Re: Issue 1110 in lilypond: Wrong octave of repetition chord with \relative and #{ #} syntax

[David Kastrup]

Peekay Ex <address@hidden> writes:

> On Sun, Oct 16, 2011 at 8:29 AM,  <address@hidden> wrote:
>>
>> Comment #14 on issue 1110 by address@hidden: Wrong octave of repetition
>> chord with \relative and #{ #} syntax
>> http://code.google.com/p/lilypond/issues/detail?id=1110
>>
>> Maybe I'm naïve, but shouldn't q be processed much earlier?  What we really
>> want IMHO is a shorthand at the input level to spare a lot of typing; thus
>>
>>  <a c e'> q
>>
>> should be equal to typing
>>
>>  <a c e'> <a c e'>
>
> But doesn't that only really work *if* the gap between the notes with
> the ' or , on them don't force the 'next' notes in the 'next' chord
> up/down the octave?
>
>> and as soon as a new <...> is seen, this new chord used as the substitution
>> for q.
>
> Isn't that what \repeat unfold is for (and isn't this just the same
> problem with \relative too?).

\repeat unfold does not progress in the octave like a literal repetition
would.

> Why not take \relativism out of < ... > together? and force absolute
> mode only when in < ... > (or rather ignore \relative for anything in
> < ... > ) then we could do away with q and just use \repeat unfold <
> ... >. I don't write chorded music - which probably shows - and am
> wondering if removing relativism from chords is such a big deal for
> typesetters in terms o 'the majority' wouldn't care if they had to
> write chords in absolute mode and use repeat unfold which would
> guarantee true 'replication' of what came previously versus those that
> needed relativism in their sequences of chords and had to add ' or ,
> as appropriate.
>
>>
>> Consequently I wonder whether handling of q can't be done by the lexer (or
>> parser?  I always mix this up :-).  Maybe such an implementation would be
>> really dumb, but I think that too much cleverness with \relative causes more
>> headaches than necessary...
>
> I think it's more of a case of worrying too much about \relative mode
> in certain musical typesetting aspects. I can't see that much is lost
> by banishing it from < ... >.

Werner proposes to rewind Lilypond to the state where q was hardly
usable, now James proposes rewinding to the state where \relative was
hardly usable.

Thanks to git, Lilypond 1.0 is still available.  And with a bit of
merging magic, one might be able to merge a number of later features in
their respective state of infancy.

I'd prefer discussing how to make Lilypond rather more than less useful.
I don't actually like the q feature at all, but there is no point in
making it worse than necessary.

-- 
David Kastrup