>> So what exactly is the problem? Your code example is faulty:
>> `Custos` grobs don't listen to the `direction` property. As you've
>> meanwhile found out, it's rather `neutral-position`.
>
> Yeah I was trying to say that it's a problem that Custos grobs don't
> listen to the direction property. Sorry if I wasn't clear, I'm bad
> at explaining things. I meant that Custos grobs SHOULD listen to
> the direction and default-direction properties (and/or that would be
> more intuitive than the current way to make them always point
> up).
Well, `direction` and `default-direction` are more coarse than
`neutral-position`, so there is a good reason not to listen to the
properties...
I consider this a documentation issue: it should perhaps be stated
somewhere that the `Custos` grob doesn't listen to `direction` and
`default-direction`.
> I meant that ly doesn't engrave custodes in that case but it
> probably should, see e.g. this edition of Vivaldi's 12 trio sonatas
> op 1 from 1705 (sonata 1, vln 2): [image: 2025-11-25 11.22.46
> vmirror.imslp.org b08421931c2e.png] (also an example of the later
> baroque tail-less custos glyph)
Looks convincing, so please open an issue!
Werner
