Discussions about key bindings

Discussions about key bindings

[Lars Ingebrigtsen]

I find the recent spate of discussions about key bindings for
third-party code rather befuddling.  It's been suggested that Magit
should move their bindings to `C-x v', and it's been suggested that
there should be a number of bindings that should be reserved for
third-party packages, and they should share them.

It's befuddling because it's irrelevant: People who write packages will
bind commands to whatever keys they want to, and people who use those
package will either accept those bindings or not (by not using the
package, or rebinding the keys).

People are free to create whatever packages they want, with whatever key
bindings they wish, and people are free to use (or not use) those
packages.  It's not up to us, so discussing this stuff here is
irrelevant.

Now, anybody's also free to think that the bindings by package X are
stupid, and are free to say so: Freedom from being "tsked" at is not one
of the freedoms set out in the GNU Manifesto.  But suggesting that an
enormously successful and well-made package like Magit is doing
something wrong by binding `C-x g' is borderline rude.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: Discussions about key bindings

[tomas]

[-- Attachment #1: Type: text/plain, Size: 2074 bytes --]

On Fri, Feb 12, 2021 at 12:54:25PM +0100, Lars Ingebrigtsen wrote:
> I find the recent spate of discussions about key bindings for
> third-party code rather befuddling [...]

:-)

I must admit that, while I find that problem space quite
interesting, I lost track of most of that discussion.

I'm impressed by those folks still following along and
staying pretty constructive.

There are many dimensions to the problem. OTOH, *my* [1]
bindings are very close to *my* heart, so a hostile reaction
on *my* part whenever tries to take away some of that is,
at least, understandable.

On the other hand, if everyone has his/her private bindings,
communication becomes more difficult. Picture someone in
a forum writing "then hit C-x C-f" -- but IMO the problem
goes deeper; some kind of "UI fragmentation" occurs, where
part of the community perhaps doesn't /understand/ other
parts, because they're seeing (or experiencing) very different
things.

Somewhere between those two poles, life happens :-)

There are more dimensions, with poles at each end (keyboard
has been mentioned: CTRL-\ is easy on an US keyboard; I
have a German one, there it's awkward; What's important on
CTRL-Y: the label on the key, or the physical position on
the keyboard (for US people: in the middle of the upper
row, for Germans in the lower left). And so on. Up to now,
Emacs hasn't taken that into account. Should it? Should it
not? What about a Thai keyboard [2]?

Emacs, btw, has accumulated an exceptional history of keeping
together an incredibly diverse bunch. Watch other UI or "user
experience" discussions to see what I mean. I think Emacs
can be proud of that.

I don't have a good idea to make the current discussion more
fruitful, except perhaps that it might be good to take a
step back and try to get a grip of the problem in a more
abstract (and perhaps non-technical) way.

Cheers

[1] *my* used as a generic placeholder
[2] Kedmanee? Pattachote?
    https://en.wikipedia.org/wiki/Thai_Kedmanee_keyboard_layout
    https://en.wikipedia.org/wiki/Thai_Pattachote_keyboard_layout

 - t

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: Discussions about key bindings

[Joost Kremers]

On Fri, Feb 12 2021, Lars Ingebrigtsen wrote:
> I find the recent spate of discussions about key bindings for
> third-party code rather befuddling.  It's been suggested that Magit
> should move their bindings to `C-x v', and it's been suggested that
> there should be a number of bindings that should be reserved for
> third-party packages, and they should share them.

The discussion has veered in various directions, some of which irrelevant, but
the original question certainly is relevant, I think. Which is whether Emacs
should have a policy on 3rd-party packages creating global key bindings.

Your stance seems to be that the status quo is good enough, i.e., that it's not
necessary to have such a policy. Gregory feels differently, and so do I, though
in the opposite direction.

In the end, though, it's up to you and the other maintainers, so I'll take your
message as a sign to stop arguing this topic. :-)


-- 
Joost Kremers
Life has its moments