Re: Chords with unequal durations

[Trevor Daniels]

Neil Puttock wrote Monday, September 13, 2010 4:48 PM

> That's an illegal merge, since it obscures the duration of the 
> minim.

>  In Nick's example, setting
> 'ignore-collision allows the noteheads to be stuck on top of each
> other with no regard for the individual durations: hence the 
> quaver
> obscures the minim.  This means the merge commands are irrelevant,
> since the merging code never gets called.

Just to avoid confusion, let's state this clearly.  Nick's original
example showed a minim (half-note) being merged with a quaver
(eighth-note).  That is allowed, as both durations are clear (in 
fact
there's an example of this in LM 4.5.3 Real music example.)  No
need to use 'ignore-collision, as this produces no error messages.

Merging a crotchet (quarter-note) with a quaver is also fine, no 
error
messages.

But merging a minim (half-note) with a crotchet (quarter-note) is 
not
allowed, and LilyPond will not do it, as it would not be possible to
tell the duration of one or other of the notes, whether the note 
head
were open or filled in.  It would be incorrect typesetting to force 
this
with 'ignore-collision.

As to the documentation bug, the documentation for 'ignore-collision
in the Internal Reference says
  "If set, don’t do note collision resolution on this NoteColumn"
That seems clear enough, but the snippet
 Suppressing warnings for clashing note columns in the LSR
is misleading and should be changed to explain clearly the
consequences of using 'ignore-collision to suppress messages,
especially as it appears in the snippet list for Simultaneous notes
in the Notation Reference.

Copying to -bug for the Bug Squad.

Trevor

Trevor