In many of the recent discussions on the topic on HN, you can easily read about the origins of Kiwi Farms, which was originally "CWCki Forums", created for the express purpose of harassing a webcomic creator, Chris Chandler - the name literally grew out of the initials of their first primary target.

Furthermore, heck, the Wikipedia page on Kiwi Farms has tons of references to primary sources that detail some of the directed harassment, e.g. from one well sourced article at https://web.archive.org/web/20190511184633/http://nymag.com/...

> In the most extreme instance of this, they goaded a 13-year-old boy into masquerading as a 19-year-old girl in order to have phone sex with Chandler (and record it, of course).

This information is so trivially easy to find (just go down the Wikipedia rabbit-hole, or again, read many of the posts on this topic on HN) that I can't believe the requests for "Source?" and false equivalence statements of "lots of companies host objectionable content" are made in good faith.

TBH, I haven't seen anyone really try to defend any of Kiwi Farms' content. The only real question is whether Cloudflare should be in the position to make judgement calls on the level of danger inherent to any of the sites that use their DDoS protection services.

None of what you said makes it different than Twitter.

Let's propose a test. We can scrape all of the tweets submitted in the last month and all of the posts on Kiwi Farms in the last month. Then let's count up the percentage of posts that contain a slur for black people, jewish people, or trans people. If the rate of slurs on KF is no more than 10x that of Twitter then we can consider these sites comparable. Otherwise, we can consider them to be in different categories.

Why should we only count slurs for these groups?