Eric Dallo

Yeah, I can see how that could leave some corner cases, and I think it's not worth the effort

Just linking related issues https://github.com/clojure-lsp/clojure-lsp/issues/1193

@borkdude I think we would need to make clj-kondo return valid analysis when we have an invalid code like `(defn foo ^ [])`, WDYT? similar to how clj-kondo handles invalid...

https://github.com/clj-kondo/clj-kondo/issues/1778

We would need to make clj-kondo return analysis for `bb.edn`, then clojure-lsp would find that those functions are really being used as bb tasks, what you think @borkdude?

@borkdude what do you mean to treat the task .clj as a library? How can we know that a specific user ns is supposed to be a library?

Yes, I think that would be a new feature for clojure-lsp where user specifies those ns, right? not sure how common is that since most cases I saw are related...

Yes, but we already have the config to allow that, right? `.clj-kondo/config.edn` ```clojure {:linters {:clojure-lsp/unused-public-var {:exclude #{my-ns}}}} ```

I'm fine with both

Maybe we should append a sufix for the created elements like `element-1`, `element-2` etc c/c @mainej