[Haskell-cafe] Alternative STM implementation

Daniel Waterworth da.waterworth at gmail.com
Sat Dec 10 13:28:32 CET 2011


This morning I have written an STM implementation as a learning
exercise. I thought I'd share it here to get some critique and because
it might be useful to others (consider it as being in the public
domain). I haven't tested it thoroughly, so there may be edge cases
that haven't been thought through. It should be a drop-in replacement
(so long as you only use TVars), if anyone has any STM benchmarks
already made, I'd love to see how it compares. Here it is:




More information about the Haskell-Cafe mailing list