[kdepim-users] nepomuk akonadi pim and all

[Martin (KDE)]

Am 27.11.2011 23:06, schrieb Andras Mantia:
> Hi,
> 
> Martin (KDE) wrote:
>> What I don't get: there are many processes running around with akonadi in
>> its name and it would be great if someone can point me to some
>> documentation what they are for and how I can configure them.
> 
> Usually there is a process for every account, every contact source, every 
> calendar source and in 4.7 three processes feeding Akonadi data to the 
> Nepomuk indexer (contact, calendar, email). In KDE 4.8 there is one nepomuk 
> indexer for all of those (and it is more optimized).
>  There is also the Akonadi server process and a small helper "control" 
> process for the server.
>  If you see an akonadi_agent_launcher, look the whole command line for the 
> real process name. This is like kdeinit4 for the rest of KDE.
>  If you want more specific answer, ask which process you are interested in.
>>
>> kmail uses akonadi and akonadi use nepomuk, so far so good. But what use
>> kmail nepomuk for? 
> 
> Nepomuk is used by KMail (and many other parts of KDE) to do searches in all 
> kind of data. Inside KMail it is used for full text mail search and for 
> addressbook search in the KMail composer.
> 
>> I saw the nepomuk virtuoso-t process running and
>> collecting data (at 200 MB I stopped it, removed the data and restarted
>> it) but what for? How is the data maintained (old stuff removed)? How can
>> the user tune the nepomuk database?
> 
> Sorry, don't know the details of nepomuk. The configuration settings are in 
> System Settings->Desktop search. How well they work, how fine you can tune, 
> can't tell. But there are still some issues with it, and nepomuk sometimes 
> goes crazy and uses 100% CPU (see a recent mail on the kdepim devel list).
> 
>> I saw a configure option in my mail folders called something like "enable
>> full text index". What is it good for and what do I loose if I disable it?
> 
> AFAIK, you will lose full text search for those mails.
> 
>> I have strigi disabled, so nepomuk data most likely are not from the
>> strigi stuff.
> 
> Nepomuk gets data from multiple places (feeders). Strigi gives that from the 
> file system, the akonadi feeders give the data from akonadi data.
> 
> I hope this clears up some things.

Definitely.

I played a little with akonadi console, removed all nepomuk feeders and
added them afterwards. As my mepomuk database grew up I stopped nepomuk,
removed the DB files and restarted it afterwards. The sad thing was that
the autocompletion from the address book stopped working (reported here
from you). removing and re adding brought functionality back.

The E-Mail feeder is something different. After re enabling it I can see
the crawl through my collections (Mail folders) but a search for some
sender brought no result (they are definitely there). The feeder told
that it has finished crawling.

Another issue is the full text search enabled by default. I use several
mailing lists. I almost never search in there (besides for
sender/recipient/subject which I do in the search bar above the
mail-list). switching index off for all these folders is a little
painful. An option to switch these data for child folders as well would
be great.

I use cyrus imap server. This server is able to do server side searches.
With nepomuk I loose this functionality right? Are there plans to use
server side search instead nepomuk in the future?

Regards
Martin
> 
> Andras
_______________________________________________
KDE PIM users mailing list
Subscription management: https://mail.kde.org/mailman/listinfo/kdepim-users