Daniel Mendler

A simple prototype: ~~~ elisp (defun embark-target-gnus-attachment () (declare-function mm-handle-filename "mm-decode") (declare-function mm-save-part-to-file "mm-decode") (defvar gnus-article-current) (when-let ((handle (get-text-property (point) 'gnus-data)) (file (mm-handle-filename handle)) (beg (previous-single-property-change (point) 'gnus-data nil (pos-bol)))...

Well, you are right. I am inclined to close #756. But then I just tried `gnus-mime-view-part-internally`, `gnus-mime-view-part-externally` and `gnus-mime-view-part`. These commands just don't work as nicely. For example `gnus-mime-view-part-externally` on...

Maybe it is just a matter of configuring mailcap properly, such that it behaves more in line with the rest (I will experiment). I think treating attachements like files in...

I think reverting should remove the overlays automatically, but I am not sure about the best way to do this.

@dwa Do you have a simple use case for testing this, something which doesn't involve large packages like lsp-mode, ideally starting with bare emacs -Q only?

Thanks. I think this is already helpful.

Okay, as far as I can tell, this PR does not improve things, at least in this test case: Before:  After: 

> The main reason for the patch is to show keybindings from lsp-mode, so a no-change/no-worse on other entries should really be a good thing here. Sure. But it is...

As far as I know, the small patch limit can somehow add up, leading to something non-trivial in the end, maybe even across parts of the overall GNU/Emacs project. It...

I think our goal is to move to a better configuration method using normal Emacs keymaps. This will make configuration easier and should prevent problems. But I am not sure...