https://bugs.kde.org/show_bug.cgi?id=477068
--- Comment #13 from tagwer...@innerjoin.org --- (In reply to Stefan Brüns from comment #12) > ... everyone else apparently knows better how Baloo works than me ... I think when troubleshooting, you get a view of problems from the "outside, in" and that comes with a fair amount of guesswork and assumptions. If you know the code, you see things from the "inside, out". That's two different ways of "knowing" - it's a question of getting them to meet in the middle. > ... no guarantee a document in the DB is currently reachable by its path, see > network mounts, see thumbdrives, etc ... That's quite clear. In this instance though I create a test file, that triggers an iNotify, and baloo indexes it - without the parent directory being in the index (with the right DocID anyway). > ... First, inotify works with paths, and second, it works with a sequence of > events ... Yes. I've seen, with inotifywait, some of the holes you can fall into (for example when a move is done with a "link/unlink") and can see that you could get synchronisation issues. I am pretty sure that is not happening here though. For what it is worth, I did the tests with and without content indexing enabled. > ... This BR is about DB inconsistencies caused by code changes, nothing else. The thread has drifted, apologies for that. I think I posted to the wrong thread to start with. -- You are receiving this mail because: You are watching all bug changes.
