tuplet bracket collides with beam and goes in wrong direction

[Paul Scott]

I had been thinking that this bug was combined with another tuplet bug
but looking at the bug listings I see I was wrong.

This code produces a tuplet bracket that slopes the opposite way as it
should and collides with the beam.

\version "2.9.19"

\relative c' { ees16[ g \times 2/3 { r16 dis cis] } c4 d e f f e d }

I would consider sponsoring a fix if necessary as this bug keeps me from
delivering an arrangement unless I change the beaming.

Thanks,

Paul Scott