I've noticed a shadowbanned behavior. If you submit a link that is a banned domain, you will notice that the 'discuss' link is missing and the post doesn't show up at all (but you see it).
Turn on show dead in your profile and visit the "new" link (https://news.ycombinator.com/newest). You'll see a lot of [dead] links, those are the sort GP is referencing. [flagged][dead] are killed by user action (flags).
There's lot of articles from this site on HN, including recently. But:
* It seems low-quality, at times almost like the Weekly World News. On HN, we sometimes replace even high-quality science writing with primary sources (when the paper is recent and the writing is just a summary).
* It's high-volume, which is problematic in that it incentivizes users to submit lots and lots of stuff from this one site; front page slots are the scarcest resource on HN and everything competes for it, so high-volume low-mid-quality sites tend to get downweighted.
* It's slathered with ads and bait-y gimmicks.
If there's something on this site that you think would make a good thread for HN, you can probably just find whatever it's blogspamming and submit that instead.
The list at https://news.ycombinator.com/topcolors is definitely not "the complete set of colors users have set" as claimed. I know this because it doesn't include my top color (namely #64BFBD).
(For the curious, this is because HN only loads profiles into memory whenever users post. Then they stay in memory until a server reboot, which seems to happen every day or two.)
Everything I hear about HN makes me feel like it's a hobby side project hosted on an old laptop in someones basement, and not literally the PR arm of a giant venture capital company.
Looks like it was updated to include that. As someone who is color blind I never noticed that the submission numbers have alternating red and green colors! That was new to me!
(dir profdir*) essentially returns an `ls` of the directory in which user profiles are stored. (Yes, they are (or were) stored to disk, not a database.) This was one reason it was hard to add a feature to rename users, which Dan somehow figured out.
So it takes the `ls` of that directory, giving a list in mostly random order, and loads them one by one, putting them into a hash table, whose iteration order is also random.
In modern times, profs* is populated in a lazy fashion whenever users post after a server reboot.
It does have a "try-order?" option, which gives a deterministic ordering if set. But whenever I set it, I sometimes get strange errors, at least on MacOS. Since it's disabled by default, I doubt they set it. Even if they did, it would be ordered alphabetically, which means your topcolor wouldn't be anywhere close to the top unless someone else whose name starts with "a" also set it.
I’m grumpy because I am 100% sure that the “automatic downweighing” is applied to me (my comments are often below faded despite multiple upvotes), but it’s never been explained.
This actually happened to me when I first signed up and after emailing dang asking why, it went away. I don’t think I ever got any actual explanation but I think it must have been some false positive which, once corrected, would never recur.
I once submitted this link to the front page, which promptly took down HN because it had no caching.
I fired off an email to Scott, frantically telling him "I think I took down HN. If you’re scrambling to figure out what’s wrong, it’s because so-and-so story is on the front page, and you can fix it by booting it off." A minute or two later the site came back to life, so I emailed saying "oh, never mind. I guess it was something else." And to my surprise he emailed back saying you were exactly right, thanks.
Presumably they either added caching, prevented it from going to the front page, froze the list for posterity, or it’s still a ticking time bomb that can take down HN again if you happen to submit it. Either way, maybe the explanation of your missing topcolor is there somewhere.
Well, given the name I assume it's a ranking? Though some colors are weird and it would be hard to believe a lot of people use it (like #fafafd or #0082a0 )
I'm not sure if this is undocumented, but I've noticed that submitted links sometimes have either the pathname or search parameters portion of their URL truncated or stripped entirely. Sometimes the truncation happens after a slight delay.
For example, if you try to submit a link to a particular Google Books page, say the one I just tried:
the link will remain intact for a short while before losing its search parameters and turning into a link to the book itself rather than to a page in it.
Possibly canonical links. HN looks at the submissions and updates the links to the canonical link if present. I'm on mobile so not able to double check the Google books page at the moment, but if you view source check to see if there's a canonical link in there. This also bites some blogging sites that end up with a canonical link to their base http://blog.example.com on all their posts.
Jtsummers's answer is correct, but we turn off canonicalization for some classes of URLs and I think you've found a good example of that. I've exempted Google Books links for the future and put your submitted URL back in https://news.ycombinator.com/item?id=38774261.
Minor nit: I've posted about this before, but I wish folks would not refer to dang as a "moderator" or say stuff about "the HN moderators". If anything, I think it's more appropriate to refer to dang as the site admin.
