Alexandre Deckner <alex@xxxxxxxxxxxx> wrote: > but the problem is that you're limited by the 4096 watch_node's limit. As mentioned by Jérôme, Tracker already needs way more than those. > So there's currently no way to catch such cases in a robust way. What > can we do about that? I think the best solution would be to introduce a flag that tells a live query that you want to monitor all query results automatically. That's the only way of preventing the race condition between sending and receiving the query changes. The code that I removed in BFS solved the problem as well (always sending query out/in events in this case), but was not without problems either (it would not know the original directory, so was not able to send a correct B_ENTRY_REMOVED messages). Bye, Axel.