Re: master e4df903f6b5 2/2: ; Fix last change

Richard Stallman <[email protected]> writes:

> That would not be better -- it would cause the same sort of problem
> that binding a function cell causes.  Indeed, the Emacs Lisp Ref [...]

I'm partially with Richard here: it's tempting to bind function cells -
but in Emacs it's bad style most of the time.  There could be a warning
for people trying to bind a function cell with `cl-letf', and a less
obvious way without warnings for the cases where it is unavoidable (but
if that would prevent people misusing it is a different question...)

Because, Richard, my own config is full of bindings of function cells.
When authors (of Emacs, of third party packages) did not foresee how I
want to tweak a behavior, this is often the cleanest way (apart from M-x
bug report RET).

The majority of uses in Emacs itself seem to be of kind (symbol-function
#'FOO) -> #'ignore.  BUT note that by far not all cases are hacks at
all: some want to evaluate an expression in a context where certain
function names are temporarily provided.  Names that are not even in use
outside of that context.


Michael.