Really cool! May be a bug: I tried Martin Luther Ling -> Elon Musk and it showed me a link from MLK -> Mark Zuckerberg -> Musk. But I couldn't find a Zuckerberg link from the MLK page.
https://www.sixdegreesofwikipedia.com/?source=Martin%20Luthe...
Show HN: Six Degrees of Wikipedia
191–200 of 354 posts
Re: Show HN: Six Degrees of Wikipedia
#192This is a lot of fun. Thanks for doing this.
Re: Show HN: Six Degrees of Wikipedia
#193This is hilarious, from 'Riemann manifold' to 'Grindcore': 196 paths with 5 degrees.
[0]: https://www.sixdegreesofwikipedia.com/?source=Riemann%20mani...
Re: Show HN: Six Degrees of Wikipedia
#194Parquetry -> Romeo and Juliet, 46 paths ( https://www.sixdegreesofwikipedia.com/?source=Parquetry&targ... ). None through Shakespeare. Parquetry -> Tromeo and Juliet, 508 paths ( https://www.sixdegreesofwikipedia.com/?source=Parquetry&targ... ). All through Shakespeare.
508 paths, nice. I was about to post Wake in Fright -> Las Vegas, which has an amazing 414 paths, but yours is a winner. https://www.sixdegreesofwikipedia.com/?source=Wake%20in%20Fr...
https://www.sixdegreesofwikipedia.com/?source=Frank%E2%80%93...
This isn't really the best measure though, because it only counts # of paths at the minimum depth level.
edit: although I found some deep searches with very few links: https://www.sixdegreesofwikipedia.com/?source=Frank%E2%80%93...
Re: Show HN: Six Degrees of Wikipedia
#195Really cool! May be a bug: I tried Martin Luther Ling -> Elon Musk and it showed me a link from MLK -> Mark Zuckerberg -> Musk. But I couldn't find a Zuckerberg link from the MLK page. https://www.sixdegreesofwikipedia.com/?source=Martin%20Luthe...
It uses a pre-downloaded database of links [1], so presumably certain pages that were once linked have been updated.
[1] https://news.ycombinator.com/item?id=16469427
Re: Show HN: Six Degrees of Wikipedia
#196Earlier quoted context omitted.
Why not use a graph database like Neo4j instead of SQLite? This seems like the perfect use case. Is it because of the resources required to run one versus SQLite?
I actually had a friend suggest it to me and the Neo4j docs happen to be one of the many tabs I currently have open. I was already so far into using SQLite for this project and I wanted to ship it, so I decided to stick with what I had. I would be interested to see how Neo4j performs with such a big dataset (the resulting SQLite file is around 9 GB with nearly 6 million nodes and 500 billion links). I was a bit worri…
I've used neo4j before and it likes to consume a lot of memory.
Re: Show HN: Six Degrees of Wikipedia
#197In case you receive the following message
"Sorry internet hipster, this little side project requires JavaScript."
here is a quick example of how to get the pages the "traditional way"[FN1]:
#/bin/sh
test $# -eq 2||exec echo usage: $0 source target;
exec curl -H"Content-type: application/json" \
-d '{"source":"'$1'","target":"'$2'"}' \
https://api.sixdegreesofwikipedia.com/paths \
|exec sed '
s/\",/\"\
/g;s/,\"/\
\"/g;s/:{/:\
{/g;s/}/&\
/g;s/\"pages\":/&\
/'
FN1. The original "web browsers" needed no GUI, no Javascript.
Re: Show HN: Six Degrees of Wikipedia
#198Could you make it possible to just test with the fun suggestions that scroll through?
Just implemented this with a quick hot fix[1]. Hopefully I didn't break anything else in my hurry to get out.
[1] https://github.com/jwngr/sdow/commit/6e42e06488a592784e5d3d2...
Re: Show HN: Six Degrees of Wikipedia
#199Really cool! May be a bug: I tried Martin Luther Ling -> Elon Musk and it showed me a link from MLK -> Mark Zuckerberg -> Musk. But I couldn't find a Zuckerberg link from the MLK page. https://www.sixdegreesofwikipedia.com/?source=Martin%20Luthe...
I get a similar case with Donald Trump - Banjo Kazooie
https://www.sixdegreesofwikipedia.com/?source=Donald%20Trump...
Re: Show HN: Six Degrees of Wikipedia
#200About 12 years ago, some colleagues (Yehuda Koren and Chris Volinsky) and I proposed a technique for measuring proximity in quasi-random (social) networks, that can handle out-of-memory databases, and more than two query nodes, also finds a visualizable subgraph that represents as much of the relationship as possible using dynamic programming. It is described here (includes some figures): http://web2.research.att.com…
Thanks a lot for sharing! BTW, I just fixed the confusing interaction with the text input placeholders[1].
[1] https://github.com/jwngr/sdow/commit/6e42e06488a592784e5d3d2...