Bastien <b...@gnu.org> writes:
> Indentation is quite sensitive: what do you think of setting a new
> default value of nil for `org-adapt-indentation' in Org 9.5?
I think that makes sense. If the controversy over the
electric-indent-mode change taught me anything, it's that a lot of
people don't expect their Org files to be hard-indented, which I fully
sympathize with (I personally soft-indent with org-indent-mode).
I don't know whether the folks that protested over the change represent
the majority of Org users, but FWIW:
- That position sounds consistent with the Org files we see in the
org-mode repo (where .dir-locals.el sets the variable to nil).
- I expect that the resulting behaviour matches what users were used to:
before Org 9.4, I imagine most folks would hit RET, get unindented
text, and not think twice about it. People who preferred indentation
would have had to go out of their way to get it, by always hitting TAB
or C-j (which is the "dumb, unindented newline" in other modes which
honor electric-indent-mode); I'd be surprised if many people worked
this way[1].
'headline-data sounds like a reasonable default too, although I think it
still has some wrinkles[2].
Thanks for following up on this, and thanks again for your stewardship.
[1] Although I'm glad to learn that the current state of affairs made at
least one user happy 🙂
<s6mr6r$17cg$1...@ciao.gmane.io>
https://orgmode.org/list/s6mr6r$17cg$1...@ciao.gmane.io/t/#u
[2] Typing "* headline RET" starts an indented line; further RETs keep
point indented until I type in something, after which RET finally
snaps back to column 0. I'd expect point to always land on column 0
when hitting RET after a headline, since "headline data"
(e.g. :PROPERTIES:, :LOGBOOK:) will probably always be entered
automatically via a command.
