A List of Hacker News's Undocumented Features and Behaviors (2018-20) | Hacker News Reader

A List of Hacker News's Undocumented Features and Behaviors (2018-20)

github.com

> If a user has 251 Karma, they can set the color of the top bar in their profile settings. The default is #ff6600. Here's the complete set of colors users have set.

It's so close, why isn't the required karma 256?

    if user.karma > 250:
      css.topbar.bgcolor = (
        user.prefs.topbar.bgcolor
        or “#ff6600”
      )

CSS only:

    [bgcolor="#ff6600"] {
        background-color: transparent;
    }

Error correcting karma

frankly it's ridiculous that there's a karma limit on a client-side cosmetic feature. I understand one can write a custom CSS rule for this, but if it's already built into the website....

Unless you have to earn it, you won’t appreciate it.

I appreciate my custom CSS just fine, thanks.

It's added flavour, I think. Not every website has to be intuitive or hand people things on a silver platter.

Ridiculous that it's so low.

    [bgcolor="#ff6600"] {
      background: #008 !important;    
    }

    body > center > table > tbody > tr:first-child > td {
      background: #008 !important;
    }

Here's some CSS rules if you want to change it yourself instead of relying on an obviously poor design choice.

Nice things shouldn't be available for throwaways or sock puppet accounts.

I've always wondered why this feature exists and I'm still not sure, nonetheless I've set a custom color (currently a fetching shade of purple - 8185E2), and change it about once a year, previously 00FF91 and FFF200.

If nothing else, I can tell at a glance if I'm logged in or not.

I tried 8185E2 and liked it.

I used a color palette generator with the black text and orange logo and it spit out: CFDBD5, which gives the site a bit of a high desert brush look.

So that each of your sock puppet accounts can have a unique top color. It helps to remind you of which facade to present.

Shoot I do that at work for admin vs non-admin logins, or different gmail accounts (personal vs side-biz). Reddit, too -- leave one permanently on night-mode.

IMO should be an SOP for anywhere that you might log in with different accounts.

How many times have you been logged in as root and didn't realize it?

Now that you mention it, I do the same for my dev installation of HN.

It’s a great visual indicator to remind of your active environment. I use it in just about every app that supports it with custom colors aligned with the environment (e.g. blood red for prod or canary yellow for dev). Gmail supports it as well. If you have multiple accounts with an “official” corporate color, it makes it easy to realize which you’re acting upon.

Same with me, I can tell immediately if I am logged in or not.

I use D8D8C0, a slighly darker shade of the background. I find it aesthetically more pleasing than the bright orange bar. Also makes the logo pop.

So that I can again drop the link that busts the "8 bits per byte"-myth ;) http://www.lispworks.com/documentation/HyperSpec/Body/f_by_b...

I've been around long enough to have had to roll my eyes and say "fine, octet" in conversation.

But I'm grateful that it's mostly stopped happening, let's not revive it!

for fellow dumbells who can't even the their laces may i ask; eh?

Can you elaborate? My older relative who studied computers thought byte= 8 bits. I also thought so. My ego is taking a hit tonight.

A byte is a collection of bits - now almost universally standardized on 8 bits because of network protocols and encodings - but older computers the byte wasn’t as important as THE WORD which was the bit width the machine could process at once. We now refer to 128bit instructions but those are more properly an instruction with a 128bit word.

Lisp provides the ability for arbitrarily defined bytes.

so i wasn't totally wrong. It's mainly semantics.

Since i'm here, why do most storage mediums i.e usb, phones sd ostensibly follow this method (1gb 2,4,8!,16,32 etc) and why are they almost always a few hundreds mb off?

A kilobyte can be defined as either 1024 bytes or 1000 bytes. Similarly, a megabyte could be 1024 kilobytes or 1000 kilobytes, and so forth. Really, 1024 is the number that makes more sense, but hard drive manufacturers promoted the “1000 bytes” interpretation so they could advertise their hard drives as having more megabytes.

Some standards organizations tried to “solve” the issue by defining a “kibibyte” as 1024 bytes and so forth, but you only ever see those in abbreviated forms (ie KiB). I personally hate that convention; the words are aesthetically atrocious and were only invented to please pedants (who annoy me) and dishonest marketing people (whom I detest even more).

Isn't this the difference between speed notation and size notation?

Some older machines had different byte lengths. I believe Burroughs had 7-bit bytes. PDP-10s had 36-bit words which could be divided into any byte size that you liked; 7 bits was common, but if you needed them you could get 1-bit bytes, 6 bit (which was used to encode filenames), 8 bits, whatever.

True, however as you are no doubt aware, one more digit or one less would double / half the karma requirement which is somewhat extreme, and it would no longer be representable by two hex digits.

The CSS colours are specifically 3 octets written as 2 digit hex, so you so while I like the broader idea that "it's a byte of karma" is sufficient, knowledge of 2 digit hex codes also applies to the privilege directly.

Plus, non-standard byte lengths would be a much more niche reference, and I don't think people would get it.

This is my favorite feature of HN. Many, many years ago there was an article posted about how the colors teal and orange go so well together. So I set my topcolor to #00cdcd, and love it.

You don't need karma to set some color, you just need to inject custom CSS.

It should be 0.

It started here: https://news.ycombinator.com/item?id=97573. I don't think that's so bad.

Here's pg talking about how he wrote the code for it: https://news.ycombinator.com/item?id=98265.

This is a really useful list, thanks for putting this together.

I was briefly shadow-banned a month or two ago, due to a few submissions of posts on my personal blog from this account (and not much else).

What was odd was that when I looked at the "newest" page while logged in, I saw my post on HN but when I looked when not logged in, the post didn't show up. I didn't know shadowbanning existed and thought there was some kind of issue with HN, so I emailed hn@ycombinator.com. dang@ sent back an incredibly thorough and thoughtful reply literally 3 minutes later (this was at 7:15 p.m. on a week night).

So while I find it a little odd to 1) shadowban someone without telling them why and 2) intentionally obfuscate the fact that they've been banned by making it seem to them like their submission was successfully submitted, I was really amazed by the care and efficiency of dang's response. So thanks dang!

But that’s what shadowbanning means. Otherwise, it’s just banning.

Meanwhile million dollar corporations buy accounts and astroturf with no penalty.

The little guy is silenced from sharing opinions/links that God Mods don't like, but the big corp can write off a marketing expense that looks like a mature organic account to God Mods.

You have no evidence of that, and neither do we (right now).

When we do get evidence of that, we crack down hard. We've done that in the past in two different $BigCo cases—but this is very rare. What's not rare is internet users making up stories about it based on their feelings. That's so common that the site guidelines ask HN users not to do it: https://news.ycombinator.com/newsguidelines.html.

Hey, how about unbanning me?

We unban accounts when they give us reason to believe that they'll follow the site guidelines in the future. You could start by following the site guideline which asks you to email such questions to hn@ycombinator.com rather than posting them to the threads. If your recent comment history looks good (spoiler: it mostly does), we'll be happy to oblige.

https://news.ycombinator.com/newsguidelines.html

If you were a low-effort spammer (Edit: I'm not suggesting you are a spammer btw) you'd never notice and keep on posting spam into the void. Whereas if you were notified you might create new accounts ad infinitum increasing the administrative burden of HN.

I once implemented a similar feature for the comments on my blog along with an embarassingly simple filter to identify the spammers initially. The volume of spam on a no-name nerd's blog and the degree to which this approach resolved it were both surprising to me.

On a later incarnation of the blog I just removed the commenting feature, but I guess that wouldn't be a sensible strategy for HN :)

The entire point of shadowbanning is making the fact that the user is banned non-obvious to that user--otherwise, they'd be liable to immediately create a new account and continue with the activity that warranted the ban in the first place.

