A favorite more recent fact about this of mine: You might have heard about reddit's several early influxes from other websites as metaphorical eternal septembers, but did you know they had a literal eternal september just over a year after the release of the official mobile app? Check out the inflection point: https://subredditstats.com/r/teenagers
If you're referring to the sharp drop at the end of the graph, no. It's an artifact of the api change making it much harder to capture comments. We can verify this by how the comments/day rank didn't budge, because it happened on literally every sub simultaneously. If you're referring to the gradual dip from its peak, I have no idea.
Since the blackouts last year and the recent IPO, it feels like astroturfing and spam have increased on Reddit, while quality contributions have decreased. All usage metrics are up according to Reddit's IPO filings, but it feels like engagement is actually down, or at least lower quality. Many niche subs feel like ghost towns now.
Is this just my subjective impression or do you feel the same?
(anecdata for n=1:) Everyone I personally know who contributed meaningfully to subs I was active in has moved on to other sites. What used to be places sprawling with activity are still places sprawling with activity… but it’s mostly reduced to reposts and posturing and especially posting photos of the thing you bought (I guess consumers can’t do much more).
This has been a trend before the blackouts but it feels like those mark the point where it all broke down.
Many of us who were writing all the decent comments there left after they made clear they want it to be another bland nothing of a website during the blackouts. A hundred dull image macros a day aren't worth wading through for the few nuggets of decent content, so loads more left or majorly cut back. Some communities such as javascript are now dead.
I love all the people who said "Oh, no, generative AI will flood the online world with crap", as if generative natural intelligence hadn't already done so: https://en.wikipedia.org/wiki/Sturgeon%27s_law
I think that while the signal/noise ratio on the internet is much lower when I joined as a wee person in the 1980s, we are simply not prepared for the poisoning of the commons that is coming.
The people who began circling wagons in the late 1990s creating invite-only “smart friends” stealth communities will turn out to have been bang on the money.
It’s probably not too late to start - but which platforms other than private email lists and RocketChat exist for the endless hunker-down?
We do pretty well with a free-tier Slack instance. They only keep a rolling 10k messages or so, but this seems to be enough for more than 100 friends and colleagues to keep comm lines open.
LinkedIn could be a model for making vouch-networks as we fight the evil bots(or more accurately the evildoers who abuse neutral bots)
IMO, an important part of HN culture that has allowed it to mostly avoid Eternal September is that this is not just a strict news site: old and well-known things can be posted and discussed again, thus introducing them to the newer crowd without making them feel bad about not already knowing something.
Interesting how you look at Eternal September from the newb's angle. As if "making them feel bad" (or unwelcome?) is the main problem of ES.
I never looked at it that way. Always looked at it from the disgruntled Unix beard view that I have entrenched myself in. "get off my lawn".
While also lamenting how new generations keep reinventing solutions that "we" have proven to not work decades ago. Or reinventing solutions to problems that have long been solved.
When the solution is not to shout "get off my lawn" harder, but to go out there, with empathy, and help the newer generations find their way around in the history and practices of ICT and CompSci so they too can reach out to that experience and knowledge when confronted with a problem (rather than blindingly running towards a cliff that "we" the seniors fell off several times already)
The problem with Eternal September isn't that n00bs are stupid lusers - but that we do a terrible job introducing people into our culture.
So many online spaces seem to have an attitude of "if you don't already know the rules, you can't play". There's no helping newbies and neophytes learn the how and why of a community.
Most organisations are desperate for new members. And yet, when people show up, they're laughed at, blocked, or sent away.
Eternal September should be a cause of joy! Look! All these new people want to talk about our hobby! New friends to play with! New people we can tell our jokes to! New blood to help give us new ideas!
Instead, they're treated as an annoyance or irrelevance. Why is that?
Because a sudden influx of new users tends to change the culture of an existig community. People don't like that. I know I was shunned for making low effort comments when first arriving here from Reddit, and in retrospect I'm glad that I wasn't coddled, I wasn't made to feel welcome, I had to learn the rules and now this is my favorite corner of the internet, because HN changed me, not the other way around.
I think we can differentiate between rules that are in place to uphold a certain standard, and those that stop change from happening at all.
The latter is guarantee for obsolence. The former a way to allow change, but not have that change affect the core of the community.
E.g. at HN, deep conversations are core. So it makes sense to protect that with rules on friendlyness, effort and topics.
But it's always a thin line and a spectrum. Something many (software) communities or community-leaders have a hard time with. So they go for absolutism. And often don't allow any change.
