Re: Page layout is affected by previous pages?

[Mats Bengtsson]

Han-Wen Nienhuys wrote:

> Mats Bengtsson wrote:
>
> > The question from address@hidden provided a nice example
> > of a LilyPond problem, namely that the page breaking and system
> > spacing on one page seem to be affected by what happened on earlier
> > pages.
>
>
> page breaking, like line-breaking, is done by looking for a globally 
> optimal solution. Apparently, we need to add some more scoring. Why do 
> you say
>
> > In my opinion, the systems on
> > a new page should be typeset without any consideration of what has
> > happened on earlier pages.
>
> ?

If you really go for a globally optimal solution, then I was clearly wrong,
but I had the impression that LilyPond at least for the moment used
some greedy algorithm that determined the page breaking page by page.
Here it seemed that some ackumulated lack of space in the previous pages
caused the troubles on the 3rd page, which doesn't make sense.

Also, after an explicit \pageBreak, I cannot see any reason why the
layout of the following pages should be in any way influenced by
the layout of the pages before the \pageBreak (unless you want to
implement some artificial intelligence that guesses that the user
inserted a manual page break since he wanted wider spacing in
the full score. Of course, such a guess is mostly irrelevant).

  /Mats

--
=============================================
        Mats Bengtsson
        Signal Processing
        Signals, Sensors and Systems
        Royal Institute of Technology
        SE-100 44  STOCKHOLM
        Sweden
        Phone: (+46) 8 790 8463                         
       Fax:   (+46) 8 790 7260
        Email: address@hidden
        WWW: http://www.s3.kth.se/~mabe
=============================================