Re: What is the problem with "\relative"? (Was: Do we really offer the future?)

[Gilles]

On Sun, 26 Apr 2015 09:58:33 +0200, Johan Vromans wrote:
> On Sun, 26 Apr 2015 05:52:04 +0000 (UTC)
> Keith OHara <address@hidden> wrote:
>
> > I wish the manual did not use the implicit \relative c'' {}
> > (sometimes \relative c' {} ) enclosing the examples.  As soon as
> > the input gets complicated, \relative becomes difficult to figure 
> > out.
>
> I've always considered \relative as an operation that should be 
> applied as
> close to the actual notes as possible. This gives the least suprises, 
> if
> any.
>
>   \relative c'' {
>     \new PianoStaff <<
>       \new Staff { \time 2/4 c4 e | g g, | }
>       \new Staff { \clef "bass" c,,4 c' | e c | }
>     >>
>   }
>
> This is, indeed, asking for problems...

I totally agree.
This kind of forgiveness from the LilyPond parser allows for
the ultimate bad practice of inextricably mixing typographical
information with musical information.

I fully understand and admit that the learning and notation
manual want to present self-contained examples (so that the
linked LilyPond code can readily compile), but I'm advocating
that readers should be made aware that it is only intended for
illustrating specific aspects of notation and that it should
not be done for any project for which maintainance is a concern.
[Some of the templates are complete counter-examples for this
aspect of best practices. The templates sections should probably
show an image of the intended output but link to "zip" file
containing the skeleton of a real project (possibly with a
"README" file).]

One should strive to separate editorial and musical information
to the utmost that LilyPond permits, not take advantage that
LilyPond allows for obfuscated code!


Best regards,
Gilles