Wouter Nuijten

Thanks for your PR and the impressive attempt at solving this issue! I think there might be an easier way to go about this. The way in which it is...

I think if you interpolate `fform` into an expression, you'll still get the actual type! The mechanism that does this in the `@model` macro is here: https://github.com/ReactiveBayes/GraphPPL.jl/blob/2f7cd70a58be2e47ea7d1932cc811ad602fb8f28/src/model_macro.jl#L640-L657 You'd really only...

I think this is a lot more like what I meant indeed! I will do a more thorough review tomorrow, but I think this is 90% along the way in...

Coming back to this PR, I see some of the tests are failing? There is a merge conflict I will solve but apart from that I think we still need...