Re: More vertical spacing confusion

[Adam Spiers]

On Wed, Aug 3, 2011 at 5:37 AM, David Kastrup <address@hidden> wrote:
> David Kastrup <address@hidden> writes:
>> Joe Neeman <address@hidden> writes:
>>>> So I have several questions:
>>>>
>>>>  - Is the behaviour I am experiencing (a system spilling over onto a
>>>>    second page) correct or a bug?  If correct, what's the best way to
>>>>    avoid the spillage?
>>>
>>> \paper {
>>>   ragged-last-bottom = ##f
>>> }
>>>
>>> By default, ragged-last-bottom is ##t, so lilypond has no incentive to
>>> avoid that dangling system on page 2.
>>
>> I don't understand this "so".
>>
>> The following should still be penalized since they are deficiencies even
>> on a ragged bottom:
>>
>> a) a page which is considerably less full than the previous page
>> b) every page break, and most certainly inside of a system, should carry
>>    a penalty, giving Lilypond some incentive to compress systems iff it
>>    helps to avoid using more pages than necessary.
>
> Otherwise, the scoring does not care whether we have 5 pages with just a
> single system and lots of bottom space, one full page, and 5 pages with
> just a single system again.
>
> Ragged-bottom or not, this is ugly and that should get reflected in the
> scores.

I agree 100% David!  This is beginning to sound like a bug to me.  I
couldn't find any documentation on how to debug the vertical spacing
algorithms, and by its nature it's impossible to reduce this to a test
case involving only a handful of notes.  Any suggestions how to
proceed?