Manuel Bärenz

> My one request would be an update to the documentation here: https://monad-bayes.netlify.app/usage/#population Yes, on it. > but would be nice to check if it's straightforward for you to update...

> Weighted is the State monad. (By the way, it's possible that one of the optimized Writer or accum monads could be used here, but that's for a separate issue)....

> I suggest we first understand better how this proposed change affects the existing inference algorithms, not just in runtime, but in the actual samples they compute, both by testing...

True. This should be taken into account when fixing https://github.com/tweag/monad-bayes/issues/260.

CC @idontgetoutmuch @reubenharry

`cabal-docspec` unfortunately is hard to install in a sustainable way. See https://github.com/phadej/cabal-extras/issues/135. The only way I can think of (apart of building from source) would be pinning a version from...

It seems `cabal-docspec` was added as an experiment in https://github.com/tweag/monad-bayes/pull/161, so I'm of a mind to remove it, since it was never tested on CI (and `haskell-nix-utils` seem unmaintained).

Can you, instead of editing in place, copy the file `Sampler/Strict.hs` to `Sampler/StrictFu.hs` maybe? And then you have two sampler types, both of which you can import in the benchmark,...

If you add a type class to the same module, the type class dictionary lookup will be optimized away by GHC. For it to have a performance impact, you have...

I did not check my statements against Core, so one should maybe look into the optimized core first to make sure that this is what's happening.