Re: [Emacs-diffs] master e82fd95: ; Comments.

Re: [Emacs-diffs] master e82fd95: ; Comments.

[Stefan Monnier]

> +;; We can't straightforwardly make this derive from text-mode, because
> +;; we need to bind (rmail-buffer-swapped) around the text-mode call.  :(
>  (defun rmail-edit-mode ()

Can't we just make a new `rmail-enter-edit-mode' function which does the
rmail-buffer-swapped dance and then calls `rmail-edit-mode' which can
then be a plain old derived mode?


        Stefan

Re: [Emacs-diffs] master e82fd95: ; Comments.

[Glenn Morris]

Stefan Monnier wrote:

>> +;; We can't straightforwardly make this derive from text-mode, because
>> +;; we need to bind (rmail-buffer-swapped) around the text-mode call.  :(
>>  (defun rmail-edit-mode ()
>
> Can't we just make a new `rmail-enter-edit-mode' function which does the
> rmail-buffer-swapped dance and then calls `rmail-edit-mode' which can
> then be a plain old derived mode?

So anyone who just called 'rmail-edit-mode' would get unexpected results?
Sounds confusing to me.
But if you can find a way to make it work, please go for it!

Re: [Emacs-diffs] master e82fd95: ; Comments.

[Stefan Monnier]

> So anyone who just called 'rmail-edit-mode' would get unexpected results?

I guess so.  Not being an Rmail user at all, it's hard for me to judge
what that means.  Do we expect end-users to call rmail-edit-mode via M-x
or via some external Elisp snippet?
If so, maybe we could just add to rmail-edit-mode some appropriate check
that the buffer-swapping was done as needed: it wouldn't make it work as
before, but it might at least signal an error before it goes on to do
something potentially dangerous?

> But if you can find a way to make it work, please go for it!

I'm 99.9% unfamiliar with Rmail, both as a user and as a programmer, so
I'd prefer if someone else does it (e.g. someone who might actually
know how to test the resulting code, for example).


        Stefan