Allow 'staff-padding to work with MeasureCounter

Allow 'staff-padding to work with MeasureCounter

 -Set 'Y-offset in the grob description to the appropriate
callback.  (This is required for the default setting of
'staff-padding to have any effect.)

 -Provide a doc snippet to illustrate vertical alignment.

[david.nalesnik, 2017-03-28 23:34:04]

Please review.  Thanks!

[lemzwerg, 2017-03-29 05:12:33]

LGTM

[simon.albrecht, 2017-03-29 11:08:17]

I’m not sure if a doc snippet is really necessary.  Using staff-padding is
explained in
<http://lilypond.org/doc/v2.19/Documentation/learning/moving-objects> and
<http://lilypond.org/doc/v2.19/Documentation/learning/fixing-overlapping-notat...>,
so maybe a verbal remark in the section on Measure counts with a link there
might suffice.  IIRC we want to be careful about extending the NR, don’t we?

[david.nalesnik, 2017-03-29 12:40:32]

On 2017/03/29 11:08:17, simon.albrecht wrote:
> I’m not sure if a doc snippet is really necessary.  Using staff-padding is
> explained in
> <http://lilypond.org/doc/v2.19/Documentation/learning/moving-objects> and
>
<http://lilypond.org/doc/v2.19/Documentation/learning/fixing-overlapping-notat...>,

> IIRC we want to be careful about extending the NR, don’t we?

I have no particular attachment to the snippet, so if it's redundant I'll remove
it.

> so maybe a verbal remark in the section on Measure counts with a link there
> might suffice.

Besides its IR page, MeasureCounter is documented right now through two
snippets:

http://lilypond.org/doc/v2.19/Documentation/snippets/repeats#repeats-numberin...

http://lilypond.org/doc/v2.19/Documentation/snippets/repeats#repeats-centered...

I could amend the second snippet to use staff-padding?  Other than that, I don't
see a convenient way to mention it.

[david.nalesnik, 2017-03-29 12:44:22]

On 2017/03/29 12:40:32, david.nalesnik wrote:
> On 2017/03/29 11:08:17, simon.albrecht wrote:
> > I’m not sure if a doc snippet is really necessary.  Using staff-padding is
> > explained in
> > <http://lilypond.org/doc/v2.19/Documentation/learning/moving-objects> and
> >
>
<http://lilypond.org/doc/v2.19/Documentation/learning/fixing-overlapping-notat...>,
> 
> > IIRC we want to be careful about extending the NR, don’t we?
> 
> I have no particular attachment to the snippet, so if it's redundant I'll
remove
> it.
> 
> > so maybe a verbal remark in the section on Measure counts with a link there
> > might suffice.
> 
> Besides its IR page, MeasureCounter is documented right now through two
> snippets:
> 
>
http://lilypond.org/doc/v2.19/Documentation/snippets/repeats#repeats-numberin...
> 
>
http://lilypond.org/doc/v2.19/Documentation/snippets/repeats#repeats-centered...
> 
> I could amend the second snippet to use staff-padding?  Other than that, I
don't
> see a convenient way to mention it.

Sorry, I mean the *first* snippet.  (The count that begins with two could be
pushed further from the staff possibly.)

[pkx166h, 2017-03-29 22:00:12]

> 
> Sorry, I mean the *first* snippet.  (The count that begins with two could be
> pushed further from the staff possibly.)

Yes that is probably a good idea (if you could add something that would normally
collide - would a slur/tie *over* one of the group of notes work? - to give such
an 'excuse' to use the padding?

Also we should have a changes.tely entry as well - use your original snippet as
the example there to use the work you have already done.

James

[david.nalesnik, 2017-03-29 23:28:40]

On 2017/03/29 22:00:12, pkx166h wrote:
> > 
> > Sorry, I mean the *first* snippet.  (The count that begins with two could be
> > pushed further from the staff possibly.)
> 
> Yes that is probably a good idea (if you could add something that would
normally
> collide - would a slur/tie *over* one of the group of notes work? - to give
such
> an 'excuse' to use the padding?

There won't be any collisions -- the numbers will automatically avoid them.

{
  \startMeasureCount
  c''4( c''''-\accent-\staccato-\marcato g'' d'''')
  \stopMeasureCount
}

The only good excuse to use staff-padding would be in an example such as in the
snippet I provided with the patch: getting all numbers at the same level where
different content in one measure would ordinarily push the number higher than
the number in another measure.  But as Simon points out, the usage of
staff-padding for this purpose is already well documented.  There's no need for
a new snippet, and I see no convenient way to slip staff-padding into one of the
existing snippets without it being an "unjustified" change.

> 
> Also we should have a changes.tely entry as well - use your original snippet
as
> the example there to use the work you have already done.
> 

This is a simple bugfix of something that should have worked since the
MeasureCounter was introduced in 2.17.something.  As such, I don't think a
Changes entry is warranted.

I would propose simply that I fix that bug only, nothing else -- no extra
snippet, no changes to the existing snippets (which are already dense enough
with information).  So I am inclined to upload a patch just with the
scm/define-grobs.scm alterations.

[Carl, 2017-03-30 01:36:48]

On 2017/03/29 23:28:40, david.nalesnik wrote:

> This is a simple bugfix of something that should have worked since the
> MeasureCounter was introduced in 2.17.something.  As such, I don't think a
> Changes entry is warranted.
> 
> I would propose simply that I fix that bug only, nothing else -- no extra
> snippet, no changes to the existing snippets (which are already dense enough
> with information).  So I am inclined to upload a patch just with the
> scm/define-grobs.scm alterations.

I don't think it needs a Changes entry.

But there should be a regression test that shows the proper function of the
'staff-padding property on MeasureCounter.  That way, if it were to break again
in the future, we would know it.

We don't need a documentation snippet, but we do need a regtest.

[david.nalesnik, 2017-03-30 14:49:16]

delete snippet, add regtest