page-breaking: allow ragged pages to be compressed; issue 3281

Old error messages:
1)
warning: cannot fit music on page: ragged-spacing was requested, but page was compressed

2)
warning: cannot fit music on page: overflow is 13.096881
warning: compressing music to fit
  
New error messages:
1)
warning: ragged-bottom was specified, but page must be compressed
warning: page 34 has been compressed

2)
warning: compressing a over-full page by 13.1 staff-spaces
warning: page 34 was compressed

[lemzwerg, 2013-11-14 05:49:27]

LGTM, thanks, with one comment.

https://codereview.appspot.com/25710043/diff/40001/Documentation/notation/spa...
File Documentation/notation/spacing.itely (right):

https://codereview.appspot.com/25710043/diff/40001/Documentation/notation/spa...
Documentation/notation/spacing.itely:411: 
Reading this it's not clear to me whether you have to use this flag in addition
to `ragged-bottom' and `ragged-last-bottom', or whether this overrides the other
two flags.

[Keith, 2013-11-14 21:48:26]

https://codereview.appspot.com/25710043/diff/40001/Documentation/notation/spa...
File Documentation/notation/spacing.itely (right):

https://codereview.appspot.com/25710043/diff/40001/Documentation/notation/spa...
Documentation/notation/spacing.itely:411: 
On 2013/11/14 05:49:28, lemzwerg wrote:
> Reading this it's not clear to me whether you have to use this flag in
addition
> to `ragged-bottom' and `ragged-last-bottom', or whether this overrides the
other
> two flags.

This flag controls how many systems fit on a ragged-bottom page, with the old
flags determining which pages have ragged bottoms.  It does not override the old
flags.

If I see an example of how the old rule against compressing ragged-bottom pages
can be useful, I'll try to work it into a regression test.  If I do not, I will
remove this flag, simplifying the code.

https://codereview.appspot.com/25710043/diff/40001/lily/page-breaking.cc
File lily/page-breaking.cc (right):

https://codereview.appspot.com/25710043/diff/40001/lily/page-breaking.cc#newc...
lily/page-breaking.cc:1503: Real ragged_compress_force = rigid () ? infinity_f :
0.0;
Oops, we need to fall-back to the natural force on the page, not 0.0.

[lemzwerg, 2013-11-19 07:57:36]

LGTM.  Just to be sure: Does the warning message regarding forced compressed
pages (if ragged-bottom is active) get emitted even for single-system pages?

[Trevor Daniels, 2013-11-19 08:59:17]

LGTM apart from a suggested doc change.

https://codereview.appspot.com/25710043/diff/180001/Documentation/notation/sp...
File Documentation/notation/spacing.itely (right):

https://codereview.appspot.com/25710043/diff/180001/Documentation/notation/sp...
Documentation/notation/spacing.itely:401: the last page in each section created
with a @code{\bookpart} block
"the last page and the last page in each section ..."
(just to help those users who haven't yet grasped
the delights of \bookpart and don't realise one is
implied.)

[Keith, 2013-11-19 09:16:34]

On 2013/11/19 07:57:36, lemzwerg wrote:
> LGTM.  Just to be sure: Does the warning message regarding forced compressed
> pages (if ragged-bottom is active) get emitted even for single-system pages?

A single system on a ragged-bottom page gets a warning from lilypond, but not
lilypond-book.

If I move the code that sends the error-message to make_pages(), so it can print
a page number, then there is there is one possible way to compress without a
warning:  if you use lilypond-book and ragged-bottom=#t, then the make_pages()
is not used.  I do not know what ragged-bottom means in lilypond-book, but it
does switches on your warning message.

I'll go back to the old error-messaging behavior, without the page number.

[dak, 2013-11-19 09:31:24]

On 2013/11/19 09:16:34, Keith wrote:
> On 2013/11/19 07:57:36, lemzwerg wrote:
> > LGTM.  Just to be sure: Does the warning message regarding forced compressed
> > pages (if ragged-bottom is active) get emitted even for single-system pages?
> 
> A single system on a ragged-bottom page gets a warning from lilypond, but not
> lilypond-book.

> If I move the code that sends the error-message to make_pages(), so it can
print
> a page number, then there is there is one possible way to compress without a
> warning:  if you use lilypond-book and ragged-bottom=#t, then the make_pages()
> is not used.  I do not know what ragged-bottom means in lilypond-book, but it
> does switches on your warning message.

I don't think that ragged-bottom means anything in lilypond-book in LaTeX:
lilypond-book then produces one image per system, and it's LaTeX's job to
assemble a page from that.  Texinfo might take whole pages from LilyPond
instead.  Not sure about that.

> I'll go back to the old error-messaging behavior, without the page number.

Can you see any effect from ragged-bottom in those runs of lilypond-book where
the warning disappears?

[Keith, 2013-11-19 09:44:25]

On 2013/11/19 08:59:17, Trevor Daniels wrote:
> LGTM apart from a suggested doc change.

Got it.  Thanks

[Keith, 2013-11-19 09:48:47]

On 2013/11/19 09:31:24, dak wrote:
> 
> Can you see any effect from ragged-bottom in those runs of lilypond-book where
> the warning disappears?

No.
Changing the warning behavior was a side-topic, so I'm putting it back and
finishing the main issue alone.