Both "telling them why" and "not obfuscating the fact" would defeat the entire purpose of a shadowban (as opposed to a "traditional" ban).

The whole point of shadowbanning is that the spammer/annoying poster doesn't know they're banned and need to create a new account to keep trying to spam/troll the board. If you tell them generally they'll just create a new account, it's especially bad for sites like HN where there's not even an email required for creating an account. As it's gotten easier to create temporary emails shadowbanning becomes increasingly more effective because it increases the amount of spam that gets blocked before the account owner notices to create a new one.

Getting less undocumented every year:

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)

just out of interest, how many active users are on hn?

dang - are there any features for making HN more accessible? A vision-impaired colleague of mine wanted to check out HN today and made me aware that the contrasts on HN make it really hard for people with impaired vision to read on HN (webaim.org confirms this: https://webaim.org/resources/contrastchecker/?fcolor=919191&...). Using his screen reader (NVDA) didn't help much as the front page doesn't seem to take into account screen reading software. Has someone built a more accessible version of HN? Also, does the HN team plan to make HN more accessible in the future?

It's a problem and yes we plan to fix it. I'm sorry it hasn't been fixed yet; we're just slow at things. Hopefully HN will be around for long enough to justify how slow we are.

That's great news and thanks for the swift reply. Have you considered asking the HN community for help with this issue? I'm sure there would be lots of volunteers.

I will when we actually get to work on it. Before that, it would be counterproductive—just having to process the inputs would cause further delay.

Markdown? Markdown-ish? Something better than the foot-gun of markup currently supported?

My wish list is not full Markdown support, but a little more than is currently implemented:

1. Ordered lists

- Unordered lists

Inline `code snippets`

Super^script

That's it. I totally get why HN wouldn't want tables, bold headers, etc. There's a danger of comments getting visually "noisy" with these. But for referencing code or mathematical concepts, or just listing things, the 4 things above would be really handy.

Yup... Lists, `inline`, and ```code``` would be fantastic. It's so annoying for n.yc to be literally the only place on the internet where you can type into a textbox but that textbox isn't some variation of markdown. (even facebook messages [on web] supports triple-ticks for code!)

> literally the only place on the internet

Haven't worked with forum software other than Discourse, have you? They all still use BBCode.

True, but it's such a mismatch in expectations where the entire tech industry has embraced markdown, but n.yc (tech mecca / water cooler) is most decidedly (and surprisingly) uses NotMarkdown(tm).

HN is an institution. The tech industry should be ashamed that they have HN modes.

I work at Vanilla Forums (we power some of the largest communities online).

I’ve worked a lot on our editors and content formatting pipelines. We support plaintext, markdown, HTML, BBCode, and a WYSIWYG editor with markdown shortcuts.

In reality is estimate the distribution is mostly the WYSIWYG editor nowadays for newer sites and BBCode for the older ones (some users on older entrenched communities really like their BBCode I guess).

You can do code blocks by putting two spaces in front of text:

  This is an example of a code block, if that’s what you meant.

