I've enjoyed all the posts in this series, but this is first time it's I've "empathized" with some aspect of the problem solving.

I've taken a couple of stabs at writing a mini/micro-kanren in a language which _isn't_ a lisp, often following along closely with some paper from Hemann, Friedman and/or Byrd -- and if the language _doesn't_ have macros it really does get awkward. As mentioned at the PS, for a good long stretch, this post is pretty directly transcribing the 'microkanren' paper, and without macros at some point Vidrun would have had to invent some of her own idioms.

The other trick to having the embedded kanren really work with the host language, which is swept a bit under the rug, is that `walk*` / `walkr` in this post implicitly assumes that everything in your universe is a variable, an atomic value, or a cons. In that sense, it's a little funny that (I think) her kanren can't really solve "inside" of the structs that she asked to be provided.