Clojurians Log v2

I feel kinda dumb because yeah, that's precisely what I expect and rely on. Maybe not for the plain eval, but I inspect "non-forms" very often (for example, symbols), and inspect being one of eval's flavor, I would think that it would get the same treatment too which would be quite disruptive for me (to the point of running a CIDER fork when the fallback option goes away).

Fair points, so I'm going back to the drawing board with this. The case when you're the final closing paren in quite tricky as this normally goes in the "form at point branch", but I understand completely that some people would prefer it work differently .

I might go back to my original idea to have an extra set of commands instead, so it would be less disruptive, although my concern with this approach was that most people would never find them.

(right now I was trying to reduce the number of form selection commands to make things more uniform and simple across the board)

I get where you are coming from. I didn't think that enough people have a problem with C-c C-e behaving in a surprising way for them. comment is a popular and universal one, but other than that I can't think of a situation where evaluating precisely one sexp around the cursor rather than the preceding form is the goal. But that's just me. Perhaps, somebody else can chime in with the examples where they will find this useful?

To be clear - until you flagged this I hoped no one would even notice the change in the defaults. 😄 Over the years the common points of confusion I've observed where what happens when people are on opening/closing delimiters of a form, so that's what I aiming for mostly. (and other thing was the people were overwhelmed by the number of similar evaluation/etc commands that only differ on the form that they target)

Fair point about the cursor being on the opening paren – no idea what the behavior should be in this case. About being on the closing paren – it might be somewhat confusing if there is just one form, but it's very straightforward to me if there are multiple forms:

I doubt anybody would expect otherwise?

I think some people expect this to act on the enclosing form. I think that's how Calva works for instance.

My rule of a thumb is "whatever paredit jumps back to with paredit-backward "😄

> I think some people expect this to act on the enclosing form. I think that's how Calva works for instance. That indeed feels like the whole other paradigm to me, that either deserves separate commands or a customizable switch.

Yeah, that's why I started thinking to just make this separate from the historical Emacsy commands. At any rate - you'll never have to run a fork for something like this. I'll figure a solution that covers everyone's use-case, and I guess the defaults will go back to how they were.

I would personally appreciate if C-x C-e and C-M-x continue to work as they currently do (i.e. similar to elisp), and people can deliberately use a dwim version assigned to another keybinding.

I'll revert everything soon and explore the separate commands angle for eval, macroexpansion, and other form-targeting actions.

FYI - https://github.com/clojure-emacs/cider/pull/4170

Now everything's back to how it used to be.

Didn't get round to trying out the changes, but I'm actually on board with the original blogpost - personally I've never gotten used to the whole last-sexp paradigm of doing things-

For whatever reason it feels more natural to place my cursor on opening delimiters and target the following-form - which I don't think cider supports out of the box, so i've had to hack around it locally with advices and custom commands.

(just thought to raise a differing opinion~ hopefully the next iteration can accommodate everyone)

@qythium last-sexp is natural with how you type. After you've typed (+ 1 2) , your cursor is after the last paren, so you don't need to move it anywhere to evaluate what you've just typed. This is probably the main motivation behind this paradigm.

Yeah, I understand the rationales but it just never made sense to my brain - besides, come to think of it with structural editing I never actually end up typing closing parens