Kai

@jwiegley > The only flaw in hdevtools that I can see now, compared to just using ghc, is that I cannot type check against modules in a Cabal project that...

@jwiegley Works for me. Can you provide a testcase?

Some warnings are not shown when compiling with -fno-code http://ghc.haskell.org/trac/ghc/ticket/8101

Try out https://github.com/estatico/scala-newtype/blob/master/README.md which is a similar, but much more fleshed out lib

@lJoublanc AFAIK NewSubType encoding there is at least equivalent to one in this library. You may want to look at https://github.com/typelevel/machinist for something that definitely avoids boxing (via macros)

The project in question: https://github.com/7mind/izumi, [build.sbt](https://github.com/7mind/izumi/blob/develop/build.sbt) is now generated by [sbtgen](https://github.com/7mind/sbtgen/) [script](https://github.com/7mind/izumi/blob/develop/project/Deps.sc)

@adamgfraser Assuming the semantic of `.onInterrupt` from its scaladoc, ``` /** * Runs the specified effect if this effect is interrupted. */ ``` In the example, `this effect` was not...

@adamgfraser Interruption being the exclusive cause of failure is not the same thing – you could have additional `Die` nodes from finalizers even if a fiber is really interrupting due...

@adamgfraser The presence of `Die` nodes does not mean that you didn't fail exclusively due to interruption, though. `Die` nodes can appear after interruption if subsequent finalizers fail, but the...

What I'm trying to say overall is: 1. `Cause` does not contain enough information to know if 'this fiber' was the fiber that was interrupted. 2. Without that information CE3...