Re: SMuFL name mapping update, 17 March

[Owen Lamb]

Hi Benkő,

On 3/19/22 02:00, Benkő Pál wrote:
> Hi Owen,
>
> Owen Lamb<owendlamb@gmail.com>  ezt írta (időpont: 2022. márc. 19., Szo, 7:38):
> > Hi Benkő,
> >
> > On 3/17/22 10:05, Benkő Pál wrote:
> >
> > re the longest rests:
> > 1. a perfect longa rest and a maxima rest are not the same;
> > 2. a perfect longa rest should take three spaces (like current
> > emmentaler rests.M3mensural), not four (as, I fear, the bravura
> > mensuralRestLongaPerfecta implies)
> > 3. a maxima rest consists of two (or three) longa rests, similar to
> > rests.M3neomensural, so much that theoretically one can't tell a
> > maxima rest from two (three) longa rests.  renaissance version of
> > multimeasure rests use groups of longa rests at different staff
> > positions, like
> > https://upload.wikimedia.org/wikipedia/commons/thumb/a/aa/Ockeghem_Prolationum_Kyrie.jpg/220px-Ockeghem_Prolationum_Kyrie.jpg
> >
> > OK. Then I think rests.M3neomensural should be given the SMuFL name 
> > neomensuralRestMaxima, and shouldn't be an alternate of anything. (It seems 
> > like a bad idea that both the longa perfecta and the maxima are labeled with 
> > the M3 duration in LilyPond, when in fact they aren't interchangeable...)
> I agree with the naming; I'm confused about potential implications of usage.

What I mean is that LilyPond's current naming system led me to think the maxima 
and longa perfecta rests were interchangeable stylistic variants: they have the 
same nominal duration (M3) and are each exclusive to a particular style 
(mensural or neomensural). I'm only worried that LilyPond's current naming 
scheme might continue to encourage the same error, especially if they are 
switching between the mensural and neomensural styles.

> > I can't check right now, but I think LilyPond doesn't use longa glyphs
> > in ligatures: longae are drawn by adding stems to brevis glyphs.
> >
> > Terminology gets a bit muddled around here. To clarify, are you talking about 
> > the glyph-on-glyph ligatures found throughout the SMuFL specification, or the 
> > note-on-note ligatures characteristic of mensural notation?
> the latter.  I haven't realized the ambiguity (the existence of
> ligatura variants of clefs should have made me think a bit though).

No worries!

I took a look at void Mensural_ligature_engraver::transform_heads 
(vector<Item *> const &primitives) (line 90, 
mensural-ligature-engraver.cc). It looks like a longa notehead smob's 
primitive property (called prim before it's passed into Scheme) does 
sometimes turn on the MLP_BREVIS flag when descending at the beginning 
or the end of a ligature, but in all other (non-flexa) cases (see if 
(general-case), line 279), it turns on the MLP_LONGA flag instead, which 
always results in a longa glyph when passed into mensural-ligature.cc 
(see Stencil internal_brew_primitive (Grob *me), in particular following 
int duration_log until the glyph name is generated). So, seems to me we 
still use noteheads.sr?M2ligmensural for those mid-ligature longae.

Owen