Pavel Sanda wrote:
Richard Heck wrote:
I guess I saw this the other way around: BOTH completion AND next-cell
should work in tables.
and both bound to tab?
Yes. And this does work if what's bound is completion-accept. If you
bind it to complete, then it works much less well.
Otherwise, you, the user, have to remember that
completion has no binding in tables. This is very unintuitive, and it is
especially nasty with mathed (which, for me, is the only place I use
completion), since next-cell also works there in arrays, etc.
...
Well, my version is essentially upside down of your version. ;-) But adding
listings is trivial, if the LFUN exists.
there must be some basic misunderstanding. i thought that complete is enabled
everywhere (maybe after some short time interval), so once bound as the first
one others will never be called.
Well, complete isn't always enabled---not in the middle of a word---but
it's almost always enabled, so you do have this problem. It's better
with completion-accept, since that is only enabled once a completion has
been shown. So you can bind that first. That's why I was proposing to
bind Tab to completion-accept rather than to complete. But I do
understand that one would like some nice binding to complete.
Richard
