* Frank Terbeck <bts@bts.grml.org>
> Well, I hope I'm not starting to nitpick too heavily.
You're a hard man to please. (o; But it's absolutely fine with me. In
fact, I appreciate your constructive criticism and all the effort
you're going through with your responses.
> But since I do like the way you're writing nice and lengthy commit
> messages, I think it's okay to see whether we can can them just
> right[tm]. :-)
Doing my best. But I'm aware of the fact that I can get rather complex
when writing in english. Therefore I'm really glad there's someone
reviewing the yackety-yak I produce.

> I don't like the wording. I know we're using 'power completion'
> elsewhere, too. But it's really just an expansion. There is *nothing*
> that's completed at all.
ACK.
> > Take '|& grep --color=auto ' for instance. Mind the trailing space.
> I guess I'd just drop this line altogether. It doesn't add much of
> value.
ACK.

> The patch itself looks fine. An updated patch with a cleaned up commit
> msg would be terrific, though. :)
Will do. But, before any further investigation, let's wait for mika's
point of view regarding the various expansions you've mentioned.

> However, I *seriously* don't know why the following are in the
> $abk[] hash at al. Where's the point in having those in addition to
> simple aliases?
> [...]
> The only reason I see would be to be able to expand aliases at will.
> Then again: are we adding new aliases to this list? Did we do it in
> the past, after we got this expansion? Can't compsys's _expand do
> something like that for us?
That's an interesting idea.

I'm not really using these expanѕion either -- exepct `e,.' for `print
-l' or a modified `mp' for `mplayer -vo xv -xy ' to specify width and
height -- and therefore I agree that several of these expansions are
rather questionable.
And a line such as
> >      'fori' 'for i ({..}) { }'
would IMHO be more useful when implemented as a widget.

But as I mentioned above, let's hear mika's point of view.