Milo

i totally forgot how this works anymore lol

a guess: nanocaml generates recursive calls so if you dont mark your passes as recursive then it will generate calls to unbound functions

right so instead of checking if there are any [@r]'s we just force the function to be recursive so that [@r] is valid no matter what

i have a similar problem: `conf-emacs` wants me to install `emacs-nox` through my package manager even though i already have gui emacs installed

How about creating a new table with an identifier relating to the naughty string that created it, so that you notice the new table in the database and can specifically...

[here](https://github.com/iitalics/macrotypes/commit/4b02839ed50c9f663cb3c29094bab0f8b74cbc17#diff-f6a3668f1fb2a02fbb91125d7d1f5fa3R866)

Yes it's unbound identifier. Which on second thought doesn't make a lot of sense because `expand/df` resolves yet its defined below `infers+ctx/erase`. :man_shrugging: However it does spit out a ton...

It's easy to introduce @AlexKnauth's changes from #10: [here](https://github.com/iitalics/macrotypes/commit/7b42684ed85cd729902afd45b53d4447d14ba5a9)

That's a good point. The problem is that the make-variable-like-tranformer call happens in phase 1, even though make-variable-like-transformer is passed as a macro. Its especially confusing to have templates inside...

*even though make-variable-like-transformer is passed as a template.