This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2022-04-19
Channels
- # announcements (37)
- # aws (6)
- # babashka (12)
- # babashka-sci-dev (16)
- # beginners (83)
- # biff (10)
- # cider (14)
- # cljdoc (26)
- # cljs-dev (20)
- # clojure (123)
- # clojure-czech (9)
- # clojure-europe (26)
- # clojure-nl (4)
- # clojure-norway (20)
- # clojure-spec (7)
- # clojure-uk (6)
- # clojured (14)
- # clojurescript (28)
- # cursive (5)
- # datalevin (8)
- # datomic (3)
- # duct (6)
- # emacs (26)
- # events (2)
- # fulcro (7)
- # gratitude (1)
- # holy-lambda (19)
- # integrant (1)
- # jobs (2)
- # leiningen (8)
- # lsp (7)
- # nyc (1)
- # pathom (70)
- # re-frame (8)
- # reagent (15)
- # releases (1)
- # sci (8)
- # shadow-cljs (117)
- # testing (5)
- # tools-deps (11)
- # vim (5)
FYI, I’m still seeing the “HashMap$Node cannot be cast to HashMap$TreeNode” issue with latest Clojure CLI tools. https://ask.clojure.org/index.php/10703/error-building-classpath?show=11784#c11784
Yeah, I had another report about it. Haven't had a chance to look at it. Doesn't make any sense to me as everything should be thread local, but obviously something unexpected Is happening
my current hypothesis is that the DefaultModelValidator is truly a global singleton in some fashion.
DefaultRepositorySystem is what makes them if you want to chase that down. Or I guess just run a tools deps system, then inspect all the DefaultModelValidator instances
I assume it's something like that, because Maven is awful
it is a giant tower of mutable maps, disguised in Java classes, glued together with 3 kinds of dependency injection. there's no reason for any of it to be mutable. why would a validator be mutable - that is a function. bleh.
https://get-coursier.io/ if scala's bytecode compatibility wasn't as odious as it is I would almost say this would be an option
I have to talk myself down from rewriting the whole thing every time I work on it
I know, but you know what I mean