maowtm

A related thought: maybe we can solve the user error problem by making sure that we have clear error messages. When I tried `a :- b()` or `a :- b::c()`...

Just my personal opinion... but I actually prefer no parenthesis if we have to enforce either one or the other, since being able to just write `thing :- other_thing` is...

I also noticed earlier that it uses tokio 0.x which can cause problem later if we add other dependencies, so I have been thinking of forking it. Haven't done so...

Actually for that we just need something that acts like a layer expression but creates no layer, rather than introducing redundant layers in the image. Maybe we can just make...

Proposal: `::cached(path)` (path can be relative or absolute, resolved based on cwd). Dir will initially just be empty.

Basically I had a `config` predicate that defined some config corresponding to variants, like: ```Modusfile config("default", "make all"). config("32bit", "make 32bit"). ``` etc... And I use it like this: ```Modusfile...

I was thinking about a warning when there are multiple valid choices for a predicate. This would also prevent mistake like ```Modusfile foo(bar) :- (bar = "1"; bar = "2"),...

I'm going to go with storing the arguments for builtin-predicates like run in ClauseID for now, since this is the easiest and it's not obvious what variable names should we...

We can already do it since we don't have type checking yet, and basically when we add type checking we will need to be able to do polymorphism.

This may actually require either finding another semver crate or writing our own semver implementation... So before we move on @mechtaev can we discuss the exact semantic we want for...