@dpsutton I was trying to use the use the cider-nrepl 0.19.0-SNAPSHOT with refactor-nrepl 2.4.0 and it didn't like it apparently.

I’m having trouble running a cljs repl

I have figwheel and piggieback setup in my dev profile:

{:dependencies [[figwheel-sidecar "0.5.14"]
                                  [com.cemerick/piggieback "0.2.1"]]
                   :source-paths ["src" "env/dev"]
                   :cljsbuild    {:builds [
                                           {:id           "ios"
                                            :source-paths ["src" "env/dev"]
                                            :figwheel     true
                                            :compiler     {:output-to     "target/ios/index.js"
                                                           :main          "env.ios.main"
                                                           :output-dir    "target/ios"
                                                           :optimizations :none
                                                           :target :nodejs}}
                                           {:id           "android"
                                            :source-paths ["src" "env/dev"]
                                            :figwheel     true
                                            :compiler     {:output-to     "target/android/index.js"
                                                           :main          "env.android.main"
                                                           :output-dir    "target/android"
                                                           :optimizations :none
                                                           :target :nodejs}}
                                           #_($DEV_PROFILES$)]}
                   :repl-options {:nrepl-middleware [cemerick.piggieback/wrap-cljs-repl]}}

and (setq cider-lein-parameters "with-profile +dev repl :headless") in my init.el

@martincharlesrichards You need cider/piggieback instead.

The project artifact was changed when Chas handed it over to us.

ah, thank you

let me give that a try

fixed! thank you

You’re welcome!

I’m working towards a deps.edn-based setup — can I start up a Cider session with -A aliases specified?

Can’t seem to find it in the docs

I'm surprised that works with figwheel 0.5.14. I thought 16 was the first with cider piggieback support

@dpsutton The first with nREPL 0.4 support. 🙂

ah i had it backwards. thanks

Hi, i updated packages and my repl is spewing a few screens of messages messages like the following: Boxed math warning, cider/inlined_deps/fipp/v0v6v13/corerrb_vector/v0v0v12/clojure/core/rrb_vector/nodes.clj:88:11 - call: public static boolean http://clojure.lang.Numbers.lt(long,java.lang.Object).

Seems to be working though.

cider-version: CIDER 0.19.0snapshot (package: 20181019.812)

how old was your CIDER before?

i bet you've been getting these warnings forever, they've just gone to the server process buffer rather than your repl

ah. i see these are the inlined cider deps

I just started seeing this too fwiw

@j0ni It’s from the upstream dep.

I also see a full exception at the end

They have to fix the warnings themselves.

is that the same deal?

Nope.

That’s different.

worth a bug report?

Seems a recent change is misbehaving.

https://github.com/clojure-emacs/cider-nrepl/pull/551

That’s the PR which introduced the problem (it added compatibility with Clojure 1.10).

Don’t have time to debug this, but I assume the problem is trivial by looking at the stacktrace.

@kharbo You can just change cider-clojure-cli-global-options.

@bozhidar Thanks

I have an instance of emacs 25.1.1 install with cider 0.18.0. My project directory looks like this

vagrant@taps-tvisher1:/opt/code/docs/scripts/tap-generate-docs$ tree -I 'bin|test'
.
├── deps.edn
└── src
    └── tap_generate_docs.clj

you can eval this (clojure-project-dir (cider-current-dir)) to see what it thinks is going on directory wise

Thanks. Sec.

and then (cider-project-type project-dir)

nil

ah. anything funky going on? sym links, remote systems, anything like that?

I guess I should also say that the buffer is TRAMP.

That would've been helpful to state up front 🙂

But with emacs 26.1/cider 0.17 this set up works fine.

i have no experience with that. i can only point you towards places

so you downgraded emacs and upgraded CIDER lol

Heh. Well this is for a friend who's not running my blessed config. 😛

But yes, essentially that's what's going on.

It would be nice if ELPA allowed installing previous versions.

check out clojure-cache-project-dir'`

variable?

that might be holding onto something before TRAMP is run. set that to nil if not already

yes

(defun clojure-project-dir (&optional dir-name)
  "Return the absolute path to the project's root directory.

Call is delegated down to `clojure-project-root-function' with
optional DIR-NAME as argument.

When `clojure-cache-project-dir' is t the results of the command
are cached in a buffer local variable (`clojure-cached-project-dir')."
  (let ((project-dir (or clojure-cached-project-dir
                         (funcall clojure-project-root-function dir-name))))
    (when (and clojure-cache-project-dir
               (derived-mode-p 'clojure-mode)
               (not clojure-cached-project-dir))
      (setq clojure-cached-project-dir project-dir))
    project-dir))

i think this is what you need to step through

possibly its cached locally and not aware to check remotely?

Oh. I have an idea now.

good luck and update us for sure

I didn't know that cider had a direct dependency on clojure-mode.

i'm sure a PR would be incredibly welcome.

yup

Their version of clojure-mode is ancient.

easy fixes are best fixes 🙂

So new nrepl claims to come up correctly but we fail to attach to it over localhost:<port>.

Which is probably definitely a weird TRAMP thing.

so you've solved the first bug and now on to a second bug?

