https://news.ycombinator.com/front?day=2006-10-09
Anything earlier gives HN didn’t exist yet.
https://news.ycombinator.com/front?day=2006-10-09
Anything earlier gives HN didn’t exist yet.
The casual conversation between pg and sama on the top link is interesting. Also neat that Sam was like 17 at the time.
I didn’t know the “HN didn’t exist yet” error. That’s neat, because it implies that this page was intended to be a secret Easter egg, rather than it being a bug.
The last interesting tidbit is to go forward starting from that day, and pay close attention to how rapidly organic interest in the site went to zero. It’s astonishing that Startup News didn’t suffer the same fate; it’s the hardest part of launching something like HN.
I still 40% suspect that lots of the early activity was artificial, and that real users eventually supplanted the artificial activity.
pg once alluded to a “secret sauce”, and I think it was as simple as “keep the front page fresh and interesting each day,” which absolutely requires manual intervention. So the secret sauce seems likely to be something along the lines of “have a big backlog of interesting links, then automatically submit them and give them a few artificial upvotes each day.”
I’ll admit, I’m probably more obsessed with HN’s early days than I should be. But it’s such a pleasing example of Lisp proving its value that I can’t help it.
I used to run a stack of private torrent trackers. The first one I made came from the fact that the admin of another site ran off with all the funds and abandoned the users. I moved them all to my site. Then every time I set up a new site I had thousands of users I could message to let them know. On the final one I had 100K+ active users - a message to them was like having your own private army to do your bidding for you.