Re: movement bindings in c-mode

[Alan Mackenzie]

Hi, David!

On Mon, Jun 23, 2008 at 08:36:14PM +0200, David Hansen wrote:
> On Mon, 23 Jun 2008 17:05:53 +0000 Alan Mackenzie wrote:

> > Just one small point, though.  The doc-string for
> > beginning-of-defun-function recommends putting point at the { rather
> > than the function header.  I think this is unright.  Stefan?

> Actually I would prefer to have it in front of the `{'.  If you just
> want to read something around the b-o-d it doesn't matter.

It matters, I think, when you do C-x n d.  Then you can see what
function you're within.

> But if you want to edit something it's far more likely that it is within
> the body, e.g. adding a variable, than the header.

I'd conjecture that any line within a defun is as likely to be edited as
any other.  That you've got an extra 3 or 4 lines to move forward to
that point doesn't seem that critical.

But I think the critical reason is that the command is called
beginning-of-defun, not beginning-of-defun-block, and commands should do
what they say.  It would be possible to modify b-o-d so that a bare
prefix (C-u) would go to the block rather than the header.  But I don't
think it would be worth the extra complication.

> David

-- 
Alan Mackenzie (Nuremberg, Germany).