Fork me on GitHub

@jr0cket that's the setup I use. I can't imagine going back.


If it eagerly evaluates it makes it feel limited to only trivial calculations


@andy.fingerhut sorry, I should have been clearer. If the default key binding in a REPL buffer for evaluating an expression was changed to C-RET, how many people would this annoy, anger or confuse. I don’t use the REPL buffer myself, hence the question. It seems Dan and I make two votes for such a change, so 2-0 so far.


I use the code buffer pretty much exclusively, unless I debug with printing. I’ve long been doing the journaling of the design choices (I call them REPL experiments) that Stuart talks about in the running with scissor talk.


However, the REPL and REPL history experience seems to be broken in Spacemacs (Evil state). While looking into that I noticed the REPL buffer didn’t do multi-line. Every other editor and even the Clojure CLI tools with rebel.readline does multi-line, so this behaviour now seems to be the norm.


i think rebel readline does multiline by unfinished parens. I don't much like this approach but I think @richiardiandrea built a clever feature that will auto close all parens and eval if that's the way you like to work


Well not even mine, but I think I would be able to dig it up from my emacs conf 😃


I thought you committed that to cider


Uhm if I did that I do not remember 😄


I always code in a source buffer and just send stuff for evaluation to the REPL. From time to time I might do some experiments in the REPL itself, but that happens pretty infrequently.


I did a (bad) demo of my approach a while ago

👍 4

> However, the REPL and REPL history experience seems to be broken in Spacemacs (Evil state). While looking into that I noticed the REPL buffer didn’t do multi-line. Every other editor and even the Clojure CLI tools with rebel.readline does multi-line, so this behaviour now seems to be the norm.


@jr0cket not sure what you mean by this. CIDER’s REPL does multi-line just like any other Emacs REPL (using unfinished expression or C-j) and that’s properly reflected in the history.


@bozhidar its Spacemacs Evil specific and I've been thinking about how to improve this in the Spacemacs clojure layer config. C-j is used for traversing history so cant be used for insert-and-indent in the REPL buffer and the REPL buffer keybindings only works in Evil insert or Emacs state


I use evil mode and find C-j behavior perfectly acceptable. Not sure what you're talking about.


Yes but the C-j is for traversing REPL buffer history in Spacemacs Clojure layer, not for cider-repl-newline-and-indent. In the Cider docs it defines C-j as the keybinding for newline interaction in the repl, but the repl history in Spacemacs over-rides this.


I have [cider/piggieback "0.3.9"] in my project.clj but emacs give me an error that it is missing cemerick piggieback. I have search my project, .lein and .emacs for a reference to cemerick and cannot find one. I also did a lein deps :tree


What version of cider are you running


How do I control the version of cider? I just checked out my spacemacs config SPC f e d and I dont see any version numbers.


so that was in the transition period of cider towards cider/piggieback from cemerick/piggieback. and that may predate cider/piggieback so it doesn't know that that is an acceptable solution


let me check out 0.17 and see if there's a knob. i think there'sa defcustom somewhere to inhibit those checks. just need t osee what the name is


you can monkey patch it here (defun cider-verify-piggieback-is-present ()) with that


and you can find that function and see how it is checking for the presence of piggieback


Yeah, that seems like a reasonable approach


In 0.19 I’ll completely drop the cemerick/piggieback stuff, so we can simplify a bit the code.


M-x configuration-layer/update-packages
154 packages to update 😛


ha. i did 127 or so a few weeks ago and it worked flawlessly. good luck


when I start the repl I see emacs using 0.19 snapshot but my repl says 0.18 snapshot?


beside that I still get the error after I start figwheel and use (cljs-repl)


I guess the dependency is in figwheel?


oh you're getting this from figwheel, not cider


which version of figwheel are you on. i think you need to be on 16+


I put back the cemerick version. I am using 0.5.16 of figwheel


I posted my real problem in #figwheel , all day I have been fidgeting with the settings, starting fresh, updating things, following tutorials. Not really how I wanted to spend my home coding time.


I think what it comes down to is the emacs repl is not calling mount-root when I save a file but lein figwheel does.


ok, figwheel updates my page from emacs when I use inf-clojure but then there is no cider so I dont get the cool evaluations.


Trying to resolve an annoyance that I’ve let go on for too long. Whenever I type the following on a new line:

(def some-var "some-string")
when I type the first character of “some-string” (after the opening double quote), the form always gets indented for some reason:
(def something "some-string")
^did not type
I assume it has something to do with “some-string” being ambiguously parsed as the docstring, but I’m not sure why it is indenting (or what is indenting it automatically). Any advice on debugging/changing? Anyone else running into this?


Active modes: clojure, clojure-refactor, cider, EditorConfig, paredit