Yep. Downgrading to 0.17 gets us past that. So some regression.

oh so the first bug is still present in 0.18? upgrading clojure-mode didn't solve it?

Sorry. No. We're past the first bug. It was a classic melpa v. melpa-stable problem.

I would've seen that if I had remembered the dependency. 🙂

oh. the bug in 0.18 is in connecting over tramp

Appears to be yes.

so you can work normally if you downgrade to 0.17

I can but he can't, it seems. 🙂

Still digging.

works on my machine is the worst

Heh.

This is why I have a carefully curated bash/emacs/etc. environment that is shared with all my coworkers so they don't have to think about it.

But my friend here doesn't feel like using that. 🙂

So we're getting an nrepl op error.

ok. that sounds like a more familiar error. any plugins? older versions of clj-refactor?

check lein profiles for offenders

Debugger entered--Lisp error: (error "Invalid byte opcode: op=183, ptr=2")
  signal(error ("Invalid byte opcode: op=183, ptr=2"))
  error("Invalid byte opcode: op=%s, ptr=%d" 183 2)
  cider-jack-in-command("clojure-cli")
  cider-jack-in(nil)
  funcall-interactively(cider-jack-in nil)
  #<subr call-interactively>(cider-jack-in nil nil)
  ad-Advice-call-interactively(#<subr call-interactively> cider-jack-in nil nil)
  apply(ad-Advice-call-interactively #<subr call-interactively> (cider-jack-in nil nil))
  call-interactively(cider-jack-in nil nil)
  command-execute(cider-jack-in)

I mean his emacs install has a bunch of stuff that I don't run so that could definitely be something. I could emacs -Q and get cider installed and see where we go.

ah i think that's easy

nuke *.elc files

that looks like incompatible emacs byte code which results from emacs X compiling elc and emacs Y running

Ah. Interesting.

you mentioned differing versions of emacs

I did. And actually now that you mention it I downgraded him to 0.17 by copying it out of my emacs 26 config. 🙂

ah yeah. that'll do it

source should be fine just wipe those elc files

Got it. We're in. Thank you for the help. 🙂

for sure

if yall could take a minute to lodge a ticket about the tramp jack in regression it would be appreciated

i've never really used tramp so i can't put the salient details in

I have a cljs / figwheel project with an ns containing a simple spec: (s/def ::content string?). Is it expected that upon jacking in I would see the ns as an option to select from M-x cider-browse-spec ? Because I don't see it.

@dpsutton I'll try for it.

I upgraded Cider on Thursday and now this happens. I can’t jack in. Any ideas?

@creese You may be mixing melpa and melpa-stable reqs.

I am trying to set the following in spacemacs in the user-config but it doesn't seem to be working. Any idea how to go about debugging it? I have tried with

(progn (require 'cider-repl-mode-map) ...)

(define-key cider-repl-mode-map (kbd "RET") #'cider-repl-newline-and-indent)
(define-key cider-repl-mode-map (kbd "C-<return>") #'cider-repl-return)

Maybe a spacemacs question?

you can't require cider-repl-mode-map

cider-repl-mode instead?

(require 'cider-repl)

if you want to require the file that has that var do that with (require 'cider-repl), else just use the var

since (defvar cider-repl-mode-map ...) is in cider-repl.el

what error are you seeing?

No error, it just isn't binding.

and those forms are in the init part of the use-package form in spacemacs?

(use-package foo
  :init
  (setq foo-variable t))

That's beyond me. I'm pretty new at this. Lol

fair enough

so where in your config are you writing these statements

I have put them inside of (defun dotspacemacs/user-config () ...) which says it's called at the very end of the spacemacs init after layers config.

ok. so what's happening (i think) is that use-package is lazy so even though the config has "run" it hasn't done anything (yet)

so that variable cider-repl-mode-map isn't avaialble since the source hasn't been required yet

so try

(require 'cider-repl)
(define-key cider-repl-mode-map (kbd "RET") #'cider-repl-newline-and-indent)
(define-key cider-repl-mode-map (kbd "C-<return>") #'cider-repl-return)

So that's where I got

(progn
    (require 'cider-repl)
    (define-key cider-repl-mode-map (kbd "RET") #'cider-repl-newline-and-indent)
    (define-key cider-repl-mode-map (kbd "C-<return>") #'cider-repl-return))

and when you evaluate that (if you know how) does it not add those bindings?

I'm not sure how to evaluate it. I just restarted emacs (or reloaded the spacemacs dotfile) and it didn't seem to bind them.

no worries 🙂

put your cursor at the closing ) and we want to run eval-last-sexp

do you know how to invoke m-x? I'm not sure how to do that in spacemacs

and in fact #spacemacs might prove more helpful

at this point

Yep alt-x 😉

I can ask there though. I didn't realize it existed before posting here.

so its the same. alt is "meta" which is the m in m-x

Yep

yeah if you have questions about cider specific stuff i'm happy to help. i just think that channel will be able to quickly get you up and running with your config

Just to answer the last bit. I eval'd it and it worked. :thumbsup: I'll move this to that channel to figure out how to get it to load with the user-configs. I appreciate the help.

awesome. yeah i'm sure there's a convention they can show you