Re: Is Elisp really that slow?

[Stefan Monnier <monnier@iro.umontreal.ca>]

> And, as already mentioned, comment-dwim is not an alternative to
> comment-region. They do different things.

I beg to differ.

> At the end, we deplete the available bindings from each mode just for
> the cause of coherence.

comment-region is not specific to C-mode, so if we want to bind it (to
C-c C-c or something else), it should be done elsewhere
(e.g. prog-mode).

The way things work in Emacs is that major-modes are written by users.
Those don't directly care about distinguishing what is really specific
to this major-mode from what is rather a personal preference or what
would be a generally useful functionality: they just want to get their
major mode to behave in a particular way.

Then, over time, as this major mode and its behavior is exposed to other
users, its maintainer(s) learn to distinguish between these differences,
so they remove some settings altogether (which proved to be personal
preferences of the author, typical examples being comment-indent,
indent-tabs-mode, word-syntax for the _ character, ...), and repackage
others as separate packages usable in different major modes
(e.g. tab-always-indent, electric-indent-mode, comment-region, smie,
...).

This second step is not necessarily in the best short-term interest of
users of this mode (not only because change is annoying but also
because often as feature are made usable in several modes they end up
dropping some very specific ad-hoc behavior which is too hard to
support in a generic way, ...), but it is the best long term interest of
Emacs as a whole.


        Stefan