Re: Poll: Who is using these commands

[Dan Davison <davison@stats.ox.ac.uk>]

Leo <sdl.web@gmail.com> writes:

> On 2010-05-09 16:59 +0100, Nick Dokos wrote:
>> I disagree: they are not parenthesis movement bindings - they are
>> structure-navigation bindings. For example, C-M-f is forward-sexp.
>> In lisp, an sexp has some relationship to parentheses, but it is
>> incidental; in other programming modes, an sexp is whatever makes
>> sense in that language and these commands are redefined appropriately.
>
> Perhaps you haven't noticed. SEXP is a useful abstract. For example,
> it allows you to move across some_long_function_name in C and even in
> the message-mode I'm currently using, not just parenthesis. Situation
> like this will arise when editing org files too.

Yes, so in other words C-M-f can be used to jump over atoms, and in many
modes that has the useful effect of jumping over words containing '_'
'-' etc. I haven't seen any disagreement yet over the general principle
of treating headings as SEXPs, so I think the issue here is: is Org
happy to declare that a childless heading is atomic? And if we are happy
with that, are we left with a convenient way to skip over
something-like-this or something_like_this when they occur in Org-mode?

An alternative view would be that when point is at the beginning of a
heading C-M-f skips over the subtree, and otherwise C-M-f skips over one
"Org atom", however defined.

Dan



> It is a key binding that you
> can rely on in various modes and they happen to do the right thing.
>
> They are not re-defined, in most modes once you have a proper syntax
> table, they just work. On the other hand, the defun abstraction is not
> as universal as sexp so redefine them is fine.
>
> C-M-f and C-M-b are keys that I use extensively.
>
> I haven't used C-M-n and C-M-p much.
>
>> I think it is entirely appropriate to use these bindings to navigate
>> structure in org-mode as well.
>
> I am not against binding suitable keys to structure movement.
>
> Leo