Re: PATCH: Re: (forw) --help able programs and completion]

[Sven Wischnowsky <wischnow@informatik.hu-berlin.de>]

Bart Schaefer wrote:

> On Jun 27,  5:02pm, Sven Wischnowsky wrote:
> > Subject: PATCH: Re: (forw) --help able programs and completion]
> > 
> > Bart Schaefer wrote:
> > 
> > > What I was expecting, for some reason, was that menu selection would start
> > > immediately if there were more than N matches, not wait for the second TAB
> > > that triggers automenu.
> > 
> > Ah. No, but that's easy to add.
> 
> Er, thanks, but you didn't really need to add that; all I wanted was a
> yes-or-no answer ...

It was fun writing it (and I needed something to distract me from real 
work).

> > I used the same form as for `select', i.e. one can say `yes=3' to turn 
> > on menu completion when there are at least three matches.
> 
> Hrm.
> 	zstyle ':completion:*' menu 1=3
> 
> Does that look a bit strange to anyone else?

Yes, to me (I wondered if I should point that out yesterday). But with 
the select[=val] in place this looked like a easy-to-understand
enhancement.

Another question is if it should mean `if there are less than <num>
matches, start menu completion', because menu completion (at least the 
*real* menu completion, not selection) works best with few matches.

That's easy to change, though (I'd like to get opinions from menu
completion users).

> > Does anybody now want support for `no=3'?
> 
> Which would mean ...?

If one has `setopt menucomplete', setting the menu style to one of the 
`false' values can be used to turn menu completion off for some
contexts. Allowing `no=<num>' would either mean to turn it of if there 
are more or less than <num> matches, whatever we decide `yes=<num>'
means.

> However, a plain-ol'-menu-completion variant of select=long-list might be
> interesting.

Hmhm, right.

Bye
 Sven


--
Sven Wischnowsky                         wischnow@informatik.hu-berlin.de