Live data from Hacker News

Stories removed from the Hacker News Front Page, updated in real time

github.com

381–390 of 532 posts

Re: Stories removed from the Hacker News Front Page, updated in real time

#381
post #310

Earlier quoted context omitted.

Why don’t you make the system transparent? This will save you a lot of effort answering questions.

"Transparent" means different things to people, but if you mean a full moderation log: I think most likely it would produce more questions and effort, for no clear gain. I've written about this over the years: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu... Here's how I look at it: if trust is present, then we don't need to publish a full log, as long as we answer questions when people ask them. That…

While you argue against transparency, keep in mind that you are doing it in full public view.

Re: Stories removed from the Hacker News Front Page, updated in real time

#382

Earlier quoted context omitted.

At a quick glance, I found several that don't match that criteria you mention, here are a few: Open Source Doesn't Require Providing Builds https://codeengineered.com/blog/2024/open-source-not-builds/ Sam Altman Says AI Using Too Much Energy Will Require Breakthrough Energy Source https://futurism.com/sam-altman-energy-breakthrough Avoid Async Rust at All Cost https://blog.hugpoint.tech/avoid_async_rust.html (Perhaps…

I don't agree with the GP at all. Most seem normal for the front page or the intellectual curiosity standard (I mean, personally I'd like a much higher standard, but I'm basing it on what HN already has). All from only one day: * Ford's new 48-inch digital dashboard is a lot of Android for one car: https://www.theverge.com/2024/1/22/24045932/ford-android-scr... * Secret Plan Against Germany (a very big story in Germa…

Ok, I'm finally getting to this - sorry for taking so long! First let's find the actual HN submissions... here they are:

Ford's new 48-inch digital dashboard is a lot of Android for one car - https://news.ycombinator.com/item?id=39089599 - Jan 2024 (78 comments)

Secret Plan Against Germany - https://news.ycombinator.com/item?id=39092116 - Jan 2024 (5 comments)

Show HN: Vx.dev – GitHub-Powered AI for effortless development - https://news.ycombinator.com/item?id=39091819 - Jan 2024 (34 comments)

Open Source Doesn't Require Providing Builds - https://news.ycombinator.com/item?id=39094387 - Jan 2024 (68 comments)

Of those 4, the Ford one and the open-source builds one set off the flamewar detector (a.k.a. the overheated discussion detector); the Germany one was flagged; and the Show HN one got moderated down. Let's look at what happened in that order.

The Ford one setting off the flamewar/overheated detector is easy to understand: hatred for modern car UIs is one of the most popular topics on HN these days and always gets people going (me too! but never mind)...which no doubt is one reason why the media sites keep playing it up. We wouldn't turn the penalty off in such a case. The discussion might not have been a flamewar but it was nearly entirely generic - for example the top comment: https://news.ycombinator.com/item?id=39090622. Given how over-discussed this topic already is, I'd say this is a case of HN's software working as intended.

The open-source one setting off the flamewar/overheated detector is also unsurprising as open source stuff is also highly discussed and also gets people going. In this case I could make a case for turning off the penalty, but in the end would probably decide against it, because the article isn't very deep, the discussion is rather generic, HN has a surplus of such discussion already, and nothing here really clears the quality bar. But it's more of a borderline call; I can see how others could interpret it differently.

"Secret plan against Germany" was flagged by users. That's a political story with a baity title, so the default would be for it to get flagged. We sometimes turn flags off on such stories but I don't know that this one clears the bar. It's more current-events than deeply-interesting, the ideological material is inflammatory and nobody is going to approach it with curiosity. The thread was already showing clear signs of turning into a flamewar. Even then, we might still turn off the flags, but only if the story were intrinsically of great significance—the sort of thing it just wouldn't make sense not to discuss at all. I doubt this clears that bar.

The Show HN, we moderated down because "star for free trial" is not a valid thing to do in a Show HN and is something the community here would strongly oppose (see the top comment). Here's what I emailed the submitter: "We downweighted the post after getting complaints from users. 'Star repo for free trial' is way too much of a hard sales tactic for HN, and even more so for Show HN, which implies that users can try out the product (see https://news.ycombinator.com/showhn.html). Asking them to star your repo first may be an ok tactic in other communities, but in the HN context it comes across as manipulative and is not in your interest at all."

