Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

So I made something that's pretty stupid but had fun doing it

hitlerhops.com

Post reply on HN
Print view View original

1–10 of 148 posts

So I made something that's pretty stupid but had fun doing it

#1

Post by cnlwsu »

I really wanted a project that I could use map reduce on and this idea came up amongst a group of friends as a joke.

Unfortunately the link graph was really stale when I ran it so it is not always correct. Also dates make it boring sometimes.

So I made something that's pretty stupid but had fun doing it
hitlerhops.com

Re: So I made something that's pretty stupid but had fun doing it

#2

Post by gdl »

It seems too eager to go through year and day pages, and doesn't always hit the optimal path:

  Mike Godwin -> Godwin's law -> Adolf Hitler
it gives:

  Mike Godwin -> 1956 -> 1889 -> Adolf Hitler
Also strange is that the "1956" it follows is in the text of, but not linked from "Mike Godwin".

Either way, I like the Web 2.0-style Hitler drawings :-)

Re: So I made something that's pretty stupid but had fun doing it

#3

Post by cnlwsu »

gdl wrote: post #2

It seems too eager to go through year and day pages, and doesn't always hit the optimal path: Mike Godwin -> Godwin's law -> Adolf Hitler it gives: Mike Godwin -> 1956 -> 1889 -> Adolf Hitler Also strange is that the "1956" it follows is in the text of, but not linked from "Mike Godwin". Either way, I like the Web 2.0-style Hitler drawings :-)

its because its using an older record of Godwin's law: http://en.wikipedia.org/w/index.php?title=Godwins_law&ol...

where it actually doesn't actually link to hitler. I would like to update the records but it takes forever to upload 5gb of entries into app engine :)

Re: So I made something that's pretty stupid but had fun doing it

#7

Post by cnlwsu »

thinkalone wrote: post #4

> but had fun doing it Well, that's more than a lot of people can say about certain projects! Want to give us a run-down of how you put it together?

Definitely, it was actually pretty simple once I figured out what I was going to do. I started from the link graph put together by Henry Haselgrove (http://users.on.net/~henry/home/wikipedia.htm) that I found when looking through the EC2 public datasets. I then had a few easy steps.

1) flip the link graph from outgoing to incoming, so from any page I can see what links to it.

2) I found all the distances and paths iteratively by exploding out from Adolf_Hitlers page. http://www.johnandcailin.com/blog/cailin/breadth-first-graph... and blogs like it were very helpful.

3) loaded the data into a large binary file that I divided into indexed parts that I compressed and uploaded to appengine to extract and load into bigtable (this took the most amount of time! both to run and to write the code to make it work)

4) ??

5) profit

Post reply on HN