The worst reason for (not) doing something is "we've always done it this way". Yet in many (software and FLOSS) communities that is exactly what is upheld. Keeping out new influences but also keeping out fresh members, ideas and users.
He never suggested that the influx of newbies should change the existing culture, precisely the opposite. The existing culture doesn't do a good enough job of helping onboard the newbies in a gentle way that integrates them. Instead, there's just indifference to this process.
> Instead, there's just indifference to this process
kind of. There are two things that often happen if people don't integrate, mostly wrt company culture:
1. The newbies don't integrate because they have created their own culture because either they don't realize that they are excluded from the "adults table" (different departments, etc) or that their experiences interacting with the existing hierarchy makes them want to reject the existing culture. They surround the "core" even though they may not often interact with it (and maybe management realizes this but most often not). Maybe they do some shadow IT but the projects that they work on are not core to the business because they are locked out from those systems and the work that they do is seen as useless because it is never integrated (maybe their code kind of sucks because they don't get mentorship, etc). The core is mostly untouched by the social group(s) that surround it. The different social groups are not as rigid as departments but if the core has any legitimate power the non-complaint people are usually shuffled around. Some people in the "wrong-group" may learn to respect the existing culture and they are promoted into the sacrosanct.
2. Similar to #1 except that the newbies are better at self-marketing or socializing with management and eventually the core culture is entirely replaced by the newer culture. The business allows this, even perhaps risking business continuity, because they start to see the old clique as stubborn, not inclusive, not open to change or suggestions, perhaps even a liability to the company.
How does HN handle its Eternal September so well? Discussion quality is still great and dang does a great job and introducing newcomers to the HN rules & culture.
I agree, he does a great job. But the last couple of years I've noticed a couple of worrying trends.
First, there has been a huge surge in activity. Back in the day, 50-100 comments on a post felt like a big thread. Now we regularly see threads that are 500+ comments, sometimes even more (the recent one on the xz back door has 1,700 comments). Because of how HN paginates long threads, this encourages new commenters to reply to the top comment rather than adding new top-level comments. (In some threads, the entire first page is just the top comment and its replies.) In short, I'm not sure HN can sustain that many comments in terms of the reading and discussion experience. As a possible improvement, I'd like to see big threads having some way of "aging out" older sub threads and promote new comments. But at that scale, how do you even follow a discussion? HN doesn't have a "highlight new comments since last visit", for example.
Secondly, there's been a significant uptick in lower-quality, low-effort comments. The regulars (some of whom are so active that I don't fully understand how they can be productive in their day job) are still keeping the quality high, and HN is still often the best place to get commentary that is superior to the article being commented on. But the quality drop is still noticeable, and I see an increase in comments that are low effort (memes, jokes, "lol", "this tbh" etc.), promote conspiracies, and so on. I suspect that with time, it's going to be increasingly harder to control the influx from other sites like Reddit without even more diligent supervision.
> but that we do a terrible job introducing people into our culture.
If you're a community of 100 people, and suddenly you get an influx of 10000 people, then there is no "introducing people to our culture" anymore. Then you got a "introducing you to the newcomers' culture".
They won't listen to your in jokes, or even care about understanding them (they might prefer to just be offended for not getting them). They won't listen to the carefully thought-through stuff already in the community, but rather shove some hype in. It is lower-effort to keep doing what you did, to keep following some hype that got you where you are, so unless you are basically forced to consider something else, you'll instead force the oldies in the community you are flooding to follow your suit instead - by sheer mass.
I'm familiar with this term for over a decade and also experienced it during the 90's and I just had the realization that Canada is experiencing something similar right now with influx of immigrants, asylum seekers and temporary foreign workers. I'm pessimistic about the future and sad about the loss of the culture. I still use Usenet but it's not the same, Canada will continue to exist but it will be very different country .
The software dev world also had its eternal september about 15 years ago I'd say. Since then we're even more being flooded with bullshit hypetrains and all.
30 comments
[ 2.9 ms ] story [ 80.9 ms ] threadIs this just my subjective impression or do you feel the same?
(periodic reminder: https://search.marginalia.nu )
The people who began circling wagons in the late 1990s creating invite-only “smart friends” stealth communities will turn out to have been bang on the money.
It’s probably not too late to start - but which platforms other than private email lists and RocketChat exist for the endless hunker-down?
