@th3j35t3r writes on his blog:
Simply put. If Jim is blocked by John, Jim can no longer even utter Johns handle/twittername in a tweet. If he attempts to the tweet simply doesn’t process or gets sinkholed. Period. The end. Forever, or until John unblocks him. This approach would not infringe on Jim’s ‘freedom of speech’, he can still say whatever he likes, but he can’t include John. This approach would be self-policing essentially allowing users to decide if they are being abused or harassed and allowing them to take immediate actions without relying on Twitter to minimize the problem effectively. This approach would not be an overhead on Twitters current infrastructure and would require NOTHING by way of extra storage capacity.
Trolls are the ugly side of Twitter, but @th3j35t3r's proposal seems very elegant to me.
So how about it Twitter?
Find out more, and check out his amusing flowchart, by reading @th3j35t3r's blog post.