Damien Cassou <damien@cassou.me> writes:

> thanks for your suggestion. I applied it in attached patch. I'm 
> not sure I like it though because the simplest call:
>
>     (current-line)
>
> is now exactly the same as (line-number-at-pos). To get a 
> different behavior, non-nil must be passed as argument:

Well, it would be the same as (line-number-at-pos (point)), yes, but I
think it's the most consistent implementation WRT `current-column'.

Cheers,
Nico