[Haskell-cafe] Why did the number of LOC in GHC nearly triple in 2014!?

Hello folks, I was looking at this: https://www.openhub.net/p/ghc/analyses/latest/languages_summary And was really surprised by the graph. Is this correct!? Did the complexity of GHC's codebase significantly increase in 2014? Can anybody shed a light on why that is? Cheers, Dimitri -- 2E45 D376 A744 C671 5100 A261 210B 8461 0FB0 CA1F