Hacker News Reader

> What if just opening the link does a destructive action, like unsubscribing or posting a comment on your website?

Then whoever created the link that does that is doing web wrong and needs to stop.

GET is, by definition, a safe method. (See RFC 7231, Sec. 4.2.1; RFC 2616, Sec. 9.1.1.) Doing destructive actions via a safe method is plain wrong.

Being wrong is very different from not existing. As long as a spec is a guideline there are people doing it wrong, I guarantee you that.

There are links that do that and if google were to follow every linkt people would notice very fast.

Google does follow every link, and those people do notice, because their forums get deleted. Very few people do it this wrong for very long.

(Or they get fed up and patch the problem without understanding it, by using robots.txt to block all crawlers from their site.)