Anyone know where I should start to debug a failing cider-connect? I’m connecting to a docker-compose env and it “suddenly” fails for me. I managed to create a never ending recursion earlier, not sure if that is related or not. Is there something that I maybe need to clear? “”helm-M-x: Sync nREPL request timed out (op eval code (seq (.split (System/getProperty “java.class.path”) “:”)) id 3 time-stamp 2019-07-04 14:37:15.403819000)” A bit rambling on my end …

That’s a strange error to see helm m-x involved

@dpsutton sesman needs the classpath early on, so I guess that’s what triggers the error, but generally this basic code should always work.

I guess there’s some unhandled exception that blocks this eval or something like this.

@kardan if you're around i'm happy to help debug

@bozhidar PR up that makes all tests pass. bug in emacs C code 🙂

So I was just about to work on my project, but when I ran cider-jack-in I got this stacktrace https://pastebin.com/Hm17dq0p. It's been working fine in the past, but I'm not sure what has changed.

@nicholas.jaunsen Not sure what exactly is wrong, but I assume that upgrading to the latest CIDER and leiningen will fix the problem for you.

@dpsutton Sounds like you had a lot of fun with that one! 😄

up till 4 trying to find out why 🙂

That’s what I call true grit! 😄

it'll be nice to see that green checkmarks everywhere. Can't figure out why it only happens in cask though

Now the only outstanding issue with the tests is figuring out why one test fails without lein being installed (something about cider-jack-in-clj&cljs).

there's a failing test? which one?

if its cider-connection-tests it was the same issue

(describe "cider-jack-in-clj&cljs" this one?

It fails if you remove the installation of lein in the image. It wasn’t need in general, we’ve added it as a quick workaround.

oh

None of the tests really run lein, so this wasn’t supposed to be needed.

Generally CIDER has no real “integration tests”, as those are a bit tricky in Emacs.

yeah for sure. i didn't think make test called out to lein. i'll look into it

ah, at one point i deleted that test. it verifies (expect 'y-or-n-p :to-have-been-called-times 3) which is just nonsense to me

it gives no insight into anything. I think its testing the params construction functions but it gives no particular insight. if that changed to 4 or 2 it wouldn't signal anything

and removing that test removes the dependency of lein.

huzzah!

i'll work on the lein stuff. annoyingly, getting weird circle test runner errors when i modify the test buffer. i think there's caching stuff going on 😕

but i'm pretty sure that test is verifying that we ask for the port...