<<< Date Index >>>     <<< Thread Index >>>

Re: mutt/1880: wish: rebind to generic



The following reply was made to PR mutt/1880; it has been noted by GNATS.

From: TAKAHASHI Tamotsu <ttakah@xxxxxxxxxxxxxxxxx>
To: bug-any@xxxxxxxxxxxxx
Cc: 
Subject: Re: mutt/1880: wish: rebind to generic
Date: Fri, 23 Sep 2005 21:14:08 +0900

 Hi Alain,
 
 * Fri Sep 23 2005 Alain Bench <veronatif@xxxxxxx>
 >   On Tuesday, September 20, 2005 at 1:56:59 AM +0200, Tamotsu Takahashi 
 > wrote:
 >  
 >  > http://www.momonga-linux.org/~tamo/patch-1.5.8.tamo.unbind.1 This
 >  > ``unbind'' might be less useful than DGC's one, because this does not
 >  > support ``unbind *'' syntax.
 >  
 >      We probably want "unbind *" functionality, if not syntax.
 
 You can make a file with a lot of unbinds and source it.
 (No, this is not a joke. I'm serious.)
 
 I don't like "*" in unbind, because it needs some tricks
 as DGC's one does. (e.g. recovering ":" and editor bindings)
 
 And I don't know how to unbind literal "*" key with DGC's.
 
 
 >  > This ``unbind'' really removes mappings instead of ``bind to noop'',
 >  > so that a little memory is freed.
 >  
 >      Fine! I seem to recall DGC removes mapping only in * case, not in
 >  individual key sequence case.
 
 But I am not sure my patch is bug-free.
 Seems too complicated for me.
 The small amount of memory does not worth the risk. ;)
 
 
 >      It seems to clearly be a wish for a new feature.
 
 Yes, but this ``new'' feature is what mutt must have,
 from the beginning. Otherwise mutt must not have
 pre-defined keymappings. (i.e. Instead of hard-coding
 them, use /etc/Muttrc to define initial mappings.
 That could decrease hundreds of lines from the code.)
 
 Without this feature, "generic" is not useful for many
 keys, because they are already hard-coded on some menu.
 
 
 >      Manual says:
 >  
 >  | Usage: bind map key function
 >  | The special function noop unbinds the specified key sequence.
 >  
 >      I agree s/unbinds/disables/ may help desambiguitation. But then
 >  reintroducing ambiguity by adding an "unbind" command is not good.
 
 So, as I once said, "bind <map> <key> generic" may be
 the best solution. It's easy and clean.
 
 -- 
 tamo