Here's a memorable jingle for that handy new emergency number...

``` https://www.youtube.com/watch?v=ab8GtuPdrUQ ```

  </snark>

I find I can make unordered lists quite effectively

• like this

• using characters like

• U+2022 BULLET

‣ Or U+2023 TRIANGULAR BULLET

Similarly, you can do superˢᶜʳⁱᵖᵗ using Unicode as well (though I would only recommend it for numbers; Unicode doesn’t include all superscript letters, and most of the ones it does include look bad when used together). But yes, I agree that proper support for these things would be very good. (Especially code blocks!)

Yeah, I’ve done it before for superscript characters and even have used fixed width code points to force inline code. But it was cumbersome. (One time my whole comment was written in fixed width characters—I don’t remember why—and I found it a few hours later transliterated to the 0-127 range)

EDIT: Also, with native list support, you don’t get the wide line spacing. I like lists that aren’t paragraphs for each item.

I've written an accessible skin for HN that's about 90% complete. It targets WCAG 2.0. I will post it here when it's released.

The text is mostly illegible, but there are LOTS of other issues too, tap target sizing for example.

Please email hn@ycombinator.com when you do post it!

> I've written an accessible skin for HN... The text is mostly illegible...

I guess everyone has to start somewhere? :P

I'm co-owner of a small Google Group for blind developers started on HN. The other co-owner is a blind developer.

Presumably, he's able to navigate HN since that's how we met.

I'm visually impaired but not blind. I have no problems navigating HN.

You are welcome to join the group and ask how other members navigate HN. Most of them joined via links found on HN.

https://groups.google.com/g/blind-dev-works

The fact that I can navigate HN does not mean that it is pleasurable. The front page makes cumbersome to navigate down the list of publications and the comments sections do not signal where a thread starts or ends and who is answering to whom. Skipping boring threads therefore becomes practically impossible unless I find the top comment and know the uncommunicated feature to collapse threads.

(Greyed out comments are not a bug but a feature)

Perhaps, but not everyone may want to enable this particular feature…

You can turn it off by enabling “show dead”.

That will show dead comments, but near-dead ones are still grey.

In fact, dead ones are gray too, so you effectively increase the amount of gray comments ;)

This is super neat. Not only for beginners/intermediate to catch up with documented and undocumented features. But also for platforms who are seeking a better moderation approach. For me, HN is a very special place - the way people interact, the way that it is moderated. I still dream that other platforms might follow suit someday.

I wouldn't attribute it solely on moderation though. The HN audience isn't very big and thread discussions never last too long because there's almost always something on the frontpage. Not saying it's a bad thing though.

HN is probably the only place that maintained its quality somewhat over the last 10 years for me. So many websites and subreddits blew up, grew too quickly, and burned out in quality.

I think a lot can be attributed to strict moderation, but also like you say being pretty niche helps in that the mod team doesn't have to be very big.

I think it’s also that HN is optimized for “interesting” rather than for eyeballs and outrage.

It's actually DEoptimized for outrage, if one reads the guidelines closely.

HN gets 1400-ish submissions and 12,000-ish comments a day, and 5M-ish users a month. Whether that's big depends on what you're comparing it to. It's big to us.

The ratio of submissions/comments is impressively high, this surely explains why the upvoted articles are almost always interesting.

Thanks for all that work!

It makes me unreasonably happy that FFOOFF is far up the top color list. Apparently it seems to be hard to Google, but for those who don't know, prior to wide spread support for alpha channels, FF00FF was one of the colors commonly used for bitmap transparency. Not exactly a commonly used color and exceptionally obvious when working with a sprite sheet.

Edit: You can probably guess what my banner is set to.

I still occasionally see it in games when a texture is missing or otherwise unable to load.

This is still my goto colour everytime I want to see if a css rule is actually applied to something.

I never understood how the vote counts works (the number that is added to your profile). It's certainly not lineair with the submission upvotes and it even seems to slow down over time (i.e. you need more votes to add 100 points to your profile)?

I think the one point default per comment isn't added? It's only additional points? At least, that's what I've noticed

There is something more than that. I noticed that the few highly-voted submissions I have didn't add the whole amount to my karma.

I have the impression that submission votes only count half of what comment votes count, or something along those lines.

It's certainly not just a factor. The first twenty upvotes count more than, say, the 501st through 520th upvote.

Earlier upvotes are more impactful in getting a post on the front page and in affecting its rank (because the rank algorithm uses time decay), but they don't count any differently towards karma.

Is there really no delay, or anything else going on? I'm sure I've seen my upvote count be out of sync with the changes I see on my scores for recent comments.

Does https://news.ycombinator.com/item?id=26868069 answer your question?

I could be wrong but I think it used to be 1:1 with submission upvotes. It may still be linear but less than 1:1, I'm not sure. It is still 1:1 with comment upvotes I believe.

That makes sense to me, a user submitting a press release or news article about a current event being the "lucky winner" where their submission gets traction doesn't automatically equate to them being a good community member in general.

I think some sites are weighted differently. More commonly posted sites require multiple votes for one point added. Less commonly posted sites require less upvotes for a point added.

Or something like that anyway.

It's 1:1 with comment upvotes, less with submission upvotes (how much less I'm not sure but my observation would peg it around 3:1). Personally I think it should be the other way around as a submission can add a lot more value than a comment.

It's not 1-1 with either comment or submission upvotes. It's more complicated than that, largely for anti-abuse reasons.

Hah well my two downvotes here and two corresponding lost points say otherwise. But I'll take your word for it.

In a weaker gravity field with a low sample frequency the algorithm does indeed appear 1:1. The complexity arises once a peripheral second gravity field can no longer be ignored or a certain noise threshold is passed.

The equation ends up looking something like this: https://slack-files.com/TEGQB8GNS-F01UU5FMKRR-820b186081

Once you start thinking about it in terms of the definite integrals of a piecewise function you'll see where the weights would be applied in more extreme situations.

I have noticed that a single downvote sometimes counts for -2 rather than -1. Sometimes I'll undo a downvote because I think a comment wasn't bad enough to cost 2 points. I have been trying to spot the pattern behind it for over a year with no luck. Can you explain this or is it secret?

I wish there was a feature that doesn't grey out the comments for a couple of hours until everybody has gotten a chance to vote. As it stands currently, a single person downvoting a new comment from +1 to 0 leads to a slightly grey comment. People here usually see this and form an implicit bias. I've seen comments that are perfectly reasonable, has no opinion, it is on topic, and follows guidelines and yet it gets downvoted. I am conjecturing that it is because of this bandwagoning effect.

Look into this please. Greying out comments is fine. Just let some time pass, even 30 mins would be a huge improvement. Alternatively, allow minimum 2-3 people to vote before greying out comments.

At the same time, I try to make it a point to vouch/upvote reasonable posts. If I had no indication it'd been downvoted, I'd never upvote it to compensate. I'd read the comments now and if I came back tomorrow to see if there were more discussion I'd be very confused to see many comments grayed out (The great comment massacre?). Based on how other comments move to gray and back I suspect I'm not alone.

Does this outweigh the bandwagoning effect though?

What about HN run an A/B test. Half of all new posts have delayed greying for a few hours. See which category accrues the most downvotes (or most net votes).

The effect of such timeout may be quite opposite.

As Jtsummers points out, people can and do give corrective upvotes in such cases. I do that all the time, and I hope people understand that it's part of being a good citizen here.

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...

It's actually kinda useful as an indicator of something that could use a corrective upvote.

I could see putting a time delay on the second or third or later upvotes, but the slight graying seems like the perfect indicator for both "maybe skip it if you're not in that mood" and "maybe take a look if you're in a generous vein".

(Minus the accessibility issues, but they're fixable with CSS.)

Are most users net positive in their votes, or are there some that only upvote, and some that only downvote?

All of the above.

I have had comments that bounce back and forth.

It seems down votes are far too easy to make. A down vote can be spent freely and carries no expense.

Expecting humans to be judicious is probably a little too optimistic ( otherwise we wouldn't need laws and governments ). Especially on the anonymous internet.

I am thankful for the upvote people who provide a helping hand when needed.

Since downvotes seem to collect quickly and possibly silence someone early in a conversation does anyone have design thoughts on ways to 'moderate' the down vote?

There was a massive conversation last week [0] about down votes (not for the first time) if you're interested. Personally I think HN largely gets moderating right in general; I don't think it's about moderating the down vote specifically, but firmly encouraging a culture of good faith debate in which the down vote plays the role of hiding bad faith/low quality content. Like GP (and anyone who uses this site at least somewhat frequently) I've seen down voted comments that seem innocuous or even thoughtful but I get the sense they're a vast exception compared to the weaponised down voting of e.g. Reddit.

[0] https://news.ycombinator.com/item?id=26810972

Thank you for the link.

