I recently read his (first) paper [1] on the Spineless Tagless Graph Redunction Machine (STG), which forms one of the lower-level backends of the main Haskell c

I recently read his (first) paper [1] on the Spineless Tagless Graph Redunction Machine (STG), which forms one of the lower-level backends of the main Haskell compiler GHC. After understanding the paper well enough to implement it from scratch (without any further literature), I have to say, albeit some 25 years late, how impressed I am by both his research and his ability to convey it. If you're interested in understanding how Haskell evaluation is done, I cannot recommend the paper enough. [1]: http://research.microsoft.com/apps/pubs/default.aspx?id=67083 http://research.microsoft.com/apps/pubs/default.aspx?id=6708...