Allows outside staff positioning for vertical axis groups with no staff

Allows outside staff positioning for vertical axis groups with no staff

[Keith, 2013-03-30 06:33:10]

Why not use the flat-line for empty VerticalAxisGroups specifically in
side-position-interface.cc when placing something against the staff (in the
clause if (include_staff) {} ) ?

If you make each empty VerticalAxisGroup into a line, wouldn't that prevent
interleaving beteen left- and right-hand piano staves, if there is a Dynamics in
between that happens to be empty ?

[Keith, 2013-03-30 08:29:39]

Simpler to just reverse the "special X-alignment" commit.

The old code, that followed the comment about a 'kldge' whatever that means, was
rather elegant.  If we are placing something alongside an empty set, place it
against the reference point for that set.  Having a special case for
StanzaNumbers and another special case for empty VerticalAxisGroups gets messy.

[mike7, 2013-03-30 09:13:51]

On 30 mars 2013, at 08:33, k-ohara5a5a@oco.net wrote:

> Why not use the flat-line for empty VerticalAxisGroups specifically in
> side-position-interface.cc when placing something against the staff (in
> the clause if (include_staff) {} ) ?

include_staff is false for VerticalAxisGroups without staves because there is no
staff grob.

> 
> If you make each empty VerticalAxisGroup into a line, wouldn't that
> prevent interleaving beteen left- and right-hand piano staves, if there
> is a Dynamics in between that happens to be empty ?

This wouldn't happen (at least I don't think it would after a quick test) -
vertical axis group spacing is controlled in page-layout-problem.cc and doesn't
use the same mechanisms.

Cheers,
MS

[mike7, 2013-03-30 09:17:02]

On 30 mars 2013, at 10:29, k-ohara5a5a@oco.net wrote:

> Simpler to just reverse the "special X-alignment" commit.
> 
> The old code, that followed the comment about a 'kldge' whatever that
> means, was rather elegant.  If we are placing something alongside an
> empty set, place it against the reference point for that set.  Having a
> special case for StanzaNumbers and another special case for empty
> VerticalAxisGroups gets messy.
> 
> https://codereview.appspot.com/8188044/

Fair enough, revert and I'll add a comment.

Cheers,
MS