Re: Is M-j broken for you in Org on Emacs 27 and 28?

[Colin Baxter 😺]

>>>>> Richard Lawrence <richard.lawrence@uni-tuebingen.de> writes:

    > Hi Org community, Some questions for those of you on Emacs 27 and
    > 28:

    > Does M-j in an org-mode buffer do what you expect?  Does it throw
    > an error?  What function is M-j bound to in Org?

    > Backstory:

    > I have long been on Emacs 26.3 (in Debian stable) but recently
    > decided to try a newer Emacs from GNU Guix. I immediately ran into
    > an issue, and now I'm trying to figure out if it's a bug, and if
    > so where to file it: in both Emacs 27.2 and 28.0.50, typing M-j in
    > an Org buffer throws (wrong-type-argument char-or-string-p nil)

    > The source of this problem appears to be that the keybinding for
    > M-j changed between Emacs 26 and 27. In Emacs 26 it calls
    > indent-new-comment-line. In Emacs 27 and 28 it calls
    > default-indent-new-line, and the call stack look like:

    >   insert-before-markers-and-inherit(nil)
    > org-comment-line-break-function(nil) default-indent-new-line()
    > funcall-interactively(default-indent-new-line)
    > call-interactively(default-indent-new-line nil nil)
    > command-execute(default-indent-new-line)

    > The error arises because insert-before-markers-and-inherit cannot
    > accept nil (the value of fill-prefix in this context).

    > I see this error in emacs -q with both Emacs 27 and 28 from Guix.

    > After some investigation, the functions involved here don't appear
    > to have changed at all recently (though see [1]); just the
    > binding. This leads me to ask: why hasn't this been discovered
    > already? Which leads me to wonder if I am using M-j in some
    > non-standard way.

    > Some time in the distant past, I internalized the idea that M-j is
    > a better way to type a newline because (a) it doesn't involve a
    > pinky reach and (b) in most contexts in Emacs, it is more likely
    > to "do what I mean" than RET is. In particular, it continues
    > comments and indents properly. (I am also an evil-mode user and
    > there is probably some part of my brain that thinks "M-j is like j
    > for insert mode".) But maybe that was always wrong, and the
    > recently changed binding is just an indication that I was not
    > using M-j as intended.

    > So which is it? Is this a bug in Emacs, in Org, or in my ingrained
    > typing habits? Many thanks for your advice.

    > -- Best, Richard

    > [1] There is a commit which changed default-indent-new-line in
    > August of this year, but the changes don't seem relevant to the
    > error I'm seeing, since I also see it in Emacs 27.2:

    > 
https://git.savannah.gnu.org/cgit/emacs.git/commit/?id=b41f31d2b60269bd0e7addd1081f3738f91e76bc

I confirm that it also appears broken to me in emacs-27.2, with the same
error as you found. I have never noticed it before, possibly because I
use C-j rather than M-j.

Best wishes,