The reason this is a pet peeve of mine is I often see people complaining about "the HN moderators" (e.g. "the HN mods are on a power trip and keep flagging my submissions!" or some such), where they are clearly taking this conceptual model from Reddit, and they seem to think there is a shadowy cabal of "mods" who control what gets downvoted/flagged (and, in fairness, this does and can happen in subreddits). In other words, when you see your submissions being flagged or comments being downvoted or flagged, 99.9% of the time it's not "the HN mods" who are downvoting you. It's just other, normal HN users (who have earned 501+ karma as TFA explains) who simply don't like what you have to say.
Not saying dang never blocks users/comments, but it's exceedingly rare and I've only seen it for the most obnoxiously egregious behavior after (usually multiple) warnings.
Speculation/intuition, but I figure it happens in ways specifically designed to evade the type of detection you're basing your claim on, akin to shadow banning, whereas on Reddit there's often more visible evidence like corpses of deleted comments.
You can see the [dead] comments if you go to your profile and enable "showdead". Most of them are very bad, but from time to time you can find a false positive and vouch it.
Yes, I keep that enabled. I guess I provided poor examples but what I'm saying is that it seems like on Reddit moderation activities are quite obvious (i.e., a giant chain of deleted comments tends to indicate moderation and elicit resentment) and here they're less obvious (there's an air of strong moderation but little to point at directly, which is great IMHO as it evades resentment).
dang, or someone at HN, can and does silently modify the "weight" of a submission. That is why some articles stay on the front page longer, and some slip off the front page in a couple of hours.
This is far more control than reddit mods have. Reddit mods have one ability, an that is the ability to remove or not remove a submission, which is a very public action. HN can put its finger on the scales in a way that is much more subtle.
Very much so and it all happens with zero transparency, explanation or acknowledgement, very sus.
Your account can also be effectively shadowbanned with everything you submit going straight to the “dead” pile. Again no explanation and the only way you realize it is cause your submissions slowly die with no comments (and when you view the site in a non logged in fashion your submissions are not there)
I've seen at least one user recently which has all his comments as [dead] automatically and they are pretty normal comments. I did not write down the name but I've seen that only once.
We do know to a reasonable degree because otherwise the ‘improperly’ banned would make a noticeable fuss. It’s not a common enough occurrence but maybe you have some in mind.
Assuming no violent (as in promoting) or particularly racist/sexist comments in their history, this hits some new accounts if they're using a VPN (some particular VPNs, I guess?). Sometimes they get caught up in the same filters used to block spammers/abusers from using HN. Or if their first (or one of their first) comments was a link without substantial text, that also gets people put into some kind of "spammer" category. You, or they if you tell them, can reach out through the email in the contact link at the bottom to get the ban addressed.
You've found an odd edge-case. baybal2 is banned intentionally, despite the fact that most of his posts are good, and some are great. Unfortunately, the remainder aren't. The compromise is that he's banned, but users like me vouch for many of his good posts to make them visible. Here's a post from Dan about him a couple years ago: https://news.ycombinator.com/item?id=29692791. And another from 4 years ago: https://news.ycombinator.com/item?id=21195898.
This could be because they got caught in a spam filter, or because we banned them and the comments they've posted since then are better. In such cases we're always happy to take a look and unban the account if they're using HN as intended.
There are some weird edge cases, like sometimes a banned user, once they're unbanned, will revert to posting abusively. Then we ban them again and they revert to posting good comments again. Who can fathom the psyche of homo internetus.
It's also common for an account to post both fine comments and abusive comments. It's the latter that determine whether we have to ban the account—that's how rule enforcement works. But if a banned account posts something good, there's no reason why the good post needs to remain [dead]. This is what vouching is for: https://news.ycombinator.com/newsfaq.html#cvouch.
Thanks, I didn't know about vouching. I guess there's indeed an history behind each ban of a real user. (Not talking about spam of course, it's a different thing)
Sometimes the history is “your account was pretty new so when you posted something a lot of people didn’t like we assumed you were a troll, banned, and never looked again”.
> Not saying dang never blocks users/comments, but it's exceedingly rare and I've only seen it for the most obnoxiously egregious behavior after (usually multiple) warnings.
Issuing warnings or reminding people of the guidelines (rules) is also part of being a moderator, so the moniker is perfectly cromulent to describe dang's role
I had never heard the word "cromulent" and so that led me to a Mirriam Webster article describing how The Simpsons made the word up for a joke and then from there "seeped into our lexical consciousness". Worth a read!
My issue isn't so much with the word, but it's that so often I see tons of users who are clearly bringing preconceived notions about how the moderator role works at Reddit and thinking it works the same here. At the very least there definitely aren't moderators, plural, on HN currently by that definition.
