Le dim. 13 sept. 2020 à 16:21, Eli Zaretskii <email@example.com
> a écrit :
> From: Thibaut Verron <firstname.lastname@example.org>
> Date: Sun, 13 Sep 2020 07:51:53 +0200
> Cc: Ergus <email@example.com>, firstname.lastname@example.org, email@example.com, firstname.lastname@example.org,
> email@example.com, firstname.lastname@example.org, email@example.com
> > The menu we have in C-mouse-3 does not show the most basic options like copy, paste, and so on
> > access them fast.
> Why should it? We show the menu for the current major mode, which is
> IMO more useful than basic editing.
> Is it? What feature of your major mode do you use more often than copy/paste?
Typing characters, of course. I hope you will not suggest that we
should have a menu item for inserting a character.
Of course not, the keyboard shortcut for character insertion is not surprising. But basic operations which are not insertion? Copy/cut/paste, undo/redo, maybe search/-and-replace... All those are very common operations, none depend on the major mode, and all require to either learn an unusual keyboard shortcut, use the toolbar, or navigate the menu-bar.
My point is that frequency of operations is not the only criterion for
what to put on the context menu, not even the main one. The most
important criterion, IMO, is what are the important operations that
would be otherwise much harder to discover and invoke.
I agree, but I'd say that ease of invokation should take priority over ease of discoverability. The menu bar offers just as much discoverability, but the ease of use is greatly decreased there.
I have only one data point at hand, but I happen to have helped a new user set up an emacs environment recently. They were happy with finding options in the menu bar, and (a selective list of) major mode commands in the tool bar.
The major-mode submenus of the menu bar were overwhelming on the other hand, and finding the same menus on the context menu was not much help.
Not finding common operations such as copy and paste in the context menu was more disconcerting (and directly led them to discovering and activating CUA).
Ditto when their spell-check (again, activated without my help via the menu) flagged some words and they didn't find the corrections in the context menu.
We decided that the menu for the current major mode is a very good
approximation to what the user would like to have at his/her
Was it perhaps at the same time as it was decided that this menu should require a modifier key in the default bindings? ;)
The context menu in emacs is underused to say the least, and I'd blame that on both the hidden binding and the redundant contents.
Sadly, by the point users know enough to know how to address/report the poor condition of the context menu, they simply don't care anymore because they can get everything done with keyboard shortcuts and menu/tool bar.
There's probably some space for improving that, but I
think the basic principle that the context menu should depend heavily
on the major mode is valid and should be preserved.
I believe that it should depend on the context in a wide sense. That includes the major mode, but also the minor modes, the thing at point, whether the region is active, etc.
Copy-cut-paste and auto-correct suggestions should be no-brainers, for instance.