3 ms·
As I was not there when all this happened, I decided to investigate and find out more about what it was like back then. I was surprised to find that the group o
by solipsist 16y ago
As I was not there when all this happened, I decided to investigate and find out more about what it was like back then. I was surprised to find that the group of people who truly affected HN the most after this article was the old-timers and not the new "digg-like users"[1]. Users reacted as if this was the end of HN and immediately grouped all new users into one that many despised. If you want, you can look for yourself: http://news.ycombinator.com/item?id=133440 http://news.ycombinator.com/item?id=133440
The new users probably did not create as much trouble as people had thought they would. Rather, the older users created the high tensions between the two groups. This all led to excessive discussions that has been prolonged for years. Unfortunately, this article has inspired many people to continue doing so (my bad!). All of this small-talk is what's really been hurting HN.
For those that want to compare HN before and after TechCrunch's article, here is what the homepage looked like:
Before: http://web.archive.org/web/20080308054301/http://news.ycombinator.com/ http://web.archive.org/web/20080308054301/http://news.ycombi...
After: http://web.archive.org/web/20080312014516/http://news.ycombinator.com/ http://web.archive.org/web/20080312014516/http://news.ycombi...
A lot of people here on HN (pg included) agree that the site quality has gone down a little over the past few years. As this is probably the result of the number of users and we know that TechCrunch has contributed to that number, we can safely say that the two are somewhat correlated. However, the heated discussions and behavior that has come as a result are probably more worrisome than the new users themselves. Just something to think about.
[1] - http://news.ycombinator.com/item?id=133446 http://news.ycombinator.com/item?id=133446
- Locke1689 16y agoI'm not sure the trouble is that new users would create "trouble" as originally foreseen, but that new users represent a wider area of the population. For a site which aims to a high standard of discussion, regression to the mean is just as dangerous as "trolling."