3 ms·

When I wrote "Don't Fly During Ramadan" (#3 post), I anticipated it'd get a few hundred hits from my friends if I was lucky. I almost didn't even post it to Hac

When I wrote "Don't Fly During Ramadan" (#3 post), I anticipated it'd get a few hundred hits from my friends if I was lucky. I almost didn't even post it to HackerNews, until a friend told me I should. ....Boy was I wrong. That aside, it's sad how many of the top posts are negative in some way, even if you exclude the death announcements. The exact count depends on whether you consider things like PG banning SOPA companies to be "negative" news, but only a small handful of the top 30 links are truly positive (eg, Light Table). It's also interesting how many (or few) of the posts are technical in their nature. I believe I count two that are completely non-technical (mine, and the other one about the TSA), and three or four that are about software that people have built (Meteor, etc.). The rest are "in the middle" - mostly about people important to the tech community (Dennis Ritchie, Aaron Swartz, etc.), or about political issues related to technology, but not talking about the technology itself in any depth. This pattern continues even if you look beyond the top 30. I wonder if this has changed over time as the HN community has grown - perhaps in the morning I'll try filtering out posts less than X days old and see how that affects the top results.

anuragramdasan 13y ago
I guess that is quite expected. I for one, do no care about LightTable or Meteor and haven't used either yet. So I probably wouldn't bother looking into such a link when it comes on HN. But on the other hand, a loss of a person important to tech community is something that probably affects everyone equally regardless of their technology preferences. I would only expect this pattern to remain this way.
ajani 13y ago
That'd be an interesting exercise. Here's the fetched dump: Might be quicker to use: http://lab.varun.io/hn_all_time_top_links/top.json http://lab.varun.io/hn_all_time_top_links/top.json
- chimeracoder 13y ago
  Thank you - you saved me having to bust out an ad-hoc web scraper. :)
derefr 13y ago
I think it's that software, even cool, amazing, never-done-before software, happens incrementally. So you see one post with a few votes announcing an MVP, another for v0.1, 0.2, 0.7, "finally 1.0", etc. Each small change is expected and there isn't really that much to see from one revision to the next. And the intial version, on its own, isn't really usable enough for anyone to immediately switch to it from what they've got, even if it does something cool, so it gets few evangelists. Light Table and Meteor are cases where that wasn't true; where the initial version came out of nowhere and did something amazing enough that people did immediately switch to it, and evangelize it. The problem is that, for most things, this sort of "Apple press event" release requires a level of development-in-secret which I don't think is possible for FOSS software developed by more than one person. On the other hand, really negative events--deaths, disasters, etc.--tend to be unexpected and happen all at once, so they get their impact (and "evangelism", in a sense) all network-effected together.
ehsanu1 13y ago
I wonder if this has changed over time as the HN community has grown - perhaps in the morning I'll try filtering out posts less than X days old and see how that affects the top results. Note that popularity in terms of votes should also be normalized with respect to traffic/users.
mbesto 13y ago
Well, interestingly enough, not one of the top posts was created earlier than 2 years ago. This is obviously due to karma inflation and slightly akin to saying that Avatar is the top grossing movie of all time (hint: it's not[1] when you adjust for inflation). It'd be cool to see a normalized version of this. If that's even possible... [1]-http://boxofficemojo.com/alltime/adjusted.htm http://boxofficemojo.com/alltime/adjusted.htm
- victoriap 13y ago
  Normalizing per total users or karma has also some side effects. The demographics of HN user base has most probably changed during the past years as well as number of interesting stuff produced also changed. Now there is more submissions more competition.

anuragramdasan 13y ago

I guess that is quite expected. I for one, do no care about LightTable or Meteor and haven't used either yet. So I probably wouldn't bother looking into such a link when it comes on HN. But on the other hand, a loss of a person important to tech community is something that probably affects everyone equally regardless of their technology preferences. I would only expect this pattern to remain this way.

ajani 13y ago

That'd be an interesting exercise. Here's the fetched dump: Might be quicker to use: http://lab.varun.io/hn_all_time_top_links/top.json http://lab.varun.io/hn_all_time_top_links/top.json

chimeracoder 13y ago
Thank you - you saved me having to bust out an ad-hoc web scraper. :)

derefr 13y ago

I think it's that software, even cool, amazing, never-done-before software, happens incrementally. So you see one post with a few votes announcing an MVP, another for v0.1, 0.2, 0.7, "finally 1.0", etc. Each small change is expected and there isn't really that much to see from one revision to the next. And the intial version, on its own, isn't really usable enough for anyone to immediately switch to it from what they've got, even if it does something cool, so it gets few evangelists. Light Table and Meteor are cases where that wasn't true; where the initial version came out of nowhere and did something amazing enough that people did immediately switch to it, and evangelize it. The problem is that, for most things, this sort of "Apple press event" release requires a level of development-in-secret which I don't think is possible for FOSS software developed by more than one person. On the other hand, really negative events--deaths, disasters, etc.--tend to be unexpected and happen all at once, so they get their impact (and "evangelism", in a sense) all network-effected together.

ehsanu1 13y ago

I wonder if this has changed over time as the HN community has grown - perhaps in the morning I'll try filtering out posts less than X days old and see how that affects the top results. Note that popularity in terms of votes should also be normalized with respect to traffic/users.

mbesto 13y ago

Well, interestingly enough, not one of the top posts was created earlier than 2 years ago. This is obviously due to karma inflation and slightly akin to saying that Avatar is the top grossing movie of all time (hint: it's not[1] when you adjust for inflation). It'd be cool to see a normalized version of this. If that's even possible... [1]-http://boxofficemojo.com/alltime/adjusted.htm http://boxofficemojo.com/alltime/adjusted.htm

victoriap 13y ago
Normalizing per total users or karma has also some side effects. The demographics of HN user base has most probably changed during the past years as well as number of interesting stuff produced also changed. Now there is more submissions more competition.