On Mon, Mar 25, 2024 at 11:47 AM Werner LEMBERG <w...@gnu.org> wrote:
>
> >> Consider this small example.
> >> ```tex
> >> {
> >> f'8 f' \autoBeamOff f' f'
> >> f' f' f' f'
> >> }
> >> ```
> >>
> >> For me it was surprising to see that `\autoBeamOff` doesn't act
> >> immediately but rather two eighths later. The documentation
> >> doesn't mention this, AFAICS, and I consider this behaviour rather
> >> counter-intuitive. Could this be fixed?
> >
> > It appears that intentionally the `autoBeaming` property is not
> > consulted when considering whether to end a auto-beam in progress.
>
> Thanks for digging! I've now done some `git blame` calls and found
> out that this behaviour was there from the very beginning of the
> auto-beamer's existence (in 1999). Honestly, I don't see any reason
> for it – who likes delayed actions like this?
>
> > Using \noBeam should let you terminate the beam early: [...]
>
> Very nice! Maybe we should add your fix to the git repository?
>
>
Seems to me like we need a documentation fix. Something like:
A beam that is started by the autobeamer will end according to the
autobeaming rules in
effect at the start of the beam. If you wish to terminate an autobeam
early, use \noBeam after
the last note you want beamed.
It may be possible to change the autobeam behavior, but it would likely
increase the computation
time for autobeaming, because you would have to continually look at both
the start and the end of the beam.
So I'd propose that making a function change could be a feature request,
but the lack of documentation
is a bug.
Carl