The other reason I think "admin" is a much better word is that it much more clearly indicates that this is a special, privileged role. That vast majority of "moderation" on HN is done by normal users with upvotes, downvotes and flagging (and also, of course, by automated systems e.g. with spam filters/bot detection).
But again, my primary point is that I see users blaming "the mods" for why their post is downvoted or hidden, when usually it's that people just don't want to acknowledge that the community, at large, didn't think they were contributing productively to the discourse of the site.
There are, I believe, multiple HN moderators; Dan is just the only vocal one. And among all the different sites on the Internet, there are multiple different moderation styles, of which Dan's is just one.
I don't believe that is correct anymore, and I think TFA is right:
> Hacker News currently has one full time moderator: Dan Gackle (dang), and formerly Scott Bell (sctb). Their comment replies provide a pseudo-log of Hacker News moderation.
I've never seen an anyone else besides dang comment that they are an HN admin in the past 5-ish years or so.
The second chance pool blew my mind the first time I got an email to resubmit. You figure if a post gets no traction, that’s just how it is. But I’ll never forget getting that email.
My post https://news.ycombinator.com/item?id=37956065 got no traction when I posted it and quickly disappeared from "/new" without many people noticing it. The next day, I received an email that this post was entered into the second chance pool. Then sometime later, it got picked for the front page and led to a good amount of community participation! Thanks to the second chance pool, I received several interesting demo submissions to my hobby project!
... which, anecdotally, I learned about JUST after getting a comment on there. Imagine my surprise at learning about this (in a manner totally unrelated to the post I wrote that made it to the list), and seeing my own writing as the most recent item on the list.
Not at present. The links aragonite mentioned all exist, but are determined by upvotes. The /highlights page (for comments) is a manually curated list. But if anyone sees a comment that's good enough to be on /highlights, please let us know at hn@ycombinator.com. We don't want to be the only ones finding these things!
> this can drive self-censorship. Which is a growing problem at large.
I suppose it depends on one's perspective. We used to call self-censorship "decorum."
The nature of human beings didn't change when we went online, and sometimes it behooves a person to just read the room before speaking. Because of the karma system, the upvote downvote system, and the user curated flagging system, I would classify Hacker News as a "decorum-friendly" message board where a certain amount of pre-filtering is not only necessary but highly encouraged.
Actually, this is the motivation __to__ nuke. It's what I mean about the environment changing. HN is one of the big places that get scraped and everyone is scraping all data that they can these days. It's very reasonable to believe you can be de-anonymized much more easily and that it's only going to get easier. It's not unreasonable to think that someone can mimic my speech and this creates a new vulnerability. I am thinking much more about self-censorship now and being more measured now than I have in the past.
I do agree with you that self-censorship is a growing problem. But I'm not sure how to solve this when we're entering a world where it is the language you use that becomes a fingerprint. I appreciate the records because there is a lot of valuable information here but at the same time these records make us vulnerable in a way we haven't been before. And that it is in a way that not just worrying about nation states being able to do this but that we're posting these records to huggingface. The processing is just getting easier and that's about all that's needed now.
It is a hard problem to solve. Switching usernames doesn't fix the issue but could be more noise. Even disappearing messages is only noise, but stronger than the former. Probably pretty helpful because I don't think people are scraping every day but that could change.
Idk, if you have thoughts I'd love to head. I've been thinking of writing them down and posting to HN but unsure and it feels like one of those things that sounds conspiratorial until years later people will say "of course this was coming" lol
I do think people are scraping the site daily, if not hourly, and rate limiting is usually trivial to overcome. But I suspect the implementation is not standard.
Scraping doesn’t even have to be hourly, scraping a thread or account once then `diff`ing the two between weeks or months or years would easily reveal any interventions.
Which would actually expose users who you should retroactively investigate. Canaries.
I presume you are aware of “fuzzed” users. With unassociated comments, submissions, and possibly even changing names. I only recently discovered this myself after a little digging, but you may be able to do this and continue posting.
I’ve been hesitant to mention this anywhere, as discussing how you would red team a site you like publicly might not be the best for the site.
There are other clever possibilities, such as using some combination of artificial & organic content to wash users identities. Like banned or shadowbanned accounts mixed with real users. An acid bath if you will. If you dissolved 100 people into each other, who can really say who is who?
Yes, I agree that disappearing messages is not a strong action.
I don't think changing usernames is a strong action either. This is because the fingerprint is your language, not the name you're associated with. See the Enron dataset and the project in Ng's intro AI course, but we'd need to scale that quite a bit (which is very hard).
I'm not aware of the fuzzed users. But I think you should mention it in the open if they are problematic.
