From:
Paolo Pr <address@hidden>
Date: Friday, December 13, 2019 at 8:38 PM
To: Carl Sorensen <address@hidden>
Cc: Urs Liska <address@hidden>, "address@hidden" <address@hidden>
Subject: Re: A _javascript_ test code for modifying ties and slurs with mouse
Hello Carl,
my script is not intended for replacing the automatic handling of slurs. I agree that the automatic handling should be the preferred way and the development of the script is somewhat "off-topic" here (but I did
not know the right ml where to post it and I needed some other help for coding in Scheme).
I totally understand that your script is just intended to simplify the manual placing of slurs. And it’s a great tool to have in our engraving toolbox. I’m glad you’re writing it, and providing a graphical way to handle things. Thanks
for doing it. But I’m concerned that we’ll rely too much on the manual tool instead of improving the automatic tool, which is the main goal of LilyPond.
However, I think that the manual placement is necessary in two cases, even with a "perfect" Lilypond:
- for very complex scores (for example: contemporary classical music)
It’s possible that contemporary music may be too complex for automatic slur creation to handle.
- when the automatic placement is correct (no collisions), but not good to see.
I don’t agree that this would be necessary if we had a “perfect” lilypond. But I don’t know that we can ever get a “perfect” lilypond. I hope we can get a *better* lilypond. Currently lilypond creates several different collision-free
placements and selects the best one based on a number of different demerits (measures of ugly slurs). I believe that we can develop better automatic slur creation. The first step to this is to have some correct, but not good to see, examples, along with
better examples. Then we can try to improve the automatic parameters.
I think then that the two features can coexist. Consider that the few lines of codes I wrote show another feature of Lilypond: it can generate svg for easy-scripting too...
Unquestionably the two features can coexist. I just hope we can collect enough examples to improve the automatic workings.
Thanks,
Carl