Hacker News Reader

Comment by chrisma0 | Hacker News Reader

Definitely a good idea to be able to specify the kind of topic areas that one would be interested in. Also good to restrict it to more "quality" articles. I can see spending a few hours reading articles :)

Anyone know how Wikipedia's own "Random article" function (https://en.wikipedia.org/wiki/Special:Random) selects articles? Truly completely random with no quality aspect?

View on HN

Hi, thanks for the feedback! Currently the topics are picked from https://en.wikipedia.org/wiki/Wikipedia:Vital_articles/Level... which are all curated articles and most of them should be interesting.

But I am planning to expand by adding more articles that have greater than 10K words. I am using word count as a rudimentary metric to qualify articles initially. But I will probably require something a bit more nuanced eventually if I have to include all of wikipedia's articles.

> Anyone know how Wikipedia's own "Random article" function (https://en.wikipedia.org/wiki/Special:Random) selects articles? Truly completely random with no quality aspect?

Here's how it works:

- https://github.com/wikimedia/mediawiki/blob/addb6649503c45f0...

TLDR: random, no quality aspect

Thanks for those links! From the code comments: "it won't cause any more bias than what the page_random scheme causes anyway. Trust me, I'm a mathematician. :)" Git blame says the author was "Ilmari Karonen" (https://commons.wikimedia.org/wiki/User:Ilmari_Karonen), 15 years ago!

Reply on news.ycombinator.com