There should be a build system with a lisp-based grammar that does what bazel tries to do
@sir serious question: which lisp?
@apg a new one
@sir I've got some ideas for a new Lisp, though I've gotta do some writing and prototyping ....
I question not adopting Racket for this, but if you've got ideas, I'd love to hear more!
@apg I don't want to use a general purpose language, I think a DSL is better suited and lisp is good for making DSLs with
@sir I’m certain I’m just not understanding. Is your intention to write an interpreter to interpret an s-exps based DSL? If so, I’d suggest you save yourself some trouble and write a DSL _on top of_ an existing Lisp. _That’s_ where the language really shines. (I happen to think that #racket provides the best tooling for making DSLs...)
@sir I agree! The great thing about racket is that your DSL, can by default, not export more than is needed. So, you could define the core forms and only expose those as part of your `#lang better-bazel` and not have more power than you want. However, because that language is implemented in racket, when it becomes necessary to integrate with other tooling you have everything at your disposal.
If nothing else, and you are serious about this as a project, check it out. 🙂
@apg thanks, I'll definitely check it out when the time comes. But I still suspect this is coming at the problem from the wrong direction.
@apg I understand where you're coming from, but I don't think that's the best approach. It's not good for build systems to be too smart.