Re: 2.13.32 not honouring margins and staves going off page

[Carl Sorensen]

On 9/5/10 12:33 AM, "James Bailey" <address@hidden> wrote:

> 
> 
> On Sep 5, 2010, at 5:31 AM, Carl Sorensen wrote:
> 
>> On 9/4/10 5:07 PM, "Nick Payne" <address@hidden> wrote:
>> 
>>> I haven't been able to reproduce this behaviour with a small example,
>>> but I'm getting some staves in a score where lilypond tries to fit too
>>> many bars and the stave goes past the margin and off the right-hand edge
>>> of the page, and others where the the right-hand barline is at the
>>> correct margin but notes go past the barline. I don't know if this png
>>> showing three staves is too large to make it to the mailing list, but it
>>> shows the problem. I put a png of that entire page at
>>> http://www.users.on.net/~njpayne/music/bwv-1004_4-page1.png, along with
>>> the two ly source files
>>> (http://www.users.on.net/~njpayne/music/bwv-1004_4.ly &
>>> http://www.users.on.net/~njpayne/music/guitar.ly). Two of the five pages
>>> in the score have this problem.
>> 
>> This is probably the same bug as 1229.  Notes on leger lines with stems
>> pointing away from the staff are apparently not considered as part of the
>> staff when considering x-extent.
>> 
>> Thanks,
>> 
>> Carl
> 
> I don't think so. Looking at the output and the warnings generated:
> warning: couldn't fit music on page: overflow is 1.795440
> warning: couldn't fit music on page: overflow is 0.221693
> 
> This looks like far worse, plus, there are no ledger lines involved. And,
> since I apparently had nothing better to do on a Sunday morning, I discovered
> that the enlarged staff is the culprit:

I still think it's the same.  In both cases, there's more music than is
available to fit within the scheduled margins.  If the notes lie within the
staff, the right-hand end of the the staff is stretched beyond the scheduled
margin (see measure 4 of the original).  But if the note is on a leger line
and the stem is away from the staff, the note is set beyond the bar line
(see measure 6 of the original).  Both of these behaviors were observed when
I played with 1229.  In the case of 1229, we got to see them by shrinking
the line length.  In this case, we got to see them by increasing the staff
size.

But I haven't actually traced the bug to a specific piece of code, so I
could be wrong on this.

Thanks,

Carl