Re: GNU Emacs raison d'etre

[Karl Fogel]

On 12 May 2020, excalamus--- via "Emacs development discussions." wrote:
>May 11, 2020, 23:12 by address@hidden:
>What are we competing for?  I feel that while other threads are
>examining "missing features", it would be helpful to examine what GNU
>Emacs does offer.  Not only in software features, but maybe also in
>philosophy, community, or tradition.
>
>What is it about GNU Emacs that makes this mailing list bustle with
>enthusiasm?  Other editors use GPL, provide source code, have
>documentation, are customizable, and extendable.  There's something
>in how GNU Emacs implements these that is different.  I feel like
>there are taters to find if we dig a little.
>
>Is it because Emacs Lisp is unique to Emacs that Emacs teaches as
>well as documents?
>Is it that by being a pseudo-Lisp machine, Emacs puts users in the
>zone of proximal development?
>Is GNU Emacs the best embodiment of the GNU philosophy? 

Sure, I'll take the bait:

To the best of my knowledge, no other editing environment rewards sustained 
user investment so well.

With Emacs, if you keep investing -- i.e., acquiring knowledge and skill by 
reading documentation, writing customizations, and exploring others' 
customizations -- Emacs keeps rewarding you with a better and better editing 
experience.  The degree to which it does this seems normal to many of us here, 
because we've been used to it for many years.  I think we sometimes fail to 
appreciate the degree to which non-users, potential ("Emacs-curious") users, 
and even many actual new users are *not* aware of it: they don't realize how 
enormous the reward can be, and how broad its scope.

This should probably affect how we think about promoting Emacs.  Emacs 
shouldn't necessarily try to attract everyone who needs to edit text [1].  Many 
people who edit text nonetheless don't view text editing as a primary activity 
worthy of investment.  Those users are not good candidates for Emacs.

Emacs's best prospects are with the sorts of people who *do* see -- or who can 
be persuaded to see -- text editing as worthy of investment.  There's a loose 
correlation in which good programmers tend to be those sorts of people, because 
good programmers are usually willing to invest in learning their tools in 
general.  E.g., they'll learn their text editor the same way they'll learn 
their debugger, their programming framework, etc.  But the set isn't limited to 
just programmers.  For example, scientists and other academics who edit LaTeX 
documents are often good candidates for Emacs usage, because by both 
temperament and life situation they are well-positioned to understand how 
sustained investment in learning their editing environment could pay off in the 
long term.

So I suggest that GNU Emacs's raison d'être is to be the text editor that best 
rewards sustained user investment.

I think Emacs actually does so right now, too, and that we just haven't always 
communicated this fact clearly enough.

Thus, instead of focusing on making Emacs easier for new users, it would be 
better to focus on smoothing out discontinuities in Emacs' investment-reward 
curve.  The long-term health of Emacs as a project will not come from a large 
number of lightly committed users who don't appreciate what makes Emacs unique, 
but rather from a smaller number of users for whom Emacs is important and 
irreplaceable.

I'm not suggesting that we shouldn't improve the new-user experience in Emacs, 
of course.  We should make it as easy as possible for newcomers *while still 
prioritizing invested users*.   In user experience design, there are frequently 
tradeoffs between making things easy for newcomers and making them rewarding 
for experts.  Unfortunately, too often in design discussions, the new user 
experience automatically wins out -- it's like some kind of magic card that 
people play (even sometimes unconsciously) in UI/UX discussions.  For Emacs, 
this would be a mistake.  Emacs's great strength will never be in its new-user 
experience, and this is in some ways a necessary consequence of Emacs being so 
great for highly invested long-term users.

This also suggests that the sorts of features that highly-invested users tend 
to want -- for example, LSP-based features -- should be more important to us 
than how square the menus are or what menu items are shown in a default startup 
configuration.  When we make decisions that disappoint the core user base, we 
endanger the project much more than when we make decisions that disappoint 
users (or potential users) who weren't likely to become highly invested anyway.

(The fact that Emacs promotes free software by being a good GPL'd program is 
nice too, and is important to many of us, but it's not unique to Emacs.)

Best regards,
-Karl