... and get a message from the source instead. Right now cooperative indexing relies a bit too much on the idea that the indexer cannot keep up with the source and the doc processor. If it outperforms them we will get into trouble.