One thing I've noticed is that when people are new one of the first things they complain about is the voting system. I've seen a lot of green names recently, so there probably needs to be some concerted effort to educate users on what they're voting for rather than redefining the system.

I would tend to agree. Learning to be part of a community should be the a goal of a new user. But the community should feel some obligation here too.

My initial experience here was I made a couple of comments which I thought were helpful and instantly had a -7 karma. Ten minutes into my experience and I was an outsider. That is a bit discouraging for a someone who has been coming to HN for years and finally decided to make an account.

Perhaps this is why people complain about the voting system. This does give the sense that you need to conform in order to be accepted.

I just had a look at your comment history. These are (of course) just my own opinions ...

You appear to have mostly commented on 'controversial' topics - you are more likely to be downvoted (and maybe also upvoted) in such discussions compared to more technical topics.

Also, you had multiple typos/errors - this always makes me think that you did not spend so much time and thought on your posts. Why are we spending time reading something that you did not spend time proof-reading?

When I consider posting, I usually ask myself - 'does my comment really add value to the discussion (or, in this case, sub-discussion)?' If not, then probably I ought to just keep reading. Heaven forbid that one writes comments just to get some points!

Yeah that's a known hurdle you will have to overcome. Observing this community and participating in it are two different ball games. There's a lot of unsaid rules to discussion here and they vary person to person, but they are within bounds consistent.

I've mentioned this before, and this seems like as good a place as any to say it again: I think HN is a bit too profligate with downvotes, and I've seen it get incrementally worse over time.

My solution to this would be, at 501 karma, a user gets 5 downvotes per 24-hour period, use them or lose them. 1 more downvote per 100 points.

This would gently teach users to downvote a bit more sparingly. Karma isn't that hard to come by, and even when I'm in a really bad mood (or there's some awful thread) I'd be surprised if I hit 30 downvotes in a day. By the time a user had as much downvote ability as they're likely to use, a habit of being less thoughtless about it would be well-established.

Human psychology being what it is, there is a subset of high-karma accounts which never ever downvote, and another subset who basically just come on here and slam the dislike button until their spleen is vented. I don't consider these approaches to be equally beneficial.

It does get confusing when someone writes a potentially interesting comment that also includes something many people could vehemently disagree with. I see it a lot on cryptocurrency posts, for example. Is it grey because it’s totally and completely wrong or is it grey because people disagree? ‘Tis a mystery!

> My solution to this would be, at 501 karma, a user gets 5 downvotes per 24-hour period, use them or lose them.

That's straying too close to the Slashdot voting system for my liking.

Can you elaborate on that?

Genuine question, karma belongs in the hall of Hard Problems next to naming things and cache invalidation.

Slashdot has this concept of meta-moderation, where you get randomly allocated 5 extra mod points to use how you see fit.

It's been a long time since I was a Slashdot user, so the system may have changed since then.

One of the unintended consequences of domain shadowbanning is that sometimes quality content gets buried. Which is an unfortunate trade-off, given the amout of spam content from such domains.

For example, here are some recent articles from https://dev.to that were submitted multiple times by different users, but were dead because the domain is shadowbanned:

https://dev.to/kprotty/understanding-atomics-and-memory-orde...

https://dev.to/yyx990803/announcing-vite-2-0-2f0a

https://dev.to/jaredcwhite/why-tailwind-isn-t-for-me-5c90

https://dev.to/lazerwalker/using-game-design-to-make-virtual...

https://dev.to/nwtgck/the-power-of-pure-http-screen-share-re...

https://dev.to/bkolobara/writing-rust-the-elixir-way-2lm8

I'm curious why dev.to is shadow banned, of all domains I'd expect this one to not be on such a list.

I guess because it's mostly tutorials, which are downranked: https://github.com/minimaxir/hacker-news-undocumented#downra...

Oh that's a good point, thank you!

https://news.ycombinator.com/from?site=dev.to

Do you have 'showdead' enabled? All the recent ones are marked as [dead].

I never knew that there was a second moderator! (sctb)

Their comment history only has some comments from 2019, and is much shorter than dangs. Did they only have a short stint as a moderator on HN, or do they now work from another account/behind the scenes?

Scott stopped working on HN about 18 months ago. https://news.ycombinator.com/item?id=25055115

Scott did leave: looks like I forgot to update it. Will do now.

(off)

For your list of downranked content: it looks like crypto is either downranked or banned.

It isn't. Such threads often get a lot of flags, though. I assume that's because it's at such a late stage in the hype cycle. The huge amount of repetition (not to mention near-spam) makes a topic less interesting.

The most interesting stuff is toward the bottom. I found this fascinating and unintuitive:

> Downranking of Tutorials

> HN submissions which are tutorials are downranked by moderators, as they gratify intellectual curiosity less.

Not that long ago, you could very quickly build a decent SEO campaign around "X vs. Y" tutorials and HN - write 5,000 words on e.g. Angular vs. Ember, get some friends to upvote it, and if you made the front page you'd be on the first Google SERP for Angular and Ember the next day.

I personally know a couple of early-ish YC companies who took advantage of that to boost their content marketing ops, and it wouldn't surprise me to learn that they were part of the reason this policy exists.

Nope - we just figured out one day that tutorials aren't intellectually interesting, in one of those obvious-once-you-see-it moments.

Edit: here's an explanation about this that I send to people who email "why was my thing downranked" when $thing was a tutorial:

Tutorials are not a great fit for HN. HN readers are looking for stories of intellectual interest that go deeply into a topic. Tutorials are great for readers who need to do the specific task the tutorial is about, but that is a small percentage of the total audience. Also, they tend to be directed toward beginners, which is wonderful, but not what the HN audience is looking for. Such articles rarely do well as HN submissions. You'd be better off writing something of general interest for an audience of smart hackers.

I believe that explicit banning is also now indicated. I'd requested something like this for long-since-shadowbanned users whose ban event was hidden in the distant mists of time (or who are sockpuppets of banned profiles). This is rare, but does occur.

Also can apply to spammers -- this is usually relevant on submissions / new queue.

The YC-company bias can also apply to negative reporting on YC companies, which is less moderated as something of a counter to charges of favouratism. See:

https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

Including from this thread: https://news.ycombinator.com/item?id=26868116

If you mean the "[banned]" indicator that appears beside some usernames, that's only for a few banned accounts.

The reason is that we were getting indignant emails about some of the most horrible stuff banned accounts have been posting, saying "how can you possibly allow this on Hacker News"? The answer, of course, is we don't. All those accounts had been banned for a long time.

The problem seems to be that (a) some people forget they have "showdead" turned on or don't know what it means; and/or (b) some third-party software (e.g. mobile apps, browser extensions, or websites offering alternate views of HN) displays [dead] and [flagged] comments as if they were normal comments.

There's not much to be done about (a), but (b) is a disaster. Many users of such software assume that what they're looking at is HN. (Previous comment about that: https://news.ycombinator.com/item?id=24335289.) Possibly we'll have to introduce some sort of standard and enforce it via API keys. Or maybe just stop serving [dead] or [flagged] comments via the API.

Yes.

There's also the confusion between "flagged" and "dead", which still gets unclear at times.

Not serving dead/flagged via API mostly makes sense, though it could make some analysis harder.

I'd add the red top bar for Christmas as an Easter Egg (... or a Christmas Egg?)

apropos undocumented:

Can I limit HNs search functionality to only include my favorites?

I often know I favoured something but for the life of me cannot find it via normal search and going through my hundreds of favorites by hand is impractical.

Alas no. It's something Algolia could build, though, if they wanted—in fact there could be an operator for searching through anyone's favorites ("fave:splittingTimes" similar to "by:splittingTimes"), as they're public.

I also have a similar need but for upvoted content. Sometimes I keep scrolling through pages of upvotes to find a specific item. The problem is that favourites and upvotes are not exposed via the API and require scraping, otherwise it would be easier to build a search tool for them.

We'd have to build that into HN itself, since upvote data is private (extremely so). But it's the sort of feature whose value grows as the site continues, so I could imagine us adding it eventually.

(2018) maybe

https://news.ycombinator.com/item?id=19212822

and discussion less than a year ago: https://news.ycombinator.com/item?id=23439437

(2020) I guess because that's the last time it was updated

I mean git projects aren't generally given a (YYYY) marker, since they're expected to be continuously updated - no reason it couldn't be updated tomorrow if someone finds something new, and it's not like the url is pinned to the specific current commit. I think it should probably just not have a (YYYY), even though it is prose.

In fact it seems this discussion has lead to a furry of updates. In fact many of the links on the README now direct to this discussion which is oddly meta.

So, um... what are the "noprocrast maxvisit minaway and delay" options on the user profile setting screen for??

I'm guessing they are documented, but I'm too lazy to find/read. ;)

No procast locks HN for you after some time (to avoid procrastination, as the name suggests ;). Max visit is the time you allow yourself on the site (starting on the first visit) and minaway is how long it is blocked. It's in minutes, iirc

