Comment by sillysaurus | Hacker News Reader

Comment by sillysaurus | Hacker News Reader

One idea for jumpstarting a new HN-type site is to spider HNSearch, gathering the first 100,000 stories ever submitted to HN, along with comments. Then set up your site so that your frontpage is a doppelganger of HN's frontpage circa 2007. I.e. today your frontpage should look how the HN frontpage looked on August 7th, 2007.

That way there's (a) the appearance of activity, (b) a constant stream of interesting content on the frontpage, and (c) interesting discussion in the comments. Before long, new real users would start to participate, e.g. by replying to doppleganger comments. At that point, it's inevitable that the new site would start to get traction as long as those new users keep coming back, which they should because the frontpage is interesting.

This could only work if someone had the balls to actually deploy the currently-released Arc 3.1 version of Hacker News, though, rather than rolling their own version in Rails. There's nothing inherently wrong with trying to clone HN's featureset, but it's interesting to note that not a single one of the HN knockoffs successfully cloned HN's entire featureset. Most of them were a halfway implementation.

Anyway. Just a fun idea.

EDIT: I just stumbled across a dump of HN from April 24, 2008: http://rapidshare.com/files/3129266675/ycombinator-news-2008...

It contains a snapshot of the first 172,575 items (submissions/comments) and a snapshot of the profiles of the first 6,519 users.

Have fun! Maybe someone can use the data to put together a cool visualization or something.

EDIT2: Just to be clear, this idea is firmly tongue-in-cheek.

EDIT3: Statistics time! According to that snapshot, when HN was 558 days old there were 38,693 submissions and 133,882 comments. The snapshot claims there were only 6,519 users. That would be an average of 20 comments per user and 5.9 submissions per user.

1. None of the HN knockoffs can really clone Hacker News because, for one thing, it's missing everything in the original HN's code. pg mentions that there's a missing "secret sauce" that ties it all together. That, and you really need to have some sort of heavy involvement in the community before you try to start it, so you'll need someone of pg-level importance to start it. Specializations could be started by very knowledgable HN users in their field but very general implementations would need an all around "ideologue." EDIT: I don't know if you're already aware of this, I'm just explaining because the way you wrote it appears as though you may think the Arc code is packaged with a full implementation of HN.

2. I don't think this would be allowed via the HN Search API...that's a lot of content.

3. This is a cool idea, but how would you get people to participate and add content to what is essentially a ghost, rather than just come back here, where there's news? I also don't think it solves the problem HN is currently facing. It merely puts a bandaid on it until we get back up to 2013 level volume.

I just ran across an old dump of HN from April 24, 2008. It contains the first 172,000 items (stories+comments) and the profiles of the first 6,500 users. http://rapidshare.com/files/3129266675/ycombinator-news-2008...

I would guess the secret sauce is spam filtering code. It seems conspicuously absent from Arc releases. The Arc code does come bundled with the HN source code, though, and all of the core features seem to work fine. There's nothing missing except some way of filtering spam.

The point for someone doing this would presumably be to hope the visitors in question would not notice that they're adding content to the ghost/doppelganger comments.

The secret sauce is that HN is the news aggregator of a successful start up combinator. The secret sauce is popularity.

The secret sauce is green.

I think a HN clone might get better traction not necessarily focusing on hackers/programmers, who likely are already perfectly happy being here. Maybe make one for all the stories and things that HN users don't want to see, or for different communities altogether.

I'm not sure why you think there would be (a) the appearance of activity, (b) a constant stream of interesting content on the frontpage, and (c) interesting discussion in the comments on August 7th, 2007.

Also, HNSearch's API won't allow you to gather that many stories.

The best way to do this would be to only take stories with years in the title. These are stories which were old during their original submission, but evidently timeless.

Reply on news.ycombinator.com