share/autostart. If you believe that this software is not of production quality then you should request that your distro not package these autostart files and have Nepomuk disabled in the KDE settings by default. Were this the case you'd never experience Nepomuk unless you turned them on. Slackware does the reverse.

Well I think most people DO think that Nepomuk is only an indexer. I think it goes without saying that if there is an indexer there is a means for accessing that data. This was not Seigo's point. The point is that Nepomuk is about managing meta data of files in addition to indexing data. Meta data is completely, or at least substantially different, than file indexing. The Nepomuk framework is about tagging and rating files. All the comments on that blog post make it clear that everyone is complaining about the resource usage of the indexing (ie Strigi). You can use Nepomuk without Strigi (ie indexing) if you want. They are two different things. Or at least to an extent: you need Nepomuk enabled to use Strigi for whatever reason.

It is using embedded mysql so it doesn't connect to your msql server.

All of you complaining about Strigi in this thread: how many of you are using Slackware 13.1 and KDE 4.4.3? How many of you are describing your experiences with this version of KDE on 13.1 and who are in fact describing what they experienced on older versions of KDE, and/or using other distros?

Eric

0 members found this post helpful.

I just tried a strigi search through Krunner. Like you, it started very slowly (krunner usually pops up immediately) and it crashed when I clicked on a hit. However, afterwards it works fine. The normal (ie not strigi) krunner matches come up first, then indexed hits show up a second or two later. It seems stable and the indexing of my data did not eat up noticeable resources. I have 5,800 files in my index, which takes up 280mb of space.

I'm not sure if I need deskstop search functionality, but if I find having it useful over the next few days, I'll leave it enabled since it is using few resources. If I decide to disable strigi/nepomuk it'll probably be because I'm not interested in seeing all these text matches when I use krunner (I use krunner to launch applications and do web searches all the time).

To be fair, most of the contributors were simply asking the question "What's it for?" which is relevant whatever version you are running. I understand that the latest version (only one I've tried) is much better for keeping these processes in the background and out of the way of the user experience. Still, the question of what they actually do to help the user is not unreasonable to ask ?

2 members found this post helpful.