Re: scratch/command 064f146 1/2: Change command to interactive ... modes

[Dmitry Gutov]

On 16.02.2021 22:30, Stefan Kangas wrote:
> But I would claim that the more verbose form in this case is less clear
> and harder to understand.

That's because 'completion' is not a good name for it: it doesn't, 
indeed, explain what the value is. Nor does it say that it will be used 
by the proposed execute-extended-current-mode-command command.

OTOH, if we had annotations like

(declare (applicable-major-modes MODE1 MODE2 ...))
(declare (applicable-minor-modes MODE1 MODE2 ...))

it would both allow some speed improvement in the completion predicate 
(don't have to apply both major and minor mode checks to each symbol) 
and more consistently allow read-extended-command-predicate to use 
arbitrary conditions (at the user's choice).

The list of annotations can be expanded with other tags, too, as well as 
with a free-form predicate

(declare (applicable-pred PRED))

P.S. I'm guessing the applicable-* names look pretty long for a number 
of readers here. Surely someone can pick a shorter naming scheme that's 
still semantic.