Hello,
well, yes. But {F<:Function, G<:Function} are “positional parameters” one and two in the type definiton while the struct has six fields. This confuses me.
The manual has only examples where all fields hvae (the same) parametric type, there is no situation like this and it is also not explained in a way I understand ?
What if I have e.g. multiple Vector{Float64} fields or a function more ? Can I leave them untyped under which circumstances ? Is there a risk of assigning the wrong type to a field ?
Edit: Thinking about it, somewhat echoing this thread may be I am looking for something more explicit like a fortran interface definition which leaves less (?) room for ambiguity.
Edit 2: So this would make the matching between the paremetric type parameters and the struct fields positional and then first match ? Is that right ? Does it take the actual types of the constructor arguments into account when matching (or is that something which cannot happen by language design) ?