Carl Mäsak

It's interesting to evaluate this issue in the light of the new contextual macros #349. It feels to me that this would need to be a "transformer macro", basically a...

Herb Sutter in [this talk](https://www.youtube.com/watch?v=8U3hl8XMm8c) (and [a talk prior to that](https://www.youtube.com/watch?v=fJvPBHErF2U)) explains how he implemented chaining operators in cpp2 _but only the ones that make transitive sense_. In other words,...

I think the simplest way to phrase it is "they don't interact". For example, the fact that we make an operator `infix` carries with it certain preconditions and postconditions for...

Just a quick addition: ``` is qtype(Q::Infix::Ternary) ``` Which will have the consequences that when someone later inspects the Qtree of a program with a `?? !!` operator in it,...

In the case of a missing `!!`, the implementation of `?? !!` must be able to emit a parser error that's *just as good* as what the compiler had emitted...

I had two scattered thoughts today: 1. The final implementation of `?? !!` should ideally be able to propagate lvalues. That is, `(a ?? b !! c) = 42` should...

> 1. The final implementation of `?? !!` should ideally be able to propagate lvalues. That is, `(a ?? b !! c) = 42` should be able to assign to...

> 1. Fix the definition of `bquote` in bel.bel so that it is not infinitely recursive (it doesn't directly call "bquote", but it does use backquotes, which has the same...

Thanks! My solution thus far was more complicated — I will try yours in my implementation and report back. Relatedly, however, I think I will need to make an "errata"...

> There should be no observable difference between evaluating expression `χ` and evaluating `(bel 'χ)` This lodestar (together with things like #11), I think it leads to the REPL using...