Seriously?!? That's hilarious! Thanks.

No problem :) It can actually be pretty useful sometimes, even though I have it disabled right now. You can also use it as a "spree breaker" by setting the values to something very low, like 60 visit/20 away.

It originated here: https://news.ycombinator.com/item?id=77173.

Ah, a data point for my future essay "Nobody Reads the FAQ".

https://news.ycombinator.com/newsfaq.html

Though an FAQ would typically be consumed by those wanting to know what the service provides and how. Not options on how to configure it.

Perhaps adding a link on the user's settings page to the 'FAQ' would be helpful.

Was the period removed? I remember a few months ago, there would be a `.` after the post age if it was the user's most recent post. Looks like it's not there anymore though.

A lot of people didn't like it so I made it moderator-only. I'm not sure it's really that useful in the end.

Why not make it a user preference, like showdead & co? Just make it off by default.

There are only so many slots available for user preferences before things become much too complicated, and I don't think that one clears the bar. It's much harder to remove these things than to add them. Better not to get on the bad side of that Maxwell demon.

To be honest, I'd take all the options you can throw at me. And a command line to manage them. It's HN after all.

PS. Just hide them behind a karma threshold and a tickbox to acknowledge that they all are ephemeral and can disappear at will. Complications, like in mechanical watches. For people who can appreciate them, to show what's possible :)

Certainly some familiar names on the top 100 accounts ranked by karma totals but seeing the username “toomuchtodo” with some 48,000 internet points gave me a pretty good grin.

What's the difference between downvoting a comment and flagging a comment?

I ask because I don't have 501 karma and I can flag a comment (but not downvote).

Downvotes can be countered by upvotes and are more for low-quality content (or things you disagree with, although many - including me - try to refrain from doing this). Flagging is for content violating the rules, for example spam or insults.

> Accounts which are less than 2 weeks old will appear with a green username.

I thought they were "approved" users.

I thought they were the poster or parent comment user. Green probably isnt the best choice for new users.

"Green" does mean "new" though. I think a bigger issue is that other admins/OPs/etc don't have colors.

"To be green" is an idiom that means "to be immature in age or judgment; untrained; inexperienced."

So it's not completely misplaced.

It is frequently used on other forums and sites to show users as being mods or admins. Why not show account age if the signal is important?

I can certainly think of some reasons why it makes sense. But I can think of plenty of reasons why it doesnt as well.

Once you decide you need a color to mark new accounts, the same reasoning can be applied to pretty much any color. Probably the solution is to add a [new] tag like it happens already to dead, flagged, and banned entries or accounts.

From a web UX perspective it's true, but green is a common metaphor for new in English.

https://www.quora.com/How-did-the-word-green-become-synonymo...

It's also not a good idea to use only color.

I assume this is about the users colored with #3c963c. A quick search tells me that's "dark moderate lime green", which I assume to most people is just "green". I have no clue if it is though, and I'm often wrong about color things. I should make a list of the different colors names can be tagged with and make a little tampermonkey script to call them out.

I never understood it - can't we just have something like "fred_bloggs [new poster]"?

(although the "undocumented" bit about this is just the time span, what green users are is documented in the FAQ)

Not sure why downvoting would require a minimum point score other than to discourage lurkers (like me) to lazily engage in discussions (even though I believe voting is in fact engaging). This added to the fact that the HN crowd is notorious for downvoting unpopular or even slightly pessimistic comments sounds to me like forced positivity.

I suppose it's more to prevent digital mobbing then to force positivity.

Greyed out comments leads to serious downvoting bandwagoning too.

The -4 maximum is there specifically to limit the impact of that. I'd say that 70% of my comments which stay in the negative, I earned it, and the remaining 30%? Well, you can't buck the hive-mind: there are simply some things I believe that most people disagree with, and my choices are to shut up about it or take my licks.

I presume it's to make it harder to create sockpuppet accounts.

Also, reaching those 500 points requires quite a bit of contribution, so there's a good chance that you understand the rules by that time. I know the downvotes here can be annoying, but in comparison to Reddit, for example, it's very calm, probably in part due to that limit.

another undocumented feature: HN shows me all of my flagged submissions when i am logged in, when viewing anonymous it shows me only one flagged submission - but that one got 114 comments before it got flagged. is this a bug or is it a feature?

Here's a thing I don't like and it's not mentioned in the list. If I submit a story I can't downvote on any of the top level comments. Why is that?

Probably the same reason you cannot downvote any direct replies to your comments, such as this one.

I can understand why downvoting replies is prohibited, it could harm conversations. But prohibiting downvoting in submitted stories assumes some kind of self interest, which I doubt is the norm.

I think it's generally "the submitter shouldn't be able to shape the discussion too much". I.e. also why you aren't supposed to put your own spin on a headline, text submissions are separate and not supposed to be used for links (with a few exceptions).

Sure, that makes sense.

It could be as simple as: coding this would be an exception, and no one has done so.

`(if (not (eq parent.user user)) (set! allow_downvotes true))` (probably not valid Arc!) is easy to code up, an exception would be more convoluted.

Presumably the same reason you can't downvote immediate replies to your comments which is probably there to discourage toxic back-and-forth downvoting.

This was very helpful. Thank you!

> https://github.com/minimaxir/hacker-news-undocumented#implic...

Disagree with this section, especially since it is basically presenting an opinion as fact in the same class as other verifiable facts on this post.

My own anecdote, but I find that diversity and inclusion posts that have a tendency to get downvoted are ones where you can see the flame war coming from a mile away: the post offers little in actual new data, so then you have 2 "sides" arguing and talking past each other, primarily because both sides are unwilling to acknowledge that it is possible that both of these can simultaneously be true: (a) women and minorities experience discrimination in tech, both overt and subconscious, and (b) other factors besides sex or race discrimination can account for different rates of participation in tech.

On the contrary, in my experience, posts that offer detailed new information around diversity in tech tend to be highly upvoted (the original Susan Fowler post about Uber was one such example).

Admittedly it's merely another anecdote but this post on GitHub changing master -> main comes to mind.

Inflammatory title, subject matter related to diversity, and yet huge karma and plenty of substantive conversation. https://news.ycombinator.com/item?id=26487854

In my experience, by and large the posts that get traction are the ones that contradict the perceived cultural trends.

So a post defending RMS will get much more traction than one explaining why it's important to hold him accountable.

A post dismissing Github's default branch name change as empty pandering will get much more traction than one explaining/defending why language and terminology matters, in all contexts.

^ These are feelings, I don't have data on them. So I think minimaxir did a really good job of presenting this problem in a more detached data driven way. D&I posts ARE more likely to be downvoted. Doesn't mean that ALL D&I posts are treated equal, though. The Susan Fowler post is a notable exception.

I wrote about this in my own comment at the same time as you: https://news.ycombinator.com/item?id=26867142

I like the feature of green usernames

> Implicit Downranking of Topics Around Diversity and Inclusion > Likewise, topics around diversity and inclusion in tech have gained lots of visibility over the past few years. However, despite these discussions not being off-topic, they tend to be flagged to death by users regardless. Unfortunately. (Moderators occasionally unkill such threads if they see it in time, although it rarely sticks).

This is a very interesting subject. I would dare say that the majority of technologists (by pure numbers) find the human, philosophical, and psychological impacts of technology less interesting than the pure technical considerations. (Older and more senior technologists gravitate in the opposite direction)

Yet, the most interesting and in-depth hackernews discussions directly touch up on privacy, censorship, employee motivation, the entrepreneurial mindset, managerial skills and lack therefor, regulation, corporate responsibility. All of these are strictly "on-topic" by the definition on https://news.ycombinator.com/newsguidelines.html:

> On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.

So clearly, this community is passionate and interested in a wide range of "fuzzy" non-STEM subjects.

Yet when it comes to D&I, the community is primarily disinterested and not intellectually curious, or has already concluded that anyone who IS interested is pushing a purely "political" agenda. Therefore, the only "interesting" stories that get upvoted are those that butt against the current cultural movement (e.g. defending James Damore, RMS, etc).

I am personally very intellectually curious about this very pattern, but I appear to be in the minority.

I wonder if people on HackerNews are aware that a huge percentage of female and minority engineers that I know find the HN community discussions completely disreputable and unappealing. And it's a direct consequence of how they treat these subjects as "uninteresting" and "off topic", because to the dominant in-group these subjects are just not relevant.

Personally, as a technologist, a "hacker" (Whatever that may mean, but i identify strongly with https://en.wikipedia.org/wiki/Hacker_culture), I consider this blind spot in this community a "bug". I don't know if it's a fixable one, as any attempts to mitigate it would clearly be perceived as intrusion and a restriction on the individual freedoms of expression. But the status quo is the status quo, to the point that even this manual of undocumented "features and behaviours" has to call it out.

