Ask HN: Is flagging being used as a form of censorship? | Hacker News Reader

Ask HN: Is flagging being used as a form of censorship?

18 pointsCieplak27 comments

This post had incredible momentum and was rapidly approaching #1, but was suddenly erased:

https://news.ycombinator.com/item?id=12328142

https://muchweb.me/systemd-nsa-attempt

Does this story violate content policy in any way?

Snowden's documents suggested that significant resources are being spent to manipulate online discourse.

Nevertheless, we must discuss and avert security issues to protect ourselves and our customers, while our government is actively subverting our security.

View on HN

"Censorship" is like "terrorism", a pejorative that's relative to what people like. So if you like a post and it gets flagged, you can use "censorship" to express not liking that it got flagged.

When users flag a story, that's the community immune system rejecting it. It might be hard or annoying to see why users would reject something you posted, but in this case it's fairly easy to explain: it was dramatic without being substantive (and combined several flamewar topics to boot). Such stories mostly don't do well here. That's a good thing.

Please, though, don't add meta-drama posts on top of drama posts. At that point you're just adding noise, and we're trying to prop the signal/noise ratio up here.

On January 2011 I read about Bitcoin as a lurker on HN.

I thought it was quite a novel idea, but needed to prove its worth to the world. I didn't want to take part in it, use it, as I saw no direct benefit to me; and without an underlying economic activity, buying Bitcoins was a purely speculative Ponzi scheme. I came to learn - through HN - about Silk Road and how people were using it to animate their illegal endeavours. Bitcoin had economic sense, so in April I bought 50 euros worth of it, to give it a try.

Price went from 7euros/BTC then to 500euros/BTC now. It has been, and will ever be, my single best return on investment. You just can't beat 70x. HN has literaly made me richer.

Had I invested 1k euros 2 months earlier, my life would have taken an entirely different meaning. It has taught me the importance of being bullish on novel ideas.

HN made me discover the Oculus Rift; HN made me see the unbelievable rise from the kickstarter to being bought 2B$ a year and a half later. I have been trying out a lot of great things thanks to HN.

To me, it is an early warning system about the best the future can offer.

-----------

It seems to me that the bitcoins of nowadays are getting buried in noise, or are no longer here. Frontpage is covered with common law, societal problems, short lived news events, click bait, and my-framework-is-better-than-yours. I am not diminishing the importance of general news, but this is not the venue for it.

Where is the good stuff? Why don't I hear more about initiatives like RISC-V (which is arguably hardware's linux moment)? Why don't I hear more about way to do some CRISPR editing mysaelf as a side project (which is arguably DNA's linux moment)?

HN is pivoting slowly towards mainstream general tasteless news; and my source of living in the future and doing interesting things in it is drying up.

-----------

So, my plea to you, Dang, is: could we have public, explicit, ruthless, white-list-style, displayed-right-next-to-the-post-button definition for flagging content?

My take on it:

* It must be about the future getting materialized.

* It must have a strong technical aspect, on which HN members can readily start to work on.

* No news events of any kind, even nation-wide ones like terrorist attacks.

* No societal articles. Not even about the gender gap, which is greatly affecting us all.

* No baseless day-dreaming about what the future could be. (How many of these can we have about "AI"?)

That's a good story and an illustration of HN's influence at its best. There are many such stories, which is wonderful. Our job as caretakers is to optimize for curiosity, so we want as many of them as possible. People's visions of what HN should be are not all fully compatible, so going far in one direction would be to choose a local optimum over a global. In some ways the global optimum (or our best approximation thereof) satisfies no one, which is a little weird if you think about it.

I'm pretty sure this isn't true though:

> HN is pivoting slowly towards mainstream general tasteless news

That's one thing we're in a position to prevent and that we put a lot of effort into. Perceptions of how HN used to be in the past are pretty unreliable; there seems to be a strong nostalgia bias in most such reports.

I'm sorry. I realize this was overly dramatic and not substantive. I would have flagged it myself as being speculatively political; I was a bit frustrated and exhausted with some kernel debugging at the moment. I'm sorry.

No prob! Thanks for receiving the point.

I'm guessing it was flagged because there was nothing of substance in the article. The author provided no evidence that systemd is malware or that the NSA was involved in its writing. So it's just idle speculation with a click-bait title. If I would have seen the article, I would have flagged it for those reasons.

Also, the "[flagged]" prefix indicates that it was killed by user flags, and several users would have had to flag a 19-point article to kill it. I'd hardly call that "censorship" (unless you think that multiple HN users are acting on behalf of the government).

Finally, the HN Guidelines say that if you have questions about moderation, you should e-mail the moderators rather than posting. So don't be surprised if this post also gets flagged.

> unless you think that multiple HN users are acting on behalf of the government

It would be naive not to think that multiple HN users are acting on behalf of the government, and I am not talking about the US government only; all governments do this on every popular site.

The question is whether there are enough users, or users with a significant impact so that any of this matters.

My personal opinion is that this is the case here, at least when it comes to stories about the NSA, surveillance, Turkey, Russia, etc. There are quite a few prominent members of the community which I personally consider government shills (and no, I won't name them, make up your own opinion).

I won't even say this is necessarily bad thing. It's inevitable, so don't take it as a criticism of the site and/or community (more than someone can criticise that water is wet).

Agree on that. And it kind of becomes too obvious once you start reading HN regularly.

I doubt that. This (apparent) phenomenon is more easily explained by people's difficulties tolerating strongly opposing views. That sounds odd, because we all think we're good at that, but really we're not.

