09:20 <Putonlalla> I can wrap the `struct` macro in this way, and it will work as expected: `(define-syntax structer (syntax-rules () ((_ arg ...) (struct arg ...))))` and then `(structer document (author title))`.

09:20 <Putonlalla> However, if I do this with my own macro that uses `datum->syntax` to create identifiers, the bindings get lost along the way. How should this be done correctly?

09:35 <Putonlalla> Looks like my minimal broken counterexample turned into a minimal working example, which is quite annoying.

09:46 <Putonlalla> This demonstrates the problem. https://pastebin.com/hnHVYufb

09:47 <Putonlalla> The two definitions look the same, but do not work the same.

09:55 <Putonlalla> I have a hunch that the `ctxt` passed into `syntax-append` is somehow wrong.

14:20 <samth> Putonlalla: using `#'id` for the context will make it work

14:20 <Putonlalla> Ah! Indeed.

14:21 <samth> right now, you're using `stx` which is the syntax object created by the expansion of `structer*`

14:21 <samth> in general abstracting over non-hygenic macros is hard because which syntax object you use for the context is part of the interface

14:22 <Putonlalla> Could I use any identifier that comes from "the user" as the context?

20:15 <samth> Putonlalla: yes, but again which one you choose is part of the interface (eg if you write a macro that synthesizes the field names then they don't come from the user)

23:21 <Tuplanolla> Okay, cool.