---

I guess the summary here is that this list is a mix of clear calls and borderline calls, but defensible ones. Anyone is free to disagree of course! No two readers, including mods, would ever make all the same calls. But if you do disagree, please keep two things in mind:

(1) You have to take each decision in larger context. A perfectly good story can be a bad fit for HN's front page if, for example, the story has already had a lot of discussion; and

(2) If we moderated cases like the above ones differently, the consequence would be letting a lot more stories onto the front page that are more repetitive and/or sensational than the median front page story is today. I doubt that most readers would want that. You can't think of this in terms of isolated submissions or topics; there would need to be some principle by which the decisions would be made differently. HN's mandate is intellectual curiosity. If there's a way to serve that better, I'd certainly like to know what it is; but given the mandate, that's the only kind of change it would make sense to implement.

Re: Stories removed from the Hacker News Front Page, updated in real time

#383
post #366

Earlier quoted context omitted.

literally quoting from that meta-analysis, which does not include many clinical trials that have demonstrated an impact: "Key messages We are uncertain whether wearing masks or N95/P2 respirators helps to slow the spread of respiratory viruses based on the studies we assessed." Example very large study published in a reputable journal: https://www.science.org/doi/10.1126/science.abi9069?cookieSe...

>literally quoting from that meta-analysis, which does not include many clinical trials that have demonstrated an impact: Yes. To their credit, they only looked at randomized controlled trials. >"Key messages We are uncertain whether wearing masks or N95/P2 respirators helps to slow the spread of respiratory viruses based on the studies we assessed." In other words: the RCTs don't show an effect to a significant degr…

no, it literally says that they do not have a conclusion. You are trying to read it differently because you have an agenda. Please do science instead.

Re: Stories removed from the Hacker News Front Page, updated in real time

#384
post #379

Earlier quoted context omitted.

I'm surprised by this thread too. I never flag anything because it's recorded on my profile, and I don't want stuff recorded on my profile that isn't useful to me. I only upvote submissions and comments that I intend to refer back to in the future. Upvotes are simply bookmarks to me, so my only tool for voting on the quality of conversation is downvotes. Which, apparently, I can't do for articles without spamming up…

Do you happen to remember if the misclicks were on mobile or not? I'm planning to add a confirmation screen to cut down on flags-by-misclick, but the current intention is to make it mobile-only.

Almost certainly they were on mobile, where the small font and use of touch screen to scroll makes mis-clicks much more likely. I'd be happy with a confirmation screen because I rarely or never intend to flag, but by the sounds of it there are a lot of liberal flaggers who could find the extra step annoying.

Re: Stories removed from the Hacker News Front Page, updated in real time

#385
post #374

Earlier quoted context omitted.

Comments like this really make me feel viscerally what we're missing out on by not having COVID fights on the front page more often. Thanks.

This isn't actually COVID specific. It's a nasty and frequent tactic on this forum, where someone makes strong assertions about one side of an argument whilst simultaneously claiming that the other side can't be allowed to speak because it would be "fighting", a "flamewar", a "trash fire", "not curious", "tedious" or whatever. It's an attempt to manipulate the site rules to suppress debate and is itself anti-curious.…

Hm. I think what I'm going to do instead is relentlessly flag them.

Check this out. It's barely on the front page, and has just 3 comments right now. How great is this post? How much more would I rather be reading comments on this than about Bari Weiss? Infinity times more:

https://news.ycombinator.com/item?id=39235484

My son is a biochemist (interviewing for grad school slots right now, as in this actual evening, I'm living vicariously through him, wish him luck). I've been for years paying attention to bio/chem/biotech experts on HN, because I'm a biochem dad. We have lots of expertise about COVID here. None of it is on these COVID threads because all of them would apparently rather eat a bug than "truth it out" with people paraphrasing Bari Weiss. The verdict is in. You're on the wrong side of it!

But these have been useful data points for me, and I appreciate you offering them up. Have a great weekend!

Re: Stories removed from the Hacker News Front Page, updated in real time

#386

Earlier quoted context omitted.

