Hacker News new | past | comments | ask | show | jobs | submit login
It's moot now. September 1993 will go down in net.history as the September that never ended. (wikipedia.org)
24 points by iamelgringo on Feb 24, 2008 | hide | past | favorite | 14 comments

Eternal September has become a pattern of mass interaction, especially on the net. It's why we have News.YC, after all: the Eternal September of Reddit.

The problem is fundamentally content overload, but that goes hand-in-hand with decrease in average quality. Quality, in an absolute sense, may increase, but it's overwhelmed by ignorance and mediocrity. So there needs to be filtering on the content lifecycle.

Filtering right up front is very effective, but bad for business. Take Google Answers. You had to pay money (mostly) to ask a question and few were allowed to answer them. Extremely high average quality content for what was there, but where are they now? Compare that, of course, with its polar opposite Yahoo Answers. It was never even August there!

But the internet's native model is to eliminate any constraint on generating content. That's why Eternal September was so grievous -- it's when that philosophy got mugged by the reality of induhviduals, spam, crazy people and the sheer volume of mundanity.

It was also when people realized that the potential of the internet was rapidly approaching. But okay, given Eternal September as a reality of content generation (don't try to set up walled gardens), what do you do? You bust out the best tools you can for dealing with it -- the PageRank, the groups, the tags, the voting, the social connections. And people seem to enjoy the internet again. And I think some money was made along the way.

So is it more interesting that Eternal September was the end of the golden age for the elite or the birth of a golden age for the mainstream?

I think a migration cycle in the nerd sites will decrease the problem - effectively establishing equilibrium.


Crazy idea - but perhaps news.ycombinator should have limited registrations. I wonder what the outcome of that would be.

Leaving aside the question of whether filtering users upfront is desirable, you don't need to simply limit registration. There are a lot of intermediate steps you can take.

MetaFilter is a classic example of reasonable but comparatively high barriers to registration: $5 for an account and you have to wait a week to post. Seems very clever, but I'm not a MeFi regular, so I'm not sure how well that's worked out.

I think paid barriers to entry just amplify the positive or negative effects of other administrative measures; in and of themselves, all they do is reduce the rate of people coming in without much affecting its make-up. It doesn't even work that well for filtering out teenagers any more with the advent of prepaid gift cards (and before that they could just ask the 'rents to stump up the cash anyway).

I say that paid barriers don't do that much good or bad themselves based on the fact that on MeFi they seem to have been an asset whereas on, say, Something Awful they seem to've just turned the place into an echo chamber, and on Kuro5hin all it's done is drive the site further into stagnation. Paid barriers aren't risk-free either; it makes account gaming even more fun and makes your site a juicier target for attacks, especially if it means you end up storing people's personal details.

By restricting registration, you cut off your community from potentially beneficial new members and ideas. While news.yc wants to keep the news quality up, I don't think anyone wants to turn it into an elitist club of people who 'got in on the ground floor'.

What if we made them pass a test before being able to register?

I think this is a very relevant article to the "Eternal September" story:


What's surprising is that some of the group dynamics that contribute to the break down of online communities was identified in the oddest of settings decades ago...

Seems to me the problem of news.yc, if there is one, so far would not be too many irrelevant submissions, but rather, not enough interesting submissions. Can we really expect dozens of stellar submissions every day, though? Maybe there are not even that many interesting new things relevant to hacker news happening in the world every day.

Ironically, this post refers to itself.

Of course "the September that never ended" ended two years ago. So this post appears to be not only a non sequitur but also self-contradictory.

"Of course "the September that never ended" ended two years ago."

How so?

Fair point. It was actually three years ago:

"On February 9, 2005, AOL discontinued newsgroup access through its service (this was announced on January 25, 2005[6][7])."

I actually found that comment a little ironic, because I thought that AOL itself died pretty much when they bought Time Warner.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact