When we wrote "Clustering of time-series subsequences is meaningless", it took six attempts to get it published.

One reviewer wrote (word for word) "you will get in bad trouble if you publish this"