Re: Substitute for s1*0

[David Kastrup]

Keith OHara <address@hidden> writes:

> I had complained that this method caused problems when instruments
> take cues from each other, but these problems are avoided so 
> long as the temporary voice is explicitly created and finished.
>
> The uses to finish spanners after the last note, 
>   { e'2\p\< d'\> s1*0\! }
> can also be handled with a temporary voice
>   <<
>     { e'2 d'! }
>     { s2\p\< s4\> s8. s16\! }
>   >>
> Other uses are simlarly avoided with parallel voices.

The latter is not a parallel voice: it is all one voice.  If we take the
output of
\displayLilyMusic
on the incriminated version and the suggested alternative (and after
fixing a totally embarrassing number of errors, some of which being
regressions from the EventChord changes, oh fantastic), we arrive at

{ e'2\p\< d'\> < >\! }

<< { e'2 d'! } { s\p\< s4\> s8. s16\! } >>

Apart from some apparent typos, the second version makes it quite harder
to correlate notes and dynamics.  And this is a simple example.  In
practice, a crescendo will last longer.  Sure, you can open and close
the << >> construct several times in order to simplify the counting.

But selling that as a sacrifice in the name of simplification really
appears absurd to me.

> David Kastrup <address@hidden> writes:
>
>> I tried discussing using "< >" in the documentation vs the possibly
>> stranger looking "<>". 
>> However, there has been no feedback whatsoever on this proposal.
>
> It has not yet been a day.

Long enough for shooting the messenger in a barrel.

-- 
David Kastrup