Re: chord-names

[Rune Zedeler]

Robin Davies wrote:
> I'm ramping up for a major assault on chord-name.scm in an attempt to get to
> work properly across the board.

Great!
Some comments:

Remember to look into the ugly spacing of chord names (spaces sometimes 
added before alterations).

> 2) Provide a substitution mechanism for "m"/"-" and Maj/M/delta/solid-delta
> within the base chord names.  So   c-(M7) in a base chord name can be
> substituted to cm(Maj7) if that's what the user has specified.

Nice. Remember to take special care of the 7 (remove it from the delta 
sign):

Maj7  M7  delta
Maj9  M9  delta9

> dim/circle-slash and aug/+ options seem to be too scattered to organize, but
> can be supported with explicit base name entries.

:-(
I would really love if that would just work so that you set flags of how 
to typeset whole and half-diminished chords.
I understand that if you add lots of alterations the the algorithm might 
run into trouble, but it would be wonderful if it would just handle the 
simple cases.
I just try and summarize the problems regarding diminished chords:

You state "dim/Ø" - but I don't think that ANYBODY conciders dim and Ø 
the same chord: Ø is by everybody I know concidered a half-diminished 
seventh, and dim is not. (correct me if I'm wrong).
I know of theese notations:

              o         dim    (b5)
a-c-es:  Ao  A   Adim  A     Am


            ø         7     (b5)
a-c-es-g:  A   Ao7  Ao   Am7


              o   dim    (b5b7)
a-c-es-ges:  A   A     Am


The entire question about when to use superscript hasn't been discussed: 
I have seen the 7 in C7 both in full size, raised full size and raised 
small size.
I like a scheme where base 3-note chords are notated in full size on the 
base line, >4note chords are notated in full size superscript and 
alterations are notated in reduced size superscript. - And hence I 
prefer the first of each of the above examples.
In order not to confuse a-cis-e-g-bes with as-c-es-ges-bes, one writes

 9b                    b9
A   for the first and A   for the second.
Notice that in alterations the alteration still goes before the number.

Also I would like you to support square instead of sus4 - so that (with 
[] = square) one would have

g-c-d: G[]

g-c-d-f:  G[]7

g-c-d-f-a:  G[]9


> 1) German note names in root, and base.

german-chords-init.ly should work. Just include that and typeset a lot 
of chords to see how it works.
Please allow for both A/cis and A/C# notation.


Dammit there are many issues to think about. Bottom line: I would like 
the scheme to be as configurable as possible - hopefully we could end up 
having just one chord algorithm so configurable that we didn't need more 
than the one.
Have to go now. Will return if I remember something more.

-Rune