The owner of the website (in this case eschnou) is always able to choose what to display on the web page. Just like you can remove unwanted comments from a Wordpress comment thread, you can choose which "indieweb" comments to display.
Ah, okay. And how does comment federation protect against (SEO) spam? If everything runs automated, naturally there cannot be anything like a captcha or other anti-bot measure in the system.
My personal inclination is to accept any comments from people I know, and notify myself of any from people I don't know. I can then approve them/block them by adding them to my address book.
So far, I let comments automaticaly and delete spammy/offending one reactively. It does not happen often. If it becomes a problem, I'll have to switch to moderating comments, just like you moderate comments on a blog.
It's not "if", it's "when". Since this is inherently an automated comment-propagation system, it seems significantly easier to spam billions of sites with a bot. Captchas suck / only work to some degree, but they work to some degree - what would be implemented in this system?
I can see a proof-of-work system working reasonably well (gets rid of mass-spams, though big sites would still get automated ones), but it seems like there should be something baked into it if this is at all intended to grow beyond 'indie'.
But I can keep the moderating effort relatively low by adding a captcha (e.g. a choose-the-kitten picture quiz) - I do not see any point of inserting a "proof of work/humanity" here.
Comments don't need to be displayed automatically. For example, comments from domains that you have previously conversed with may be displayed automatically, but the author may choose to queue up other incoming comments before displaying.
