Re: \relative proposal: putting absolute pitches anywhere within \relative block using @-sign

[David Kastrup]

Kieren MacMillan <address@hidden> writes:

> I feel that the original proposal, as specified elsewhere, violates
> the principle of least surprise for existing users starting new
> projects (not using convert-ly) with the new LilyPond.
>
> I'm not sure I understand why that's true… As I understand the proposal,
>
>     \relative { c''' d e f }
>
> would come out as a rising step-wise scale, starting on c''' — which
> is not in the least surprising to me.

The premise was "existing users", with a focus on those knowing
\relative to be relative to c' by default.  At least that's how I
understand his objection.  There has been agreement quite a few years
that silently defaulting to c' is confusing enough that we don't want to
see it in LilyPond's code base.

Non-silently specifying f as reference is somewhat startling.  It is a
nice candidate for a silent default, in correspondence with your "not in
the least surprising" remark above.  Nice enough?  Not everybody is of
that opinion.

> Please explain how it's surprising, or how I am perhaps
> misunderstanding the original proposal.
>
> Thanks,
> Kieren.
>
> p.s. In the end, I'll almost certainly still stick with the
> emminently-obvious/logical [to me]
>     \relative c''' { c d e f }

It gets less update-friendly the more material creeps between c''' and c
since its redundancy is spread to two places.  When used consistently
(in the "octave of first note" sense rather than "octave of c", which
are indistinguishable in your example), it's a bit more cumbersome.

For me the litmus test was how I'd feel about converting the entire
LilyPond code base consistently to \relative x''' { x notation as
compared to converting it consistently to \relative { x''' notation.
The complexity of the involved convert-ly rule is quite the same.

For me, the second conversion would feel more satisfactory/justifiable.
Whether or not one actually does it, this feeling for me means that
having this notation available would be a reasonable addition to
LilyPond's music input.

-- 
David Kastrup