The problem is that D&I threads are _not_ intellectually curious. People are too emotionally charged to have a genuine, dispassionate and reasoned discussion about it.

Why does one preclude another?

One can be intellectually curious while also being emotionally charged and passionate.

There is genuinely outrageous behaviour out there in the world. China is orchestrating a modern genocide. Russia is maneuvering to violate international sovereignty. The USA has a humanitarian crisis at its border. And there is a climate change emergency being mostly ignored. It's reasonable to have an emotional charge to all these discussions even while looking for logical solutions to these problem.

The other issue is that a lot of people don't classify behaviours as "emotional" consistently. The prototypical example is feminine expressions ("tears") are deemed emotional while masculine expressions ("anger") are not.

That may seem ludicrous but it happens on this site too. People will go on tense emotion-laden tirades defending personal freedom or outrage over the treatment of <insert problematic public figure>, and they are heavily upvoted and deemed "intellectually curious". While those that are pointing out the problematic behaviour and the impact it may have had through heartfelt experience stories are deemed "emotionally charged".

You've brought up a number of good and subtle points in this thread, so please don't take this comment as a generic disagreement—this stuff is complicated. It's not only hard to get right, it's hard to even discuss precisely. I just want to mention that

> People will go on tense emotion-laden tirades defending personal freedom or outrage over the treatment of <insert problematic public figure>, and they are heavily upvoted and deemed "intellectually curious"

... isn't really accurate. Those tirades are not deemed intellectually curious, either by mods or by the majority of the community. The vast majority of such threads have been flagged. I wrote about this here: https://news.ycombinator.com/item?id=26713636.

Just my two cents of feedback: amongst my friends and colleagues in industry, I'm one of the last people that still participates here. A few of them only participate to the minimum necessary to help their startup. The rest absolutely abhor this place, and that varies from people I'd describe as ranging from leftists to neoliberals.

There are many people who have been alienated by what the poster above is talking about, and unfortunately there is a very clear demographic slant in who is being alienated.

So not trying to be spammy here, but upon reviewing my prior comment I realized I really was not specific enough:

You cannot use the data from hnews to understand the audience hnews is alienating. While I think you've overall been a very positive force here, and moderation has improved from what I remember some years ago, your link is not an adequate response to this issue.

You need to think more about the people who will never tell you what you're getting wrong, because they regard this whole place as too toxic to bother with.

I hear this from time to time but it's hard to know how to assess it. Sometimes it comes from people with strong ideological commitments; in that case the word "toxic" sometimes carries an ideological charge, and in that case it becomes difficult to disentangle how they personally feel on the site from how they feel it ought to be moderated ideologically. Those are different questions, but it's not always possible to discuss them as different questions. It gets complicated quickly—which doesn't mean we don't care. I don't want people to feel badly about HN.

There's another aspect too. In some tech subcultures it became fashionable to diss HN. For a while it was a sport on Twitter to pass around links to something awful someone ran across on HN and use words like "cesspool" and all that. People would unify around how much they abhorred it. This used to bother me, because typically the something-awful would actually have been moderated or flagged or banned here quite quickly, but people would pass around screenshots anyway and talk about it as if that was HN. It isn't a good-faith discussion; the conclusion is predetermined, no one corrects false statements even when they find out the truth, and so on.

Eventually this started bothering me less because it is simply the way subcultures work—when a larger entity becomes perceived as mainstream, subcultures differentiate themselves in opposition to it, and people bond around abhorring the opposed object. It's no different in music, say. That variety of criticism can't be satisfied because the abhorrence itself serves a valuable function, and it would be a category error to try.

I'm not saying that these are the only two dynamics—on the contrary!—but they are factors that make this sort of criticism difficult to assess. Actually, the point you're making shows up often enough that it seems even to be a bit of a trope in certain circles. From that place it's not so easy to communicate further. The only thing I know that has much chance of working is a specific connection with a specific person around their specific experience. We're open to that and do it all the time.

p.s. Re "your link is not an adequate response to this issue" - sure, that link was just responding to what the GP (indirectly) said about Stallman threads on HN. One reason I replied on that narrow point is that there have been many mistaken claims about it. But there's another reason too: it's orders of magnitude easier to respond to narrow points than general ones. I spent over an hour writing this comment, and still I have to be paranoid about whether and how it may be open to misunderstanding. It's nearly impossible to have this sort of exchange in public—there just aren't the hours, and there's so much more downside than upside. Yet it's important to try anyway, sometimes, because otherwise we seem aloof and uncaring, which is also a dead end.

> The prototypical example is feminine expressions ("tears") are deemed emotional while masculine expressions ("anger") are not.

For what little it's worth, I do classify both as emotional and usually disengage when either happens.

For what it's worth (And I've had plenty of disagreements with dang), I also think he has an impossible job.

By seeing the worst of both extremes, it can be easy as a moderator to think "well there are people who think HN is too right wing, and people who think HN is too left wing. That must mean we're in the center!"

Just in this very thread you have https://news.ycombinator.com/item?id=26868622 who is an agitator convinced that HN is a liberal cesspool.

Is theknocker genuine? Or a provocoteur? Does it matter?

> By seeing the worst of both extremes, it can be easy as a moderator to think "well there are people who think HN is too right wing, and people who think HN is too left wing. That must mean we're in the center!"

Indeed, and all the more so because (a) the numbers of people who make such claims are roughly balanced, and (b) their comments resemble each other so closely. But I would never use the word 'center' in this context because it connotes political centrism. People sometimes misinterpret my moderation comments on political-bias claims as an implicit advocacy for centrist politics, but that's jumping to a different orbit.

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

You're right, and I didn't mean to suggest that's your intent either.

I think the point I was getting at is that no online community has equal participation across the full spectrum of social and political opinions, or representations.

But having an equal volume of opposing but comparable magnitude feedback from the extreme positions doesn't mean that both positions are equally represented amongst the community.

Going deeper would require me to enter conjecture and flamewar-baiting territory but my broad opinion is that the Online Right is far more sophisticated in its ability to persuade than the Online Left.

> I wonder if people on HackerNews are aware that a huge percentage of female and minority engineers that I know find the HN community discussions completely disreputable and unappealing. And it's a direct consequence of how they treat these subjects as "uninteresting" and "off topic", because to the dominant in-group these subjects are just not relevant.

I don't quite see the connection here, too be honest. Just because D&I discussion is mostly not welcome (without judging either way) does not mean the community overall is unwelcoming. I can see why one dislikes the style of discussions on HN and that's fair, but "allowing" D&I topics without changing anything else would probably not fix it IMO. Or am I missing something?

I don’t think anyone here is accusing the community as a whole of being unwelcoming. Rather the problem seems to be that it is disinterested. And when some toxicity arises the lack of interest to address this toxicity comes across as unwelcoming.

I am not a member of a minority group so I’m only guessing here, but I can imagine the frustration of seeing something you perceive as hostile against you, and then see some backlash, but then witness the backlash being deemed uninteresting. If it happens once, okay whatever! But then you see it again, and again, and after a while the frustration is not worth it and you leave the platform.

You only need a few unwelcoming individuals and a community that doesn’t care about correcting it, then the community as a whole can come across as unwelcoming. I think that is the problem here.

The pseudonymous Internet allows us to not wear our increasingly politicized, some might say fetishized identities on our sleeves. Don't assume that minorities agree with the woke agenda. Many find it condescending or uninteresting at best.

> If the comment desaturation makes Hacker News difficult to read

...then make it accessible. Sighted users do not have to click a timestamp or install extensions to have legible text. Why should the visually impaired?

The extension linked to does nothing to improve the legibility of "X points by Y 30 minutes ago..." or "help" etc either.

Unable to reply to comments -- edited to remove.

Response to dang: I disagree with the no-delete policy (our comments are ours and we should have the ability to delete them), but more importantly I refuse to leave a comment up when my ability to reply has been disabled.

I'm curious about the potential for comment bots that specifically aim to gain upvotes. Not with long comments, but with shorter more supportive "me too" comments. Target the "I can't be bothered to reply but I agree with this comment so I'll just upvote it" reader.

