Starting Calva and jacking in currently means creating three clj-kondo caches and then some .lsp. cache files as well. The .clj-kondo caches are:
.clj-kondo/cache.clj-kondo/.cache (because the .clj-kondo directory exists because of 1.).clj-kondo/.cache (different version subfolder)There's a picture here that tells this story: https://github.com/BetterThanTomorrow/calva/issues/917
Although this is to some extent going to be dealt with upstream, (hopefully as upstream as well https://github.com/clj-kondo/clj-kondo/issues/1113, but we'll see about that), we should probably consolidate our use of clj-kondo some. I know we are heading towards using the linting via clojure-lsp. Is there also a way we can use the clj-kondo analysis via the lsp, maybe? That would make it easier to use the debugger too.
clojure-lsp creates as .clj-kondo/cache here
Thanks. Updated the issue with the correct path.
Is there also a way we can use the clj-kondo analysis via the lsp, maybe?
I've been thinking about this for some time. Created an issue.