[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Fwd: Changes for emacs 28

From: Göktuğ Kayaalp
Subject: Fwd: Changes for emacs 28
Date: Tue, 08 Sep 2020 19:38:53 +0300
User-agent: mu4e 1.2.0; emacs 28.0.50

[ I’ve originally sent this reply using the reply button on the web
  interface, so it probably ended up as a private reply.  Sending it
  again to the mailing list, hopefully will end up showing up under the
  correct subthread.  Sorry if this somehow inconveniences anybody. ]

> WDYT??

Every single one of the suggested changes are breaking changes.

Emacs is essentially a programming environment, so all the packages and
themes are built either directly on its defaults or on other packages
and defaults.  When you e.g. change the default theme, you’ll break all
the other themes because they all build on the default theme. Or you’ll
break people’s configurations.

I’m wary of this ‘attract users’ rhetoric that’s rather popular lately
in Emacs discussions.  A software package should service its existing
users first.  Newcomers either come because they like the thing or they
use something that serves them better.  We’re not in competition with
other editors.  And chasing new users at the cost of annoying existing
users is potentially disastrous. See e.g. the sad demise of Firefox and

Also, breaking things has a huge cost in introducing schisms that are
hard to overcome. See e.g. Python 2 vs. 3 shenanigans which continue to
this day.

With Emacs, there’s the possibility of "canned configurations" like Doom
and Spacemacs, and IMHO having them exist as a separate thing is very
healthy.  One nice thing could be to advertise them on
https://gnu.org/s/emacs as potential entry points to Emacs, and help
them document their packages better.  Also maybe modifying the usual
welcome screen to note that Emacs is decades-old software and the user
might want to set some "better defaults".  Or, maybe adding a sort of
"21st-century-mode" which enables many breaking customisations that are
conventionally accepted as better defaults, but explicitly.

In sum, move slow and _never_ break things.  The main reason I stick
with Emacs is that I _know_ it won’t be pulling the rug from under my
setup in this day and age of just breaking everything and telling people
to GTFO if they ask for some stability and backwards compatibility.

İ. Göktuğ Kayaalp / @cadadr / <https://www.gkayaalp.com/>
pgp:   024C 30DD 597D 142B 49AC 40EB 465C D949 B101 2427

reply via email to

[Prev in Thread] Current Thread [Next in Thread]