A shame that acme(1) only gets a passing mention when the original paper on acme’s predecessor help really drives the point home for having a mouse-friendly interface: https://doc.cat-v.org/plan_9/1st_edition/help/

Unfortunately it is uniquely difficult to make any nuanced point about mouse usage because of the inertial pull of the decades old keyboard vs mouse argument, and the equally vapid “use both” argument. Talking about Acme’s design would have derailed things further.

I wasn’t advocating for mouse-friendly interfaces at all, by the way. My goal was only to point out that (i) there are times when using the mouse with Emacs is natural, although the specifics depend on your use of Emacs, and (ii) Emacs includes some features to improve mouse expressivity (gestures, configurable drag-and-drop, etc.)

This isn’t so much a problem with use-package as it is the primitives (autoload and with-eval-after-load) and the way Emacs manages load order with load-history. I don’t know that you can write a better macro to handle composite package configurations under these constraints. I also haven’t tried to, so maybe there is.

An actually declarative system will need something like a fixed point evaluator, and be much more complicated than use-package internally.

Practically speaking, the only solution right now is (i) proactive measures by package authors to shove all the bits meant for integration into separate file(s) and provide separate features, and (ii) knowledge of this organization on the part of the user.

Even this can be too much, magit-extras (which contains magit-project-status) is 900 lines, most of which you don’t need.

Github, MELPA

In the meantime I added explicit support for GPT4All, the above instructions may be incorrect by the time you get to it. The Readme should have updated instructions (if it mentions support for local LLMs at all).

In that case you can use it right now with gptel, which supports an Org interface for chat.

Enable the server mode in the desktop app, and in Emacs, run

(setq-default gptel-model "gpt4all-j-v1.3-groovy"
              gptel-host "http://localhost:4891/v1"
              gptel-api-key "--")

Then you can spawn a dedicated chat buffer with M-x gptel or chat from any buffer by selecting a region of text and running M-x gptel-send.

I can add this to gptel quite easily, but I can’t find the instructions on how to use it. Does it use a local http server? Where can I find these details?