Dan's stats are hard to audit because not every project has contributors squash commits before they merge, so some of his counts are likely pessimistic because he's counting intermediate states that were never the true state of master. On Mar 15, 2015 1:35 PM, "Johan Larson" wrote:

Dan Luu recently dug up some build performance data about various open source projects, and found that many such projects have broken builds more than half the time.

http://danluu.com/broken-builds/

Unfortunately Luu's data set doesn't include Haskell. Does anyone have the facts on this?

(A brief search didn't turn up anything.)

-- Johan Larson -- Toronto, Canada _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe