I'd be interested to know more about the rationale behind this. As I see it, the change means: 1) The blocked person doesn't know that they're blocked, so the

I'd be interested to know more about the rationale behind this. As I see it, the change means: 1) The blocked person doesn't know that they're blocked, so they cannot be certain as to whether their tweets are being seen 2) The blocked person can now mention the username of the person who has blocked them (obviously this is required by #1 - if they couldn't mention them, then the existence of the block would be obvious) 3) The person who places the block cannot see any tweets by the blocked person, except in search results or by visiting that person's profile page So, it seems like a trade-off between the old situation in which blocking someone would prevent them from mentioning your username, but would also inform them of the block, and the new situation where the blocked person is not informed. There's really no way of secretly blocking someone unless you allow that person to continue tweeting at you, because if that were prevented then the block would be revealed. I may lack the necessary perspective on this, but to me this looks like a not-unreasonable trade-off to make. If you inform someone that they're blocked, they will know that they need to sign up another account in order to follow you. Informing someone of the existence of a block may also be a trigger for escalating harassment. On the flip side, if someone can mention your username then they can make it appear as if they're engaging you in conversation and you're actively choosing not to reciprocate, where in fact you have blocked them. It's a difficult balance to get right. An alternative long-term solution might be to have different privacy levels per-tweet, or to allow a person to have multiple personas, with one being private and another public, so that you can have some tweets which are for everyone and some which are for family and friends only, which would prevent a third-party harasser from hijacking personal conversations. But any move in that direction dilutes the simplicity and directness which is what differentiates Twitter from, say, email or IRC, and also raises the stakes in situations where people mistakenly tweet something to the wrong persona or privacy level. EDIT: When I said that I may lack the necessary perspective, I was correct. It seems that the main benefit of the old block system was that, if you have a follower who actively dislikes you, you could prevent them from seeing your tweets in real-time in their main account, which is often enough to stop them responding, or retweeting you in order to encourage others to respond. It's not entirely rational, but then abusing people on Twitter isn't rational either, so perhaps this is one of those situations where behaviour of the wetware nodes on the network needs greater consideration.