Hacker News Reader

WTFs per comment would seem to me like a much more interesting measurement. Eg. I'm pretty sure CSS's low rating has more to do with the typical sparsity of comments than the awesome sensibleness and consistency of CSS renderer implementations.

Edit: Also takes no account for the quantity of code in each repository. Languages that are used for large projects are at an automatic disadvantage.

View on HN

I agree, but I have no idea how to get that information out of github, aside from perhaps iterating over every project there.

I think it would be cool if they made some aggregate data like that available though. =)

Ah, I commented in the comments of the article itself that CSS was surprisingly low given the existence of IE6. This may explain it.

Reply on news.ycombinator.com