Re: What to do wanting a 4th order Bézier?

[Urs Liska]

Am 09.10.2016 um 12:07 schrieb Thomas Morley:

2016-09-18 15:38 GMT+02:00 Simon Albrecht <address@hidden>:

On 18.09.2016 15:15, Kieren MacMillan wrote:

And finally with a better user interface

Under what circumstances would you NOT want the end of one curve to match
precisely the beginning of the next one?
If “none”, then I would say an even better user interface would not
require typing that set of coordinates twice.

I thought about that too – it also makes the data structure less complicated
if it’s a list of seven pairs. I added the new version to the LSR
<http://lsr.di.unimi.it/LSR/Item?id=1045>, awaiting approval.

Best, Simon

Hi Simon,

I'm not sure what to do with this snippet.

I had the impression Urs started to implement it in the source. Is
that true? Urs?
Otoh, the LSR runs 2.18.0. With this version the lsr-snippet is the
only available possibility to do slurs this way, warranting it's
approval, I'd say.

Opinions?

I've started an implementation in https://github.com/openlilylib/snippets, but with the intention of proposing it for LilyPond proper. My approach doesn't require the redundant definition of control points (both doubling the start/end and ensuring the two opposing control points are in opposite directions), but actually I could even imagine integrating it into the \shape function, in a way that allows you to optionally add inflection points when shaping a slur.

If it makes a difference (and I can imagine it does) my code will only run with latest LilyPond and would unreasonably complicated to backport (due to the really useful angle/vector ly:XXX functions that are available now).

Urs

Cheers,
  Harm