I have thought of AI to rewrite my language but that's difficult and only fuzzing too.
I do put a lot of value on privacy, but I actually don't really find that problematic. I don't have the expectation to be able to just erase everything I wrote online, I assume it is public and persistent. So for things where I want to preserve private aspects I have to take that into account from the start.
There are certainly individual cases where some post might turn out to be problematic later, but that's more of an exception and can be handled by manual intervention.
I do find it annoying in some places how users can just erase their posts without a reason. Because they don't just erase their own content, they take the responses to it with them or at least make them less understandable. And those people responding did put some effort into it.
Seems to be missing comment vote manipulation. Users can see the number of votes their own comments get but can't see the number of votes other comments get, or even if their vote changes the vote count on a comment, which allows shadowbanning of unwanted comments by vote count alteration by mods/admins.
That only covers downvoting comments, it doesn't mention why comment vote counts aren't publicly visible. Keeping comment vote counts invisible makes sense if upvotes on some comments are not being recorded as a form of shadowbanning those comments/users.
Other people's vote counts are invisible because they led to constant bickering about the merits of any given comment's "spot" vote count (this is also the reason there's a guideline that asks you not to write these kinds of comments, but changing the affordances so you can't see them at all was a much more effective solution to the problem).
The conspiracist rationale here doesn't make much sense, because the site operators also control the displayed comment count and the registration of votes.
A lot of these backend features/behaviors are super interesting to read through, especially in regards to the "flame war detector" section. I've only ever seen link aggregator websites such as Reddit relying on community moderation & spam detectors rather than anything to improve the quality of the discourse.
I'm curious about the full list of automatic headline changes. They seem reasonable, but it's worth knowing that if they change a headline for the worse, you can immediately edit the submission and restore the original headline (or an "in the spirit of" edit that fits within HN's character limit).
Very cool, thanks. How does one become aware of this class of magic? It seems completely undiscoverable, not being in the FAQ, nor in the present "List of Hacker News's undocumented features..."
Thanks. For those wondering how someone may have missed this: Evidently I had not in fact accidentally done a "hide", so nothing was in my hidden list, and thus my profile page cleverly didn't have a "hidden" link at all.
Things expire off the hidden list over time as well, presumably based on when the link itself disappears, but such links will then show up on hn's historical view.
Yeah it's very easy to accidentally hide (or flag!) posts on my phone. Wish there was an option for a confirm screen. Every once in a while I check what I've accidentally flagged and unflag it. I'm sure I'm not the only one...
1) I enjoy figuring out how HN works over time by noticing certain patterns. Still haven't figured out (please don't spoil) why I can't downvote certain comments in threads where I see both arrows on other comments. HN feels like games in the past - you're given some docs but over time you'd go "ooooh" about something undocumented. Good stuff.
2) voting system works well for just users but HN is infested with corporate voting manipulation. It seems that an entity monitoring eg. brand's name is able to use just a few 501+ karma accounts to bury, or a few more to strongly boost, any post
3) based on what I learned about posts' scores (more than a grain of salt - as I mentioned - I'm still figuring things out) it seems a little out of place that sometimes a few hours-old posts (not as old to qualify for the "second chance" thing) with barely any comments and points are positioned higher on the top page than younger posts with more activity (higher score and more comments). I've noticed it more than once with substack and my assumption is that HN may be boosting YC-backed products
It may not even be official corporate manipulation. HN is frequented by a lot of industry insiders, and if they see something that doesn't resonate with what they know to be true they'll downvote.
Exactly: there's a ton of people here who work at the big ad-tech companies, so of course they're going to downvote anything that promotes ad-blocking since that harms their own paycheck in a way.
Imagine a discussion forum for medical topics (science, research, etc.) where 1/3 the users work for companies that sell homeopathic "remedies", and how that would influence moderation there.
I don't think Dang is advancing YC interests - I think it's just bog-standard astroturfing. Especially with the low vote count of most stories, it's not hard to make your story pop or disappear.
I did warn him that people might complain about side effects
Edit: believe it or not, we got more complaints about this. Some people picked their top color specifically to go nicely with the orange Y! So I've reverted this change, except for the Christmas case. Users who have set their topcolor don't see the Christmas top bar anyhow, so don't have to worry.
is it possible that NetNewsWire (possibly other rss readers as well) don't show the favicon for https://news.ycombinator.com/rss anymore because of this change? I noticed this a few days ago
HN is still serving the orange Y for favicons. Only the Y to the left of "Hacker News" in the top bar now has a transparent background.
