hi!
I'm using kde since 4.1 and have to say that strigi never worked for me when I wanted to index my whole home partition (default setting)
I have to say that I have quiet a collection of pdfs (approx 500 books with each about 1000 pages) and obviously strigi can not handle this?

with how much data do you feed strigi so that I works?

There are alternatives to strgi if it will not meet your needs though I have no idea if they will be able to index your collection.

recoll (kde4 & kio interface), tracker, Beagle (requires Mono, should be kde gui) and Google Desktop Search (you might need to look at the Enterprise version as the free used to have words per doc limits).

You might need to consider something outside the realm of "desktop" tools, something more enterprise focused

well actually I don't even want all those books in the search index, at least not with fulltext-search. I like to tag them and search for their filenames nothing more than that, but I do of want fulltext-search for other docs. An option for not taking the whole pdf, but rather only its filename into the index would be handy

First why not set up your search index so that the pdf books are in a directory not scanned by strigi (alleviates the scanning/indexing issue)

and for the book titles (I assume the title is the file name) either:
1) tag the books and use nepomuk
2) use Dolphin and it's filter to search for title
3) use locate (with kio_locate as the interface) for the titles
4) use kfind to search the book directory

Or what about Calibre (an ebook reader and management system) see http://calibre-ebook.com/about & http://calibre-ebook.com/user_manual/fa ... management. There might be other "management systems" out there.