I should have added more context about my issue to the ticket because I'm not remembering the full context anymore either. I was trying to get repls connected to webworkers using the vanilla build machinery (not shadow or figwheel) and brepl was getting injected into the worker script, I believe. My thinking with regard to the bug was that [this change](https://github.com/clojure/clojurescript/commit/5b2d4b8a33355dfcde16e54b9175b97cc6d76c91) was conjing :browser-repl into clojure/known-opts but that hash-map already contained that key, so I think I thought that the intuition there was accidentally inverted, and that we wanted to be removing :browser-repl from the known-opts if the user does not supply the "-r" cli option. And IIRC my change did enable/fix repling directly to webworkers using the vanilla stuff.

^ @dnolen

At the time though, I certainly didn't think I broke normal repling, as I wouldn't have been able to do normal repling and it would have been broken for everyone else as well, I'd think. So either there's something more specific about the way the user is launching the repl here: https://ask.clojure.org/index.php/13279/clojurescript-issue-in-1-11-121-and-master-when-compiling Or maybe there was a change since then that isn't compatible with my change? Not sure.

Perhaps I was only testing against the repling launch methods described in the vanilla tutorial and the above user's setup is a more specific launch case that I missed. When I follow the above user's repro steps though, I do get a hung repl, as described in the bug report. I'll try to dig in more tonight.

thanks!

w/ a bit more info happy to help figure out a way to solve your original issue

Okay cool, thanks @dnolen!