Having strongly opposing views in your face is unpleasant, even a sort of painful, and we react to this by constructing stories about how the opponent isn't just wrong but bad (disingenuous, dishonest, astroturfing, a shill). The alternative is to accept that honest, smart, committed people might hold an opposite view to yours, but because our own views are so strongly held, that feels much less likely than it should.

In other words, claims that HN is rife with shills, etc., seem mostly to be un-self-reflective indications of how diverse the community is.

Flagging is done by the HN community as a kind of self-policing. Some of us are more aware of and sensitive to the guidelines[0] (and more likely to enforce them) than others. So a post can be popular, but also violate guidelines blatantly enough that folks flag it. I think your question gets more interesting if a submission violates guidelines, but the readers who choose to flag it are actually the ones also opposed to the content on its own merits, and use their flagging rights to jettison the post.

I've seen the same situation where posts get several up-votes shortly after posting and land on the front page before disappearing suddenly. It made me curious whether up-votes can override flags in some ratio, or, whether there is a fixed trigger count that drives [flagged], or, if flagging simply gets the attention of moderators who make the call whether to [flagged] a post or not.

[0] https://news.ycombinator.com/newsguidelines.html

Upvotes do not counter flags, users with high enough karma can vouch for flagged posts tho. If enough users vouch a post can be unflagged.

I'm curious where you got that impression. Upvotes do counter flags. You might be confusing flagged posts with dead ones (those are closed to new comments and, I think, upvotes). In such cases a post is explicitly marked [dead].

Just by casual observation but without knowing the flag vs upvote ratio/formula it's hard to tell anything.

We've had posts with 100's of upvotes being flagged to death (usually 'political' or 'polarizing' issues) so it didn't look to me like upvoting does much to keep the post alive.

And since the amount of flags isn't visible to us it's not that we can see that posts get flagged once they reach 5 flags for example or once the flag count reaches 10 over the upvote count etc. and make a better observation from that data.

But thanks for the correction :)

Your description seems accurate to me. Upvotes do counter flags (and vice versa). Once the flags go above a certain threshold, [flagged] appears. That's always something users do, not moderators. Moderators sometimes downweight posts as well, but using a separate mechanism. Flagging does get the attention of moderators, even before the [flagged] threshold is reached. If there are enough flags to also kill a post, you'll see [flagged] [dead]. The software only kills a thread if it isn't very active, though.

Does that answer your questions?

Yes!

But then begs another...

Moderators may 'downweight' a post... I'm curious what types of submissions a moderator might downweight. 'Popular but violates guidelines', or 'Popular but too distasteful/offensive', or 'popular and really funny, but not consistent with HN ethic'.

How much subjectivity is there in downweighting and to what extent might it be perceived as censoring, if any.

Also, I'm a bit put off that OP changed the 'Ask HN:' question after I'd commented. Formatting and spelling is fine, but the question changed significantly from (when I commented):

===

This post was briefly on the front page in the past hour, but was suddenly erased:

https://news.ycombinator.com/item?id=12328142

Does this story violate content policy? Or do the ideas threaten the military-industrial complex?

===

To:

===

This post had incredible momentum and was rapidly approaching #1, but was suddenly erased: https://news.ycombinator.com/item?id=12328142

https://muchweb.me/systemd-nsa-attempt

Does this story violate content policy in any way?

Snowden's documents suggested that significant resources are being spent to manipulate online discourse.

Nevertheless, we must discuss and avert security issues to protect ourselves and our customers, while our government is actively subverting our security.

===

HN courtesy dictates that after the fact modifications be indicated with 'EDIT:' or some such wording.

I apologize, I will ensure my posts are immutable going forward.

PS: I also made the edit very shortly after the post, so I think the eventual consistency issue worsened my mistake.

Too late, this article is flagged to.

Thank goodness for http://hckrnews.com/

I saw it when it was live but only noticed it was flagged because of http://hckrnews.com. It's really the only way to browse at this point. So many flagged or dead stories (some that even have value) I never would have seen otherwise.

If you want to see all the flagged/dead stories, you can also turn on "showdead" in your HN profile. That will make them appear on the "new" page when you're browsing HN. But the vast majority of dead stories are spam.

Most dead stories I see on hckrnews.com are worthwhile.

It looks like hckrnews shows only stories that have been upvoted, even if you select "all" (I couldn't find any stories with less than 3 points). That would exclude most of the junk that you'd see on HN's "new" page (like ads for escort services, etc.).

The disadvantage of that filtering is that you'd never have the opportunity to upvote good stories that nobody else has upvoted yet.

Happens all the time. It only takes a handful of people to flag something early for it to get blown down the rankings and effectively consigned to oblivion. Some of the more interesting and technology-focused stories do appear to rub some cohort the wrong way and disappear rapidly.

You could get tin-foil-hatty, and think that the NSA or CIA has a handful of interns that are paid to watch reddit and hn and influence opinion. But it's probably just as likely that there are enough civilians whose livelihoods depend on inconvenient discussions not occurring.

Whoever wrote this had a point: Any time I do something even slightly nefarious, I reboot into Linux.

Systemd embeds its own web server:

https://www.freedesktop.org/software/systemd/man/systemd-jou...

... I didn't realize that. Can it read email yet, so we can just say that it finally fulfills Zawinski's law?

Flagging is a form of censorship.

OTOH, "Censorship" isn't, inherently, a bad thing.

Reply on news.ycombinator.com