However, there was a brief period where we were serving the transparent Y for favicons before I noticed that and corrected it; I wonder if the transparent Y somehow got into a cache? Is there a way to force-refresh?
Unfortunately the issue still persists and it doesn't seem to be related to caching.
I think the problem might be that NetNewsWire doesn't support SVG favicons (/y18.svg) and that /favicon.ico isn't available as a fallback. Were there any changes regarding these files recently?
Don’t make me start emailing you every day to update the default font size from 12px to a modern default like 16px. This isn’t 2007 when everybody had 1024x768 monitors. I had to dust my monocle out of storage before adjusting my browser zoom level.
Thank you for compiling this list. I have been using the site since 2007 and had never looked into the classic page (assumed it was an old stylesheet or something similar). Also had no idea about some of those special URLs such as noobcomments.
If a user has 31 Karma, they can also vouch for a [dead] submission/comment. A vouched submission/comment has its rank restored (and potentially improved as the vouch can counteract the effects of flags).
Is this consistent? I recently vouched for something on the New page and it wasn't restored.
One other thing I would like to ask: I currently bookmark about 20 users whose comments I find especially insightful across a wide range of subject matter expertise. Is there another way/better way to do this?
> I currently bookmark about 20 users whose comments I find especially insightful across a wide range of subject matter expertise. Is there another way/better way to do this?
I'd propose participation for X days and a very loose karma metric (e.g. non-negative karma to block out consistent hecklers) as a compromise. You don't want someone who posted once and disappears for months to be given elevated permissions, because the idea of the karma gate is to adjust to how the community works before getting more features.
semi-artbituarily, I'd say 2-3 months worth of distinct days making a comment or post should be enough time.
Yeah, the karma requirements mean you frequently see people adding noise to discussions rather than signal, just to get out there and seen. I rarely comment and haven't hit the 501 threshold to downvote despite being here for a number of years, mostly because I just upvote what I agree with rather than posting a "me too".
(And I'm now concerned that this comment will feel like a "me too" :) )
One undocumented heuristic that I wish was documented better is that repeated submissions to the same domain, especially without discussion activity, leads to an automated shadow ban. This is just from what I’ve observed browsing /new with Show Dead turned on. There are lots of people submitting their own blog posts or repeated posts from their favorite news outlets who have been shadow banned. Vouching provides a mechanism for restoring posts of particular quality, but most of these people aren’t trying to spam, they just don’t know this unspoken rule.
That's connected to this guideline (from https://news.ycombinator.com/newsguidelines.html): "Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity."
If people submit primarily their own stuff, our software eventually starts classifying the account as "promotional" and starts filtering the posts.
People are totally welcome to submit their own work—that's great—but to use HN as intended, they should do that as part of a diverse set of submissions on unrelated/interesting things. There are too many bloggers/marketers/writers out there that use HN merely as yet-another-channel for the content they're pushing, and that's definitely not in the curious spirit which is supposed to animate HN.
there's a slight difference between "your own stuff" and "same domain"...
I don't know how many false positives would fans/people with narrow sharing interest make, but slight editing of the guideline might help make them less "false"
I'm not sure that makes much difference from a quality-of-HN point of view. That is, if an account is (mostly) posting a single domain, and also is (mostly) the only account posting that domain, that doesn't seem to me so different from using HN "primarily for promotion". Or did I misunderstand you?
299 comments
[ 2.7 ms ] story [ 276 ms ] thread* It seems low-quality, at times almost like the Weekly World News. On HN, we sometimes replace even high-quality science writing with primary sources (when the paper is recent and the writing is just a summary).
* It's high-volume, which is problematic in that it incentivizes users to submit lots and lots of stuff from this one site; front page slots are the scarcest resource on HN and everything competes for it, so high-volume low-mid-quality sites tend to get downweighted.
* It's slathered with ads and bait-y gimmicks.
If there's something on this site that you think would make a good thread for HN, you can probably just find whatever it's blogspamming and submit that instead.
(I'd emailed mods after noting that my submission seemed to have been autokilled. They revived it, though the submission didn't see much action.)
EDIT: You can see for yourself how it's ordered here: https://github.com/shawwn/arc/blob/arc3.1/news.arc#L2611-L26...
This is the original code, which survived untouched till at least 2019 or so. I've wondered if they changed it since then.
The key is the (users) function:
https://github.com/shawwn/arc/blob/arc3.1/news.arc#L144-L145
This returns the keys of the profs* hash table (short for "profiles"). At the time, profs was populated each time the server restarts:
https://github.com/shawwn/arc/blob/arc3.1/news.arc#L88-L91
(dir profdir*) essentially returns an `ls` of the directory in which user profiles are stored. (Yes, they are (or were) stored to disk, not a database.) This was one reason it was hard to add a feature to rename users, which Dan somehow figured out.
