Re: fixed vs relative

[David Wright]

On Thu 09 May 2019 at 21:40:30 (+0200), Gianmaria Lari wrote:
> On Thu, 9 May 2019 at 21:32, David Wright <address@hidden> wrote:
> 
> > On Thu 09 May 2019 at 20:20:36 (+0200), David Kastrup wrote:
> > > Gianmaria Lari <address@hidden> writes:
> > >
> > > > I saw the discussion about fixed vs relative few times in the lilypond
> > ml.
> > > >
> > > > I think it's a lot about personal taste and habit and personally I
> > decided
> > > > to stick to fixed mainly because I find fixing mistakes in relative
> > mode is
> > > > much more annoying than writing in fixed way....
> > > >
> > > > But what if the editor would help a bit? What if you enter music in
> > > > relative mode (don't worrying to specify the octave) but your editor
> > would
> > > > propose the "correct" (nearest) octave to add?
> > > >
> > > > For example suppose your cursor is immediatly after
> > > >
> > > > b'
> > > >
> > > >
> > > > and then you type
> > > >
> > > > c
> > > >
> > > >
> > > > What if the editor proposes to autocomplete with '' ?
> > > >
> > > > This is just an idea.
> > >
> > > You know that Frescobaldi can convert absolute to relative and vice
> > versa?
> >
> > As can the   ly   standalone (derived from F~ possibly).
> >
> 
> Sorry I don't understand.

ly, packaged as python3-ly in Debian, has a number of commands hived
off (I assume) from Fresco including: re-indent, reformat, translate
the language, transpose, abs2rel, rel2abs, simplify-accidentals, etc.

> > But if you can coerce an editor into doing this trick, I would suggest
> > that absolute is a better target than fixed. Perhaps you meant that anyway.
> 
> What's the difference? I was thinking they were synonimus in lilypond
> talking.

AIUI, in \absolute, c' generates middle C and only middle C; in
\relative, c' generates a C which is at least a fifth above the
previous note; and in \fixed, c' generates a C within the octave
above the reference octave.

> > However, there may be a downside. With relative, a wrong decision on
> > one note has a dramatic effect on the following music, but is easily
> > corrected with one tick. When this trick has been applied, many
> > notes may have to be individually tickled after the mistake is
> > discovered.
> 
> Maybe you're right. Or maybe the fact to see clearly what the editor
> propose you (for each note) make you very self conscious of what you write.

The "wrong decision" that I was talking about was where you think a
note needs a tick (that the editor proposed) when it doesn't¹. The
mistake is only revealed when LP has run and the notehead (and those
following) is an octave off the correct position.

> Generally I appreciate the autocomplete feature even if I found it making
> my though less fluent (but maybe it's better :))

¹ or vice versa.

Cheers,
David.