To a vanishing degree, most people agree that hosting a site that encourages vulnerable people to kill themselves (and each other) makes the world worse. That's the context here, not a site debating the merits of different dog breeds.
It is a broader consideration. It is easy to decide to stop servicing murderers, nazis, and other widely derided groups, especially on a one-off basis whenever objections are raised. It is another thing to actively investigate your customer base and police them. There are industries where this absolutely should happen. And there are others where we probably are better off with a laissez-faire approach (my local grocery is not a regulated utility, but I think they should still roughly operate that way - serve anyone who comes in the door). Cloudflare needs to have a clear TOS, and remove this site for violating it.
But we should still be cautious about how broad a TOS it might be, what evidence is required, what due-process and appeals are available. Every other day there is some post on HN saying "Big Service provider shut down my site overnight without warning!" but that is consequence anti-abuse systems policing customer activity and badly implemented systems are going to have negative consequences. The obvious cases make it seems easier than it actually is to build fair systems.
We learned this the hard way during the Civil Rights Movement: there needs to be some public participation in the restrictions that private businesses can place on their customer base.
Of course they don't. Look at all the objections to removing a forum explicitly created to encourage stalking and harassment of marginalized people, which has lead to multiple suicides by its targets.
Sometimes you have to move forward without 100% agreement and stop listening to trolls making slippery-slope arguments.