> The original forum style was much better.

How was it before? It would be constructive if you could point out some differences. I understand that they can be subtle and hard to discern but I would appreciate any pointers to satiate my curiosity.

>our comments are ours and we should have the ability to delete them

I strongly disagree with this. When you're part of a conversation your comments are critical to understanding the conversation. By having a conversation in a public space, you're allowing anyone who wants to listen in, and your comments are necessary for that. I can't tell you how frustrated I've been searching through old reddit threads in particular where the user has completely nuked their account, and the only trace of what they said is in replies. (Worst is when it's a technical issue and their comments are the only relevant search results: xkcd.com/979 where the answer has been actively removed.)

If you want to have a private conversation, please use a private medium. Otherwise you're gaining the benefit of public spaces (you can view anyone's conversation, their ideas and wisdom, and watch the development of discussions and understanding) while preventing others from gaining those same benefits.

There’s an uncanny sameness to the point you’re making and the cartoon even without the deletion modification. Ultimately your point is about missing context (albeit through deletion). Here the missing context is when the cartoon was published. (Google tells me 2011, but there isn’t anything on the page that I can see that says so and certainly nothing in the cartoon itself.) Although the joke still works, I feel there is a subtle difference reading it in 2021 compared to if I had read it in 2011 (which I probably did).

Not only do I find it somewhat rude[1], I also think the ability to delete comments is offering a false promise to the user - if you've publicly posted something on a notable site for any length of time, there's a good chance there's a copy of it somewhere else (as has happened with this post).

[1] subject to obvious exceptions like accidentally revealing personal/private information, it being spam/off-topic, or a misunderstanding (editing with a correction is my preference though)

>There's a mod who curates comments based on what's politically expedient according to leftist valley dweebs

https://news.ycombinator.com/item?id=26148870

Don't insult my intelligence. Discourse here is confined to the limits of acceptability according to the neoliberal mainstream. This is abundantly obvious to literally everybody. Of course the liberal mainstream endeavors to enforce that.

I won't be treated like a moron by a moron.

"Abundantly obvious to literally everybody" is something that all the commenters who I quoted at that link would agree with. They just disagree violently about what is abundantly obvious to literally everybody.

How about this: There is no doubt in my mind -- none -- that you know exactly what I'm talking about.

And all anyone really needs to know, in order to suspect dishonesty and censorship from you, is an awareness of your shadowbanning tactics.

-----

It's ok though. There are other culturally influential websites where people/groups can market their ideas, cults, viewpoints, albums, etc. And some of them don't shadowban. Some of them do something a little different instead.

It all works out. Have a nice day!

I won't be treated like a moron by a moron.

That's just way inappropriate. Probably the only reason your comment isn't [dead] is you're in a discussion about moderation.

I wouldn't attack Dan so personally. Instead, think of the community here. There is a hint of who they are in the following quip. Actually it explains things well enough that it tends to get downvoted whenever I post it. Perhaps because it hits a little too close to home:

HN is open minded about intellectual inquiry as long as that inquiry doesn't challenge anything an average Californian already believes

I don't remember exactly where I found that. But just think about it. There are so many people on HN from the Bay Area. The submissions, discourse, and moderation all reflect that "bias". What you're observing is simply an emergent behavior.

Only about 10% of HN users were anywhere close to the Bay Area, the last time I ran the numbers. And only about 50% were in the US.

Do those numbers also hold true for % of submissions, % of comments, % of upvotes/downvotes, etc?

If you have a cosmopolitan group of "users", but the most active ones are largely from the Bay Area, then the point about Bay Area groupthink still stands.

I checked that at the time and there wasn't much difference - maybe a percentage point or two.

This was a few years ago already though.

Haha, that's an awesome compilation. Nice work!

I landed on HN from within the tech world (at the time, I'm largely no longer), though critical of the VC element of it. Circa 2011. The site then and now resembles (in both good and bad elements) Usenet and Slashdot of yore.

My sense then was a strong pro-Valley bias, though with some criticism. That criticism has been increasing here, as it has in the media generally ("techlash", "tech backlash", general antitrust / antimonopoly viewpoints, privacy / surveillance concerns, and the like).

I remain somewhat disappointed that there seem to be topics HN is unwilling (in the collective hivemind sense) to discuss, though that's more a case of unable to sensibly discuss than a specific moderator-led political bias. I've shared these frustrations with the mod team (mostly dang, though there are and have been others) over the years, sometimes agreeing, sometimes differing strongly.

Given the state of discourse elsewhere, in the press, in the US, in the rest of the world, and if I may suggest, in other online venues, Hacker News does not do badly. It may not be the best venue, but it is very, very, very far from the worst. Viewpoints do get heard (many of which I disagree with, or simply find based on ideological mantras or a failure to grasp fundamental facts). I try to express my disagreement as respectfully and clearly (and concisely) as possible. For the most part, that seems to be received reasonably well.

Accusations that HN is entirely of the Left, Right, techbro, woke, or other communities strike me as quite false. The culture quite likely skews US, educated, male, and wealthier, though other voices are heard. (I'd like to see more, I suspect HN would as well.)

I generally don't try to win given arguments at all costs, though I'll do my bit to nudge discussion in what I see as more interesting areas, both in terms of topics and insights on those. This includes comments, voting and flagging, submissions, and (where it seems advised) pinging the mods (hn@ycombinator.com) for egregious issues (mostly spam, titles, link disambiguation, and the like, though occasionally behaviour against guidelines).

> I generally don't try to win given arguments at all costs, though I'll do my bit to nudge discussion in what I see as more interesting areas, both in terms of topics and insights on those.

This is pretty much my approach.

I don't try to "win" arguments on HN at all. I try to provide my perspective when it differs substantially from the prevailing, but I try to do that in a way that's not confrontational. I rarely get comments in reply that are directed at me (as opposed to the idea I've presented), but when I do, I almost always just ignore them.

I have noticed a lot of the downvotong of things critical of YCombinator and it's bias against topics it doesn't like. I even have emails where the moderator or employee refuses to acknowledge or explain why I was downvoted and not others. It was eye opening that experience and reading this list doesn't surprise me. Everyone has an agenda.

What you're "noticing" is not accurate. HN gets tons of comments that are critical of YC and YC-funded startups. We don't censor those, and in fact we moderate less, not more, when YC is involved: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

Your description of those emails doesn't match my memory. I spent hours trying to explain things to you. That was not a refusal to explain.

Hours? It was like 6 or 7 emails. How many hours does it take you to write a 5 or 6 sentence email?

Don't be upset people are rightly criticizing Hacker News.

Edited: Nevermind, I guess HN users prefer to remain unaware of certain undocumented behavior.

To dang: I disagree with the no-delete policy (our comments are ours and we should have the ability to delete them), but more importantly I refuse to leave a comment up when my ability to reply has been disabled.

HN is an extremely authoritarian web forum, maintained this way solely to prop up the reputation of YC as an organization. You're not allowed to have certain opinions because it will make the platform look bad, but of course censorship goes against the core beliefs of the western world so they can't exactly be transparent about it.

The more experience I have with this site, the more I've found myself hating it. And the smugness of the users here who believe HN is the one-true-forum is the turd on top of the shitcake for me.

Oh boy - I have been here for 12 years or so, and I don't really agree with this take at all. Maybe I am biased, but I think the conversations here are more open and more frank than most (or any) other places on the internet.

The parent comment is already greyed out, so yes, maybe you're biased.

The parent comment made some bold assertions with no facts (anecdotal or otherwise) to back them up. I'm not surprised that some people feel that the argument is in bad faith.

Yes, it appears that way because of the way the moderation works. Comments that HN wants to be hidden just get sent to the bottom of the comment stack, then grayed out, then hidden. You probably believe that only comments that users downvote get sent to the bottom, but no, there is a secret rating system controlled by the mods that causes certain comments to behave as if they were already downvoted. It's an opaque system that is selectively applied so as to conceal its nature.

If the system were honest and fair, the mods would have no reason to hide their behavior.

Now I have to be careful, I've posted 3 comments in the last hour and I'm running up against my rate limit since the mods have highlighted me as a troublemaker. Can't say too much exposing the secret system of moderation here, they don't like that.

This sounds like conspiracy theory stuff. There's this secret bad stuff that I know about, but you don't because "they" hide the evidence.

So, how do you know? What is your objectively verifiable evidence that this secret system exists?

Or are you just referring to the editors' ability to kill a post that they deem to be against the site guidelines? They have that power, but it's not a secret.

