Yeah, the pie chart doesn't quite cover it - I'd like to see both swear words per commit per language (if, say, Java has 10% of the swear words but 3% of the commits) and complexity of the swear words - a simple "Fuck" implies far less frustration than a "Motherfucking Cocksucker!"

Could develop quite a nice Programming Language Pain Index…

From what I remember there was only one or two "motherfuckers".

I will post up some more data if anyone is interested.

I run a slang dictionary website which lets users assign an offensiveness score to each term. That would be an interesting bit of data to add: not only the raw word count, but how offensive the swearing is for each language.

(For sample data, the 100 most vulgar words on the site are in a table here: http://onlineslangdictionary.com/lists/most-vulgar-words/ )

If it's legal, it would be awesome of you to make this commit message dataset available on Infochimps or something.

i'd say alter the list of swear words in general to a list possibly more tuned to programming. agreeing with above, word breakdown would be nice as well.

actually, maybe just a github swear browser.