bug#16071: Erc re-binds [home] even if not beginning-of-line

bug#16071: Erc re-binds [home] even if not beginning-of-line

[Ivan Shmakov]

Package:  emacs
Severity: minor

	As of 0dd4ebbcf94d, erc-mode-map unconditionally binds both C-a
	and Home to erc-bol.  Which is problematic, as:

	• either (or both) of them may happen to /not/ be bound to
	  beginning-of-line (as, for example, when using an older,
	  Emacs 20-or-so, binding of Home to beginning-of-buffer);

	• the user may’ve previously bound additional key sequences to
	  beginning-of-line, which are not rebound by the current
	  erc-mode-map.

	It’s arguably more semantically correct (and convenient) to
	re-map beginning-of-line to erc-bol, instead of using an
	explicit list of key sequences to be bound to erc-bol.

-- 
FSF associate member #7257

bug#16071: Erc re-binds [home] even if not beginning-of-line

[Stefan Monnier]

> 	As of 0dd4ebbcf94d, erc-mode-map unconditionally binds both C-a
> 	and Home to erc-bol.  Which is problematic, as:

Good points.

But please note that doing it with remap is intrusive because the user
may actually like to have a key bound to beginning-of-line.

I guess rebinding with substitute-key-definition is OK.


        Stefan

bug#16071: Erc re-binds [home] even if not beginning-of-line

[Ivan Shmakov]

>>>>> Stefan Monnier <monnier@iro.umontreal.ca> writes:

 >> As of 0dd4ebbcf94d, erc-mode-map unconditionally binds both C-a and
 >> Home to erc-bol.  Which is problematic, as:

 > Good points.

 > But please note that doing it with remap is intrusive because the
 > user may actually like to have a key bound to beginning-of-line.

	Indeed, it’s intrusive either way.

 > I guess rebinding with substitute-key-definition is OK.

	Does that mean binding only to the sequences which are
	beginning-of-line in global-map /and/ also not overridden by the
	user in erc-mode-map, or something like that?  As per my reading
	of substitute-key-definition, it implies more or less the same
	effect as the use of [remap] with define-key.

-- 
FSF associate member #7257

bug#16071: Erc re-binds [home] even if not beginning-of-line

[Stefan Monnier]

>> I guess rebinding with substitute-key-definition is OK.
> 	Does that mean binding only to the sequences which are
> 	beginning-of-line in global-map /and/ also not overridden by the
> 	user in erc-mode-map, or something like that?

Yes.

> 	As per my reading of substitute-key-definition, it implies more
> 	or less the same effect as the use of [remap] with define-key.

No: people can still bind a key to beginning-of-line in erc-mode-map and
it won't be remapped.


        Stefan

bug#16071: Erc re-binds [home] even if not beginning-of-line

[Lars Ingebrigtsen]

Ivan Shmakov <ivan@siamics.net> writes:

> As of 0dd4ebbcf94d, erc-mode-map unconditionally binds both C-a
> and Home to erc-bol.  Which is problematic, as:
>
> • either (or both) of them may happen to /not/ be bound to
> beginning-of-line (as, for example, when using an older,
> Emacs 20-or-so, binding of Home to beginning-of-buffer);
>
> • the user may’ve previously bound additional key sequences to
> beginning-of-line, which are not rebound by the current
> erc-mode-map.
>
> It’s arguably more semantically correct (and convenient) to
> re-map beginning-of-line to erc-bol, instead of using an
> explicit list of key sequences to be bound to erc-bol.

Well, I think that users who want to rebind keys in a mode should use
the mode hooks -- there's nothing special about beginning-of-line and
friends.  So I'm closing this report.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no