Re: line-move-visual

[Johan Myréen]

On Friday 10 July 2009 09:12:23 Stephen J. Turnbull wrote:

> Hey!  Not that way!  By attracting a New Generation of users who have
> never seen the "carriage" that "carriage return" refers to.  "What,
> were typewriters horse-powered?" ;-)

Ok, I seem to have stirred up some conversation here, and at the same time 
made me appear a grumpy old man. I hope to present more rational 
arguments in this message.

I still haven't quite figured out what the goal here is:

(1)  to keep the mental model of a buffer containing logical lines, but change 
the behaviour of C-n when moving within occasional continued lines

or

(2)  to move away from the "lines" model to a "paragraph" model, where there 
are no logical lines, but Emacs presents the paragraph using visual lines, 
like a word processor does

If (2) is the goal then ok, the behaviour of C-n makes sense. But you should 
go all the way and change the behaviour of C-a, C-e, C-k and probably a lot of 
other stuff too. You should also get rid of the continuation marks, and make 
the lines wrap at a more sensible place than the right edge of the window. 
Wrapping should also occur at word boundaries.

Also note that you can't completely get rid of the lines model: programming 
language source code files consist of (logical) lines. The above mentioned New 
Generation users who haven't seen the "carriage" still bang on their Carriage 
Return keys even in more modern environments than Emacs, like Eclipse or 
Visual Studio.

If (1) is the goal, I don't see how this change makes editing that much more 
convenient. Continued lines are normally quite rare, and I find movement by 
visual lines to be both nonintuitive and inconsistent with the behaviour of C-
a and C-e. (By the way, if the behaviour of C-e would be changed to respect 
visual lines, what would it do--move the insertion point _in front_ of the 
last character of the visual line?)

Johan Myreen