LinkedIn could be a model for making vouch-networks as we fight the evil bots(or more accurately the evildoers who abuse neutral bots)
When the blooming of the perennials no longer moves you, perhaps it's time to take a walk outside the garden ;-)
I never looked at it that way. Always looked at it from the disgruntled Unix beard view that I have entrenched myself in. "get off my lawn".
While also lamenting how new generations keep reinventing solutions that "we" have proven to not work decades ago. Or reinventing solutions to problems that have long been solved.
When the solution is not to shout "get off my lawn" harder, but to go out there, with empathy, and help the newer generations find their way around in the history and practices of ICT and CompSci so they too can reach out to that experience and knowledge when confronted with a problem (rather than blindingly running towards a cliff that "we" the seniors fell off several times already)
So many online spaces seem to have an attitude of "if you don't already know the rules, you can't play". There's no helping newbies and neophytes learn the how and why of a community.
Most organisations are desperate for new members. And yet, when people show up, they're laughed at, blocked, or sent away.
Eternal September should be a cause of joy! Look! All these new people want to talk about our hobby! New friends to play with! New people we can tell our jokes to! New blood to help give us new ideas!
Instead, they're treated as an annoyance or irrelevance. Why is that?
The latter is guarantee for obsolence. The former a way to allow change, but not have that change affect the core of the community.
E.g. at HN, deep conversations are core. So it makes sense to protect that with rules on friendlyness, effort and topics.
But it's always a thin line and a spectrum. Something many (software) communities or community-leaders have a hard time with. So they go for absolutism. And often don't allow any change.
The worst reason for (not) doing something is "we've always done it this way". Yet in many (software and FLOSS) communities that is exactly what is upheld. Keeping out new influences but also keeping out fresh members, ideas and users.
kind of. There are two things that often happen if people don't integrate, mostly wrt company culture:
1. The newbies don't integrate because they have created their own culture because either they don't realize that they are excluded from the "adults table" (different departments, etc) or that their experiences interacting with the existing hierarchy makes them want to reject the existing culture. They surround the "core" even though they may not often interact with it (and maybe management realizes this but most often not). Maybe they do some shadow IT but the projects that they work on are not core to the business because they are locked out from those systems and the work that they do is seen as useless because it is never integrated (maybe their code kind of sucks because they don't get mentorship, etc). The core is mostly untouched by the social group(s) that surround it. The different social groups are not as rigid as departments but if the core has any legitimate power the non-complaint people are usually shuffled around. Some people in the "wrong-group" may learn to respect the existing culture and they are promoted into the sacrosanct.
2. Similar to #1 except that the newbies are better at self-marketing or socializing with management and eventually the core culture is entirely replaced by the newer culture. The business allows this, even perhaps risking business continuity, because they start to see the old clique as stubborn, not inclusive, not open to change or suggestions, perhaps even a liability to the company.
This happens with open source projects as well
First, there has been a huge surge in activity. Back in the day, 50-100 comments on a post felt like a big thread. Now we regularly see threads that are 500+ comments, sometimes even more (the recent one on the xz back door has 1,700 comments). Because of how HN paginates long threads, this encourages new commenters to reply to the top comment rather than adding new top-level comments. (In some threads, the entire first page is just the top comment and its replies.) In short, I'm not sure HN can sustain that many comments in terms of the reading and discussion experience. As a possible improvement, I'd like to see big threads having some way of "aging out" older sub threads and promote new comments. But at that scale, how do you even follow a discussion? HN doesn't have a "highlight new comments since last visit", for example.
Secondly, there's been a significant uptick in lower-quality, low-effort comments. The regulars (some of whom are so active that I don't fully understand how they can be productive in their day job) are still keeping the quality high, and HN is still often the best place to get commentary that is superior to the article being commented on. But the quality drop is still noticeable, and I see an increase in comments that are low effort (memes, jokes, "lol", "this tbh" etc.), promote conspiracies, and so on. I suspect that with time, it's going to be increasingly harder to control the influx from other sites like Reddit without even more diligent supervision.
If you're a community of 100 people, and suddenly you get an influx of 10000 people, then there is no "introducing people to our culture" anymore. Then you got a "introducing you to the newcomers' culture".
They won't listen to your in jokes, or even care about understanding them (they might prefer to just be offended for not getting them). They won't listen to the carefully thought-through stuff already in the community, but rather shove some hype in. It is lower-effort to keep doing what you did, to keep following some hype that got you where you are, so unless you are basically forced to consider something else, you'll instead force the oldies in the community you are flooding to follow your suit instead - by sheer mass.