It's not that hard to notice, there's just no official documentation or explanation of what these mechanics are and that's the extent of how "secret" it is. I think if you search dang's comment history he mentions it a few times, and obviously if you look for anecdotes you'll find plenty of users on this platform that have been rate limited, and deprioritized on other ways to have their opinions filtered out of the conversation.

Selective moderation is also a big problem on this platform. Rules only apply to certain groups of people, basically at dang's discretion.

The biggest problem with this platform is that there is no moderation log. The mod team works in an opaque manner, making it hard to prove or disprove the claims I was making. If we could just see a list of all comments dang had deleted, and the true vote count of each comment, we could all see exactly what the mod team was doing and be confident in its fairness.

That's obviously not the case though, and HN keeps this information from us to protect their image.

Authoritarian is a bit harsh. That implies the moderators delete and ban anything that goes against a particular set of narratives. I have not seen that to be the case. There are topics and opinions that the user-base will downvote and some topics that are likely to start flame wars. In the case of down-voting, it helps to know your audience. In such cases one could just create their own forum that caters to topics that don't resonate here. I've done that for several topics and kept it entirely separated from HN. It would be unfair of me to assume I can change the acceptable use policies of a site or the reaction of its audience based on my preferences.

If I may, allow me give you one unsolicited advice I've been slowly learning through the years:

Most things in life have their downsides and shortcomings. We should strive to minimize the impact of those and instead capitalize on their strengths. Take what good you can from HN and run with it. HN is great in many ways.

I disagree. I think HN is good in some ways, but awful in many other ways so much so that it overshadows the good. Unfortunately the good can trick honest people like yourself into thinking the platform as a whole is good, but hey, you're entitled to your opinion.

Neither dang nor hu3 went over to your place, pointed a gun at your head, and forced to to hang out on a website you hate. If you think HN is awful in so many ways, vote with your feet. Go somewhere you like better. Why hang out here to complain about here?

Because I have bad impulse control, and dang won't let me delete my account :)

Apparently once you sign up for HN they have the right to keep your soul locked up here forever.

I've asked for them to delete my account or permaban me for over a year, but users just don't have rights here and that's just the mentality that dang has always had.

I'm guessing this doesn't apply to HN jurisdiction?

https://en.wikipedia.org/wiki/Right_to_be_forgotten

https://news.ycombinator.com/item?id=26561884

https://news.ycombinator.com/item?id=23008642

https://news.ycombinator.com/item?id=22892517

How about you guys just implement an account deletion button and treat your users like actual people?

> Even if you send them an email, apparently they're "too backed up" to handle account deletion requests

I've searched and found no record of you emailing us. Did you mention the account name? We process these requests literally every day, so I'm not sure what you're trying to achieve with these claims.

I think it's pretty clear what I'm trying to achieve. I'm trying to get you guys to implement an account deletion button in user settings like you should have. I should not have to expose my personal email by contacting you in order to do so. How would you even know the request was legit? I have no email associated with my account.

There are many other ways to establish that someone owns an account.

"Trying to get you guys" to do something that we've explained in the FAQ that we don't do and why, is off topic, and repeatedly posting the same complaints and false claims in the threads is trolling. Please knock it off. If you don't like how HN works, you needn't use it. It's just an internet forum.

It's not off topic, this article is about HN's features and I'm discussing HN's features.

Just because you have something in your FAQ doesn't mean it's off-limits for discussion. You can post whatever you want there, and I can choose to disagree with it. It's one of many reasons why my disdain for this site grows with each passing day.

And they're not false claims. You guys absolutely do not have an account delete button, and you guys absolutely do deprioritize accounts and comments in an opaque manner to limit their exposure.

If you want me to stop complaining, stop using subterfuge to accomplish nefarious objectives. Be transparent about the moderation on this platform.

Please don't delete-edit your posts like that. It leaves the thread incoherent and isn't fair to the reader or to the commenters who replied.

This begs the question: why were these features undocumented, and left to an unpaid volunteer to weed out?

The combined valuation of the top YC companies was over US$300 billion as of January 2021. Certainly their news aggregator could get a little bit of love.

I’ve always seen these features as something akin to easter eggs. Stuff you learn over time as you become part of the “in” crowd.

FWIW, I've never liked this argument, although that's not why I initially made this list. The best case scenario from the community perspective is that it increases elitism, although companies like Snapchat went viral with teens for hiding away features.

Note that relatively few companies hide features for serendipity nowadays.

It isn't lack of love, it's that it isn't in HN's DNA to spell everything out. The site guidelines, for example, make no attempt to be comprehensive (https://news.ycombinator.com/newsguidelines.html).

The currency on HN is interest, not money, so 'unpaid volunteer' isn't the best way to frame this. If someone is interested enough to put together and maintain a list like this, that's in the spirit of the community in a way having an official list would not be.

(Btw we're all in favor of it and I believe I've sent Max some details to add to the list. If pages like this are going to exist—and they will, because online communities are endlessly interested in themselves—better that they be accurate.)

Re "better that they be accurate", I figured I'd check it again, since I haven't in a couple years. That led to the following errata, which I will send to Max. I thought I'd share them here too in case people are interested in this level of detail.

> it can be [dead] again at which point it can't be re-vouched

I don't think that's right? Pretty sure it can be re-vouched. I'd have to check the code to be sure.

> the [+x] number on the right

Since changed to [x more], to make it easier to click.

> will usually be at the top of the front page at that time

It would be better to say "on the front page". The "$Person has died" thread is usually at the top for a while, but we leave the black bar up for longer than that.

> the line between technology and politics is blurred

I've written a lot about how we approach this; it might be helpful to point to some of that. I usually direct people to these links first:

https://news.ycombinator.com/item?id=22902490

https://news.ycombinator.com/item?id=21607844

https://news.ycombinator.com/item?id=17014869

There's lots more at https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...

> The web button next to submissions

Removed in July 2020. I removed 'past' as well but there was more pushback about that so I put it back.

> what the front page looks like at any point in time

Not quite - it's a blended view of the front page for a 24-hour period, sort of like a weighted average of a bunch of snapshots. That turns out to be unsatisfyingly hard to get right; different algorithms produce very different blends, and none of them feels like a real front page. You can always look at https://web.archive.org/web/*/https://news.ycombinator.com to see specific snapshots.

> https://news.ycombinator.com/lists

We added one for whoishiring and one for Peter Roberts' immigration AMAs. I should also add the Launch HN list for YC startups. (Edit: added now.)

> All public URLS with user-generated content

I don't have a comprehensive list but the quantifier "all" there can't be right.

> I believe I've sent Max some details to add to the list.

For transparency, I'll confirm that dang sent over some things when I first wrote the list in 2018.

This does not seem to cover HN’s silent rate-limits, apparently for selected “troublemakers”: https://news.ycombinator.com/item?id=20997593

This is what seems to cause “You're posting too fast. Please slow down. Thanks.” to appear for up to two hours. Fittingly, I had to wait two hours to post this comment.

https://github.com/minimaxir/hacker-news-undocumented/issues...

Also https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Anti-spam/Anti-malicious-posters algorithms will, by necessity, never be transparent, nor even consistent or remain the same over time.

Fighting spam and bad-faith commenters is a cat-and-mouse game.

You’re twisting the issue by calling it “bad-faith commenters” though. Just take a look at this thread, where even the most polite, yet critical comments are downvoted and quite fittingly really hard to read.

It needs a section on muting. If you receive too many downvotes in a certain time frame you'll find yourself muted for a while and get a "you're posting too fast" message when you try to comment. This can be triggered well after the comments were made. For example, if you you have a discussion thread where you have four or five comments they can sit on 1 even overnight but if you come back the next day to find the tide has turned and all of them are in the negatives you'll probably be muted for a few hours.

This is why it can be smart to take the temperature of the crowd before replying to a thread. Your liability for a bad opinion or picking the wrong side scales linearly with the number of comments.

"Damn the votes, full speed ahead"

Posters should go full speed ahead if issue is with a topically relevant opinion, not the way it is expressed. We're all here for different reasons, but I like to presume we want to discuss a topic online.

Gaining Internet points or avoiding moderation penalties shouldn't be a priority. If it is, the result is sycophantic conformance. This makes for uninteresting discussions.

That's great up until you have something you really want to say on a post you're very interested in and find you're not allowed to say anything at all.

Self-censorship or outright muzzling. Pick your poison. I'd rather people speak their mind, even unpopular opinions.

Reply on news.ycombinator.com