Re: rehearsal marks: separate from content encoding and display above strings

[David Wright]

On Mon 15 Aug 2016 at 17:20:09 (+0200), Johan Vromans wrote:
> On Mon, 15 Aug 2016 09:39:39 -0500
> David Wright <address@hidden> wrote:
> 
> > I'm sorry, but your wishlist has nothing to do with the statement
> > "Specifying the structure of a score in \global is wonderful, it just
> > doesn't work with midi."
> 
> There is something called the principle of least surprise. I can put bars,
> rehearsal marks, time signatures, repeats in \global and they all end up
> nicely in all staves of the printed output.
> 
> Then it surprises me that they do not all end up nicely in the midi — in
> particular, the repeats do not. So my first thought is "something doesn't
> work". That there is a good explanation [thanks, David K] why it
> doesn't/cannot work as I would have expected it doesn't change the surprise.
> 
> I tried to make clear that I do not consider this to be a bug
> but that it would be 'nice to have'. I see nothing wrong with that.
> 
> > If you're going to make a statement like that, and be challenged,
> > please supply some evidence or withdraw the remark.
> 
> I change my remark to: "Specifying the structure of a score in \global is
> wonderful, it just doesn't play nice with MIDI, because no voice can impose
> a MIDI repeat structure on any other."
> 
> > It's not as if this is the first time you've said that writing
> > repeats in LP doesn't work in the MIDI output:
> > http://lists.gnu.org/archive/html/lilypond-user/2016-02/msg00284.html
> 
> In this article I wrote "LilyPond doesn't deal with segno, coda, d.s. etc
> either in the MIDI." Apart from being a totally different topic, would
> you challenge this? Aand also, I wrote it as a statement, not a complaint
> and definitely not "reports of a LP failure."

That sentence was in a paragraph within a thread. Looking backwards:

...you wrote "LilyPond doesn't deal with segno, coda, d.s. etc either
in the MIDI. The LP generated MIDI only matches the sheet music in
most but definitely not all cases."

...commenting on my "This will unnecessarily prevent LilyPond using
the \repeat volta construction and so, once again, no MIDI rendition
among other concerns,"

...where that construction was what I presented as rpt1 in
http://lists.gnu.org/archive/html/lilypond-user/2016-02/msg00260.html
and this complete example would give the correct MIDI output.

I criticised your suggestion because it would not only produce an
unconventional score, but it couldn't handle the MIDI automatically.
So implying that LP's MIDI only matches the music in some cases (for
reasons that are never spelled out) by throwing in a few distractions
(segno, coda, d.s., which are *markings*, not repeats) is a way of
tempering my criticism.

That's what you just left out of your *full* paragraph above.
My contention there was that the volta construction can produce the
correct score and the correct midi when applied correctly (ie to all
the voices). The repeats are correctly observed in the MIDI.

This present thread discusses the same topics: repeats and MIDI.
http://lists.gnu.org/archive/html/lilypond-user/2016-08/msg00222.html
"For midi, \unfoldRepeats is necessary. Last time I tried, the repeats
(in \global) were note applied to the music parts.
Specifying the structure of a score in \global is wonderful, it just
doesn't work with midi."

The words midi and repeats spring to the eye. In fact, I thought
you introduced them yourself (I haven't checked) at
http://lists.gnu.org/archive/html/lilypond-user/2016-08/msg00216.html

I don't understand why you seem reluctant to give credit to what LP
does correctly, and in accordance with the documentation.

Cheers,
David.