John Nimmo, Internet Troll, Jailed Again

The Guardian report misses out a lot of information, which is available in the linked to sentencing remarks.

Here's one example:

Guardian: "Throughout, Bonehill-Paine, of Yeovil, Somerset, was on bail awaiting sentence for making claims on Twitter that several people were paedophiles."

In fact, it's much worse:

"You are currently serving a sentence of 40 months imprisonment, imposed on 17th December 2015 for stirring up racial hatred against the Jewish community in Golders Green. "

and

[On] "28th October, you appeared in court at Yeovil, your home town, and pleaded guilty to six offences of sending malicious communications and an offence of harassment, all arising from false and hateful allegations against innocent members of the public that they were paedophiles. Sentence was adjourned, and your bail was renewed. "

The Guardian says she got 2500 tweets. The sentencing remarks say that at its worst she got 2500 tweets in just 3 days, meaning she got more than 2500 tweets.

Filtering has worked pretty well with email for spam. I don't see anything fundamentally different about harassment that would make it less responsive to filtering than spam, nor anything fundamentally different about an incoming social media stream than an email inbox.

Spam filtering works because the spam email has characteristics such as word frequencies that differ significantly from your non-spam email. E.g., none of my friends, acquaintances, co-workers, or companies that I correspond with ever have occasion to discuss my penis size, and so it was easy for my spam filters to learn that mail about my penis is spam.

I would expect that most people who get harassed on their social media feeds do not normally receive messages from friends or other wanted senders describing how they are going to rape them or kill their pets or children or things like that, so why can't a filter learn to block those?

This is not censorship, is responsability. There is not a constitutional right to stealth and harass women and ruin their lives, or to emit sinister death menaces publicy while you hide under a mask. Try to cry loud in a street that you will cut the throat of somebody in the fair name of free speech and you'll end exactly in the same place.

If you let someone to borrow your living-room for making a bomb, you are a cooperator. If you let someone to use your address to send letters with death menaces or to spread a racist message, you are a necessary cooperator. Is as simple like that. Twitter was repeatedly informed by the victims and it seems that they did nothing for months. Twitter is guilty here also.

Filtering all messages directed to one single twitter account after the harassment started (and the victim asked for it) should be technically trivial. They can implement a "panic button" in a day.

It seems that everybody can implement or had implemented yet such simple filter. Is twitter dumber than other companies in this sense? I seriously doubt it.