So it takes the `ls` of that directory, giving a list in mostly random order, and loads them one by one, putting them into a hash table, whose iteration order is also random.
In modern times, profs* is populated in a lazy fashion whenever users post after a server reboot.
Arc uses "maptable" to iterate over a hash table, which under the hood used hash-table-for-each: https://github.com/shawwn/arc/blob/arc3.1/ac.scm#L1082
Nowadays in Racket, it probably uses hash-for-each: https://docs.racket-lang.org/reference/hashtables.html#%28de...
It does have a "try-order?" option, which gives a deterministic ordering if set. But whenever I set it, I sometimes get strange errors, at least on MacOS. Since it's disabled by default, I doubt they set it. Even if they did, it would be ordered alphabetically, which means your topcolor wouldn't be anywhere close to the top unless someone else whose name starts with "a" also set it.
I only know all this because I've spent five years updating Arc for the modern era over at https://github.com/shawwn/sparc
Past discussion: https://news.ycombinator.com/item?id=38586832
I’m just guilty without knowing the crime.
I fired off an email to Scott, frantically telling him "I think I took down HN. If you’re scrambling to figure out what’s wrong, it’s because so-and-so story is on the front page, and you can fix it by booting it off." A minute or two later the site came back to life, so I emailed saying "oh, never mind. I guess it was something else." And to my surprise he emailed back saying you were exactly right, thanks.
Presumably they either added caching, prevented it from going to the front page, froze the list for posterity, or it’s still a ticking time bomb that can take down HN again if you happen to submit it. Either way, maybe the explanation of your missing topcolor is there somewhere.
And yes my choice is there (#ffaa33)
For example, if you try to submit a link to a particular Google Books page, say the one I just tried:
https://www.google.com/books/edition/The_Albert_Einstein_Col...
the link will remain intact for a short while before losing its search parameters and turning into a link to the book itself rather than to a page in it.
https://en.wikipedia.org/wiki/Canonical_link_element
The reason this is a pet peeve of mine is I often see people complaining about "the HN moderators" (e.g. "the HN mods are on a power trip and keep flagging my submissions!" or some such), where they are clearly taking this conceptual model from Reddit, and they seem to think there is a shadowy cabal of "mods" who control what gets downvoted/flagged (and, in fairness, this does and can happen in subreddits). In other words, when you see your submissions being flagged or comments being downvoted or flagged, 99.9% of the time it's not "the HN mods" who are downvoting you. It's just other, normal HN users (who have earned 501+ karma as TFA explains) who simply don't like what you have to say.
Not saying dang never blocks users/comments, but it's exceedingly rare and I've only seen it for the most obnoxiously egregious behavior after (usually multiple) warnings.
Speculation/intuition, but I figure it happens in ways specifically designed to evade the type of detection you're basing your claim on, akin to shadow banning, whereas on Reddit there's often more visible evidence like corpses of deleted comments.
This is far more control than reddit mods have. Reddit mods have one ability, an that is the ability to remove or not remove a submission, which is a very public action. HN can put its finger on the scales in a way that is much more subtle.
Your account can also be effectively shadowbanned with everything you submit going straight to the “dead” pile. Again no explanation and the only way you realize it is cause your submissions slowly die with no comments (and when you view the site in a non logged in fashion your submissions are not there)
This is not hard to figure out plus when it was frequent, they did.
I'm going to contact the moderation indeed, it's the first time I was seeing non-spam comments being shadow banned automatically.
The behaviour of the account holder.
There are some weird edge cases, like sometimes a banned user, once they're unbanned, will revert to posting abusively. Then we ban them again and they revert to posting good comments again. Who can fathom the psyche of homo internetus.
It's also common for an account to post both fine comments and abusive comments. It's the latter that determine whether we have to ban the account—that's how rule enforcement works. But if a banned account posts something good, there's no reason why the good post needs to remain [dead]. This is what vouching is for: https://news.ycombinator.com/newsfaq.html#cvouch.
Issuing warnings or reminding people of the guidelines (rules) is also part of being a moderator, so the moniker is perfectly cromulent to describe dang's role
https://www.merriam-webster.com/wordplay/what-does-cromulent...
The other reason I think "admin" is a much better word is that it much more clearly indicates that this is a special, privileged role. That vast majority of "moderation" on HN is done by normal users with upvotes, downvotes and flagging (and also, of course, by automated systems e.g. with spam filters/bot detection).
