Re: Lilypond lobbying?

[Joseph Wakeling]

On 08/25/2011 02:30 AM, PMA wrote:
> *If* my LilyPond output PDF were to match what Schott wants to see
> (in other words, a correct Schott-targeted style-sheet would not have
> changed it), then would Schott print my original PDF *as-is*?

It's important to understand what the _real_ requirements are from the
point of view of the publisher.

Generally speaking the publisher (whether of music or other texts) does
not expect to receive from the author a version of the document that
corresponds exactly to how it will appear in print.  Of course, it's
very nice if the author can follow as closely as possible the
publisher's style guide, but even where this has been followed to the
letter (and the text itself is impeccable and entirely error-free) the
publisher will typically plan on further editing the author's text.

The publisher's main requirement from a software solution is therefore
that it makes _editing and tweaking_ a text very easy, because even in
the best-case scenario they expect to have to do a lot of manual
intervention.

Lilypond is excellent (and highly extensible) when it comes to
implementing general stylistic rules, but very finnicky when it comes to
the small manual tweaks that are the common currency of editorial
intervention.  Take as an example the tweaks described in the Lilypond
Notation Manual on slurs:
http://lilypond.org/doc/v2.14/Documentation/notation/expressive-marks-as-curves#slurs

... and compare that to the ease of editing in Finale: click, drag,
click, and see _straight away_ that you've got it right.

Now consider that relative difficulty scaled up across the number of
times you might have to implement an individual custom tweak in a
50-page orchestral score, and you begin to see the issue from the
publisher or engraver's point of view.  The fact that Finale may get
more things wrong initially is not an issue when correcting them is
simple; the fact that Lilypond may get so many things right initially is
not an issue when it's so much more tricky to make (and validate) small
manual corrections.