Re: Add function to rotate/transpose all windows

[Pranshu via "Emacs development discussions." <emacs-devel@gnu.org>]

On 25 December 2024 5:33:04 pm GMT+05:30, Eli Zaretskii <eliz@gnu.org> wrote:
>> Date: Wed, 25 Dec 2024 07:05:19 +0800
>> From: Pranshu <pranshu@bauherren.ovh>
>> CC: rudalics@gmx.at, emacs-devel@gnu.org
>> 
>> On 24 December 2024 8:58:10 pm GMT+08:00, Eli Zaretskii <eliz@gnu.org> wrote:
>> >
>> >Why are we allocating key bindings to commands before they are enough
>> >time in use to know whether they will be used frequently enough to
>> >justify a default binding?  There's nothing wrong with a command which
>> >has no bindings, there are gobs of them in Emacs.
>> 
>> how will we know if or not a binding is used frequently?
>
>By listening to users who ask for a binding.  Also, by using this
>command ourselves and paying attention to the frequency of such uses.
>
>> IMO binding commands makes it more discoverable
>
>How so?

Lots of people use which key, also I can imagine people assume some specific prefixes have all the relevant key bindings for a purpose, so just do C-h to list bindings for the map. Ik I did this for project and vc and window prefix map.

>> and a specific window prefix map exists solely for this purpose. 
>
>The prefix map exists for the purpose of binding window-related
>commands, but it has no bearing on the decision whether to provide a
>binding in the first place.

The C-x w is very empty, and it would be good if it had new keys. I honestly don't ever imagine that binding having 26 commands, so any binding is better than no binding. IMNSHO these commands would be by far the most useful in that map after C-x w (2|3).

The problem is basically people will assume all relevant bindings are in that map, since it is so empty, new commands would be added.

>> >Let's not rush into reserving key bindings we are not sure will be
>> >needed, because it will be nigh impossible to change or remove them later.
>> 
>> I didn't know situation was that bad, maybe someday I will read M-g change mailing list just for entertainment.
>
>I don't think I understand what you are saying here.

IIUC the problem is that curmudgeons storm emacsdevel when there is a change, IIRC most recently this happen with the Thierry register drama. Has there been a case of this happening with any binding released after emacs 24?