But again, my primary point is that I see users blaming "the mods" for why their post is downvoted or hidden, when usually it's that people just don't want to acknowledge that the community, at large, didn't think they were contributing productively to the discourse of the site.
> Hacker News currently has one full time moderator: Dan Gackle (dang), and formerly Scott Bell (sctb). Their comment replies provide a pseudo-log of Hacker News moderation.
I've never seen an anyone else besides dang comment that they are an HN admin in the past 5-ish years or so.
My post https://news.ycombinator.com/item?id=37956065 got no traction when I posted it and quickly disappeared from "/new" without many people noticing it. The next day, I received an email that this post was entered into the second chance pool. Then sometime later, it got picked for the front page and led to a good amount of community participation! Thanks to the second chance pool, I received several interesting demo submissions to my hobby project!
It’s missing the relatively recent new list called “Classic”.
https://news.ycombinator.com/classic
Classic: Frontpage as voted by ancient accounts
... which, anecdotally, I learned about JUST after getting a comment on there. Imagine my surprise at learning about this (in a manner totally unrelated to the post I wrote that made it to the list), and seeing my own writing as the most recent item on the list.
Truman show vibes for sure.
https://news.ycombinator.com/lists
https://github.com/minimaxir/hacker-news-undocumented#hacker...
https://news.ycombinator.com/item?id=38779197
But I appreciate the fact the consequences for everyone are not the same, and potentially even fatal in current times.
I believe I understand why it's done in the way it's done, but also this can drive self-censorship. Which is a growing problem at large.
Without speaking out of turn, which may just create more work, they can help.
Contact hn@ycombinator.com.
I suppose it depends on one's perspective. We used to call self-censorship "decorum."
The nature of human beings didn't change when we went online, and sometimes it behooves a person to just read the room before speaking. Because of the karma system, the upvote downvote system, and the user curated flagging system, I would classify Hacker News as a "decorum-friendly" message board where a certain amount of pre-filtering is not only necessary but highly encouraged.
Actually, this is the motivation __to__ nuke. It's what I mean about the environment changing. HN is one of the big places that get scraped and everyone is scraping all data that they can these days. It's very reasonable to believe you can be de-anonymized much more easily and that it's only going to get easier. It's not unreasonable to think that someone can mimic my speech and this creates a new vulnerability. I am thinking much more about self-censorship now and being more measured now than I have in the past.
I do agree with you that self-censorship is a growing problem. But I'm not sure how to solve this when we're entering a world where it is the language you use that becomes a fingerprint. I appreciate the records because there is a lot of valuable information here but at the same time these records make us vulnerable in a way we haven't been before. And that it is in a way that not just worrying about nation states being able to do this but that we're posting these records to huggingface. The processing is just getting easier and that's about all that's needed now.
Maybe the answer is to let trusted users become polymorphic, but that defeats some of the spirit of the site.
I'll have to think more on it, get back to you.
Idk, if you have thoughts I'd love to head. I've been thinking of writing them down and posting to HN but unsure and it feels like one of those things that sounds conspiratorial until years later people will say "of course this was coming" lol
Scraping doesn’t even have to be hourly, scraping a thread or account once then `diff`ing the two between weeks or months or years would easily reveal any interventions.
Which would actually expose users who you should retroactively investigate. Canaries.
I presume you are aware of “fuzzed” users. With unassociated comments, submissions, and possibly even changing names. I only recently discovered this myself after a little digging, but you may be able to do this and continue posting.
I’ve been hesitant to mention this anywhere, as discussing how you would red team a site you like publicly might not be the best for the site.
There are other clever possibilities, such as using some combination of artificial & organic content to wash users identities. Like banned or shadowbanned accounts mixed with real users. An acid bath if you will. If you dissolved 100 people into each other, who can really say who is who?
I don't think changing usernames is a strong action either. This is because the fingerprint is your language, not the name you're associated with. See the Enron dataset and the project in Ng's intro AI course, but we'd need to scale that quite a bit (which is very hard).
I'm not aware of the fuzzed users. But I think you should mention it in the open if they are problematic.
I have thought of AI to rewrite my language but that's difficult and only fuzzing too.
There are certainly individual cases where some post might turn out to be problematic later, but that's more of an exception and can be handled by manual intervention.
I do find it annoying in some places how users can just erase their posts without a reason. Because they don't just erase their own content, they take the responses to it with them or at least make them less understandable. And those people responding did put some effort into it.
(My confidence in this explanation comes in part from the fact that I'm pretty sure I had a hand in this coming to pass: https://news.ycombinator.com/item?id=2403716).
