That's why few people are using things like parinfer. This demo might look cool to non-lisp programmers, but in lisp-land, everyone is using semantic editing (in Emacs or otherwise), and not line-based editing.
That's why few people are using things like parinfer.
I don't think this is true. I develop Cursive and I use it, despite being competent at paredit. A ton of my users do too, even experienced ones. It's just easier, and it turns out that a lot of editing is actually line based, even in a lisp. And when it's not, I can still use the paredit commands in parinfer mode, at least with Cursive.
It's referenced in an edit to noctuid's document, but many of those criticisms are no longer valid since smart mode became the de-facto default for parinfer. It's now really intuitive. Parinfer's worst problem now is its difficulty inferring parens when there are multiple expressions on a line (i.e. when there's no indentation to infer from), but it's surprising how infrequently that's actually a problem.
Paredit is really not the closest or best comparison. If you want simplicity, lispy is often simpler than or at least as simple as parinfer. At the same time, it's more powerful than paredit. Even if you never used its special keybindings, I'd argue that lispy is more intuitive and has more useful default keybindings. The indentation-based inference doesn't add anything major as far as I can tell; it seems to just cause unnecessary problems. I think parinfer is more popular in editors besides Emacs because of the lack of alternatives more than anything else.
What percentage of users is that, assuming that you don't make parinfer the default choice in your IDE? What percentage of Cursive users knowingly say: I'm a bit confused by how these parens work (my guess: all lisp newcomers), what percentage goes to preferences to see what they can do about it (my educated guess: a lot, but less than, say, a third), and what percentage of those conciously look for parinfer (I don't know - I'm asking)?
Now, Cursive is used by something like a third of Clojure users. Half of them use Emacs, and in Emacs land there are rather few Clojurians who seek something like parinfer. Even when they do, it's a usual "how we can make this more palatable to outsiders" type of quest rather than "I'm really frustrated by how Emacs does Lisp editing. I'd like something like parinfer"...
I don't know exact numbers, since I don't do any user tracking. But I know of numerous users who are already experienced with Clojure who have switched, myself included. Cursive currently has paredit on by default, but I'm going to switch that to parinfer.
FWIW Emacs users do ask about parinfer in #parinfer on Clojurians because they actually want to use it. There currently isn't a smart mode for Emacs though, and that's what I consider the game changer. I suspect that uptake will be good once that's available.
Parinfer seems popular in the clojure community (even among those who have been writing it for a while) and among people new to lisp. In its favor, parinfer's listed selling point is simplicity and not power. I don't think it does simplicitly perfectly, but I think it does it better than paredit, for example.
