Re: Proposed new available and recommended behavior of \relative

[David Kastrup]

Jim Long <address@hidden> writes:

> On Sat, Mar 09, 2013 at 11:40:14AM +0100, David Kastrup wrote:
>> 
>> Well, the new mnemonic would be "first pitch after \relative is
>> absolute"
>
> I'm not sure whether this is profound or profane, so please
> excuse, but....
>
> For just the case of \relative WITHOUT a reference pitch:
>
> I.  Am I correct that the CURRENT definition of \relatve 
> without a reference pitch is that the reference pitch 
> defaults to the absolute pitch c'?  In other words, 
> \relative { \music } = \relative c' { \music } in the current
> stable version?

Yes.

> II. Is it also true that for any absolute pitch { X } that
> \relative f { X } is the same pitch?

Given a standard Western scale, yes.

> So if somehow I've made two consecutive correct postulates,
> wouldn't a user who used the mnemonic:
>
> "If no reference pitch is given, then the first pitch after
> \relative is relative to f"
>
> be equally correct at predicting the output of the newly proposed
> \relative?

Yes, but he might take longer and make mistakes.

> "\relative f" was mentioned in discussion a few days ago, so
> maybe I'm just slow on the uptake, but it seems that a devil's
> advocate could say that the proposed change just changes the
> default reference pitch from c' to f.

In a standard Western scale.  In any scale, it picks out the middle.  I
don't know whether people use LilyPond with things like Indian scales,
and if they do, whether they would be as audacious as to use \relative
at all, so the difference might be academical.

> Under the proposed change, would not:
>
> \relative { c d e f g a b c }
> \relative { c' d e f g a b c }
>
> be the same as
>
> \relative f { c d e f g a b c }
> \relative f { c' d e f g a b c }

Yes.

> I realize that may not be the way one would want the
> documentation to read, but it works as a mnemonic.

In the same way as "to draw a circle, just draw an ellipse where the
main axes have the same dimension" works as a mnemonic.  Yes, it will
make you arrive at the correct result, but you have more opportunity for
error and confusion.

> As for which is clearest to the user, document both mnemonics,
> and let the user decide which way they prefer to think of it.

I'm getting there in a wording proposal in a different posting.

-- 
David Kastrup