RE: read-face-name PROMPT arg should be self-contained, including ": "

[Drew Adams]

> >> `read-face-name' still blindly appends ": " to the prompt it
> >> is supplied.
> >>
> >> I thought of `read-face-name' as an internal function, not for
> >> users to use.
> >> But I agree that it could be useful for user code to call.
> >> So I'm not against making this change.  I see that all the callers
> >> within Emacs are easy to change.
> >>
> >> If someone wants to do it, please do it.
>
> > Thanks, I hope someone does.
>
> > BTW, for sometime after the release, we might consider adding
> > an optional history-list arg to `read-face-name', so accessing
> > history entries doesn't include non-face stuff.
>
> I was recently thinking that rather than adding new history vars
> everywhere
> all the time, we could change the history navigation to automatically skip
> entries which are not completion candidates (at least for those cases that
> are `must-match').

Hmm, maybe. It's worth thinking about, anyway.

Considerations:

. That would tie history cycling to the current typed input, rather than
letting you access any history entry. That is, history cycling would be
among matches only. Would that be good or bad? For non-`must-match', at
least, that might be a nuisance, requiring you to first clear the
minibuffer.

. We already have one way to match against the history, though it is
limited, giving you only the first match.

. It would not really be "rather than", because a history var, if available,
is more efficient. IOW, it would be good, in any case, to _allow_ use of a
history var.