The conspiracist rationale here doesn't make much sense, because the site operators also control the displayed comment count and the registration of votes.
Nice to see that some of the suggestions from there, like listing the hidden URLS (/leaders, etc.) seem to have been implemented.
A List of Hacker News's Undocumented Features and Behaviors - https://news.ycombinator.com/item?id=33076053 - Oct 2022 (68 comments)
A List of Hacker News's Undocumented Features and Behaviors - https://news.ycombinator.com/item?id=30459276 - Feb 2022 (64 comments)
A List of Hacker News's Undocumented Features and Behaviors (2018-20) - https://news.ycombinator.com/item?id=26866482 - April 2021 (255 comments)
A List of Hacker News's Undocumented Features and Behaviors (2018) - https://news.ycombinator.com/item?id=23439437 - June 2020 (266 comments)
A List of Hacker News's Undocumented Features and Behaviors - https://news.ycombinator.com/item?id=20292361 - June 2019 (25 comments)
A List of Hacker News's Undocumented Features and Behaviors - https://news.ycombinator.com/item?id=19212822 - Feb 2019 (183 comments)
Hacker News's Undocumented Features and Behaviors - https://news.ycombinator.com/item?id=16437973 - Feb 2018 (391 comments)
Shame I'm not in the top 50 anymore, but there just ain't a lot of good content to submit anymore and karma whoring seems like a waste of time.
https://news.ycombinator.com/hidden?id=drfuchs
2) voting system works well for just users but HN is infested with corporate voting manipulation. It seems that an entity monitoring eg. brand's name is able to use just a few 501+ karma accounts to bury, or a few more to strongly boost, any post
3) based on what I learned about posts' scores (more than a grain of salt - as I mentioned - I'm still figuring things out) it seems a little out of place that sometimes a few hours-old posts (not as old to qualify for the "second chance" thing) with barely any comments and points are positioned higher on the top page than younger posts with more activity (higher score and more comments). I've noticed it more than once with substack and my assumption is that HN may be boosting YC-backed products
I don't see much evidence of voting manipulation (and a lot of evidence of HN being hair-trigger about suppressing vote manipulation).
Imagine a discussion forum for medical topics (science, research, etc.) where 1/3 the users work for companies that sell homeopathic "remedies", and how that would influence moderation there.
You mean to say that next week's Kagi thread won't be as organic and spontaneous as it purports to be? Impossible.
There are so many notable tech people on HN, it would be useful to be able to recognize their posts and comments.
Btw, that's not the record for how long a bug fix has taken. We get there eventually.
Edit: believe it or not, we got more complaints about this. Some people picked their top color specifically to go nicely with the orange Y! So I've reverted this change, except for the Christmas case. Users who have set their topcolor don't see the Christmas top bar anyhow, so don't have to worry.
However, there was a brief period where we were serving the transparent Y for favicons before I noticed that and corrected it; I wonder if the transparent Y somehow got into a cache? Is there a way to force-refresh?
I think the problem might be that NetNewsWire doesn't support SVG favicons (/y18.svg) and that /favicon.ico isn't available as a fallback. Were there any changes regarding these files recently?
What do I need to do to make /favicon.ico available as a fallback?
So this really did break my workflow, hah
If a user has 31 Karma, they can also vouch for a [dead] submission/comment. A vouched submission/comment has its rank restored (and potentially improved as the vouch can counteract the effects of flags).
Is this consistent? I recently vouched for something on the New page and it wasn't restored.
One other thing I would like to ask: I currently bookmark about 20 users whose comments I find especially insightful across a wide range of subject matter expertise. Is there another way/better way to do this?
I don't know if it still works, but I used this a few years ago for this exact reason: https://github.com/veggiedefender/hn-friends
If DMs about web site bug > X then push fix lol
[1]: https://github.com/minimaxir/hacker-news-undocumented/pull/6...
semi-artbituarily, I'd say 2-3 months worth of distinct days making a comment or post should be enough time.
(And I'm now concerned that this comment will feel like a "me too" :) )
If people submit primarily their own stuff, our software eventually starts classifying the account as "promotional" and starts filtering the posts.
People are totally welcome to submit their own work—that's great—but to use HN as intended, they should do that as part of a diverse set of submissions on unrelated/interesting things. There are too many bloggers/marketers/writers out there that use HN merely as yet-another-channel for the content they're pushing, and that's definitely not in the curious spirit which is supposed to animate HN.
I don't know how many false positives would fans/people with narrow sharing interest make, but slight editing of the guideline might help make them less "false"
(I apply those using the Stylus CSS manager browser extension.)