On Mar 13, 2011, at 3:17 PM, Nicolas Sceaux wrote:
> Le 13 mars 2011 à 19:53, Nicolas Sceaux a écrit :
>
>> Hi Mike,
>>
>> First of all, thank you very much for your precious work on footnotes.
>>
>> The following example demonstrates a strange behavior wrt footnotes:
>>
>> %% For each empty \fill-line in header and footer markups,
>> %% footnotes together with separation line are added.
>
> Actually the header markup has no impact, and the problem is not due to
> \fill-line, but to \column:
>
> \paper {
> %% no problem with \markup { \column { } