Re: Following voices in chords?

[David Kastrup]

Carl Sorensen <address@hidden> writes:

> On 5/7/10 8:29 AM, "David Kastrup" <address@hidden> wrote:

>> Well, looks like a fair piece of work.  And if one invests all this
>> work... I guess it would be nicer if one could write <c\glissando
>> e\glissando g\glissando> <d e f> and notes got matched one by one.
>> And possibly let <c e g>\glissando be the same as that spelled-out
>> first chord.
>> 
>> Putting aside the obvious "patches will be thoughtfully considered"
>> to a later point of time, anybody with a hunch why this would be a
>> bad idea and/or terribly complicated to implement and/or leading to a
>> lot of unpredictable behavior?
>
> I don't have any ideas why it would be a bad idea.  I'd be happy to
> have the behavior you describe.
>
> The reason it doesn't work now is that \glissando inside a chord construct
> creates an articulation, while \glissando outside a chord construct creates
> a separate event.  For me, it was much easier to create a music function
> than to dive in and do the repairs necessary to get to the state you
> describe.  So I did it.

In this case, a music function basically is casting a snippet's essence
into something one can use.

> I appreciate your consistent questioning as to how we might be able to
> get LilyPond to behave in a way that seems consistent with our
> expectations.  I wish I had the time to understand the internals of
> parsing better so I could contribute more in this area.  But I don't,
> so I do the best I can with the time I have.

And it's appreciated.  In an ideal world, the snippet corpus would be
without things one would label "hack".

I don't know when I'll have time to look at that one, but it seems
tempting.

-- 
David Kastrup