The reason I dislike this, is that besides obviously racist or otherwise inappropriate comments, there are regularly normal seeming comments that are dead. Now, there may be reasons for this I am just not privy to, but I feel it hinders discussion when some opinions that don't seem problematic are just silenced like this.
[Edit]: And rereading my own comment, I think I want to clarify that I am not talking about free speech issues or political discussions here, but I regularly encounter factual seeming or explanatory comments in technical discussions that are dead and it just baffles me.
Usually when I see a perfectly good, but dead, comment I check out the user's history. I usually see one of two situations here:
1. It's clear that they have a pattern of making nasty, off-topic, often vulgar comments. I'll still vouch for their good comments, but at least I can understand why they were dead to begin with.
2. If their history is of mostly normal comments except one bad joke at the start of their posting history that got them hell banned, then I'll vouch for them and let them know what's happened.
Hacker News is a lot less hospitable in reality than it seems to an outsider. If you just register a lurker account and enable showdead and look around it's a veritable zombietown in here.
I knew there was an option to vouch for comments but just assumed I did not have enough Karma to do so. But another comment in this discussion mentioned the "hidden" flag button, which you only see if you click the comments timestamp.
And sure enough, the vouch button is just there. So the problem may just be that you have to know where to find it. Which I now do. And hopefully everyone who didn't and reads this now as well.
I do appreciate that you can undo most of these actions now. There was a time when downvoting (and upvoting) were instantaneous and permanent. It was very frustrating when scrolling on my phone and accidentally downvoting someone because I used the wrong kind of press in the wrong place.
Moderation is guesswork, and even if we guess 99% accurately about which accounts to ban in the early stage—a necessary thing to do, because otherwise the site would be overrun with spammers and trolls—that still leaves quite a few false positives who later turn out to be neither spammers nor trolls.
That's not true.
New accounts are subject to extra software restrictions because of past abuses by trolls and spammers, though. I don't see any way around this.
I'll guess again! My VPN just happened to be IP banned which I had no way of knowing before registering.
I'm sure new people appreciate being labeled as abusive trolls and spammers just for registering and I’m sure new people don't stay either.
If you or anyone has suggestions for how to better write software to distinguish between genuinely new accounts and serial abusers, I'm sure we'd be very interested.