Re: Invalid input I expected to be valid

[David Kastrup]

"Phil Holmes" <address@hidden> writes:

> ----- Original Message ----- 
> From: "David Kastrup" <address@hidden>
> To: <address@hidden>
> Sent: Sunday, September 09, 2012 5:10 PM
> Subject: Re: Invalid input I expected to be valid
>
>
>> Again, just _what_ did you expect the \p to be attached to when
>> expecting this to be valid input?  I don't see a really convincing
>> solution.
>
>> From a user perspective, I would have expected it to sit on the last
>> note of 
> the triplet.  Again, as a user, I don't really see anything different
> between
>
> \times 2/3 { g'4 g'4 g'4 } \p
>
> and
>
> \times 2/3 { g'4 g'4 g'4 \p }

\p attaches to the note coming before it, and there is no note coming
before it but rather a large expression.  Note that in the second case
\p happens simultaneously with the _onset_ of g'4, effectively moving
back a quarter note in time.  I am not convinced that the first version
has an obvious target just how far back in time \p should be moving.

> although from following -devel, I realise the parser (?) would likely
> take them as completely different.

Not really "completely different" but rather "can't make a decision in
conformance with the rules normally applied for \p".

> It was actually auto-generated from my noteworthy converter, so I've
> done a minor code change to write dynamics before triplet conclusions.

In my book, the only thing making sense.

-- 
David Kastrup