This is accurate, per dang's comment on the Gary Tan thread the other day: > We didn't flag the post; users did. When it comes to submissions, that's nearly always the case - see https://news.ycombinator.com/newsfaq.html . https://news.ycombinator.com/item?id=39169622

There are stories on this list that deserved to be seen, were popular, were important, and were not in fact dumpster fires in the comments - but a particular crowd with a particular bias decided to flag them. Example 1: https://news.social-protocols.org/stats?id=39142094 Example 2: https://news.social-protocols.org/stats?id=39130652 Example 3: https://news.social-protocols.org/stats?id=39214844 Does this crowd think…

I don’t think these things, or even a lot or the other political topics are uninteresting. I’ll often still flag them, however, since I’m really very uninterested in what the HN crowd who responds to these sort of things have to say about it.

Part of this is because I’m European, and the whole “red vs blue” team sort or politics a lot of Americans seem to do these days is just silly, and often hateful. But part of it is also that we’re a bunch of people who know tech and business, but not international politics. I guess I could just ignore them, but I’d frankly rather they were kept to other places on the internet.

Re: Stories removed from the Hacker News Front Page, updated in real time

#387
post #345
post #115

Earlier quoted context omitted.

We haven't touched those stories except reduce the penalties on them (user flags mostly) and moderate them less than we normally would (per https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu... ). I put one back on the front page last night despite this contradicting every principle HN stands for—every other principle, that is, than the first one, which is that we moderate HN less, not more, when YC or a YC…

@dang Thank you for the info. One questions I do have–I would guess posts critical of HN/YC are going to get a log of flags and have not the best discussion. This has a side affect of biasing the home page to not have posts critical of HN/YC. Do you see this as a problem?

The home page already has a massive bias in favor for pretty much any kind of negativity, but doubly so for anything involving a tech company. It doesn't matter whether the stories are untrue, unverified, repetitive, etc, they'll still get voted up, and the comment threads will half full of low quality complaints repeated from past discussion, often only tangentially related to the submission.

And it creates a very visible feedback loop, as users start to think that this is what HN is supposed to be. They're probably the biggest quality problem of HN.

Re: Stories removed from the Hacker News Front Page, updated in real time

#388
post #364

Earlier quoted context omitted.

There's a lot of expertise about COVID here! The problem is, in a variational autoencoder discussion, that's mostly all there is, and in COVID threads there is lots of energy from non-COVID experts. This isn't complicated. You can just look at any COVID thread and see what a shitshow it is. That's not for lack of COVID expertise, though most of that expertise is probably Homer-melding-backwards-into-the-hedges when t…

>This isn't complicated. You can just look at any COVID thread and see what a shitshow it is. I hardly see any covid threads here. I happened to see the one of this week. It got 8 comments before being flagged into oblivion. >That's not for lack of COVID expertise, though most of that expertise is probably Homer-melding-backwards-into-the-hedges when they see the thread. You cannot have it both ways. Either you flag…

Sometimes the signal to noise ratio is so bad that I can't blame the experts for no longer engaging and/or me disengaging before I encounter an expert steering the discussion towards more fertile grounds

Re: Stories removed from the Hacker News Front Page, updated in real time

#389
post #357

Earlier quoted context omitted.

The flags on the last item don't seem to be made in good faith. This looks like abuse of the flag system to me. Is there a system for monitoring flag abuse?

I assume you mean this one: Avoid Async Rust at All Cost - https://news.ycombinator.com/item?id=39102078 - Jan 2024 (62 comments) I can make an argument either way there. The argument in favor of flagging it could be: Rust is one of the most-discussed topics on HN; Async Rust in particular has had a ton of discussion [1], including a major thread just a few days earlier [2] - therefore this post was very much in the…

dang, thanks for taking the time for such a thoughtful response. I didn't know about the policy regarding topics that have been on a lot lately, that makes sense. I've not been around as much lately and hadn't noticed that this topic was well-trodden.

Re: Stories removed from the Hacker News Front Page, updated in real time

#390
I wish we had meta-moderation and voting and flagging reasons like on Slashdot. Flagging as a super-downvote seems user-hostile and passive-aggressive for a site like HN, but I’ve seen it tacitly acknowledged as allowed as okay on HN by dang, which seems counter to the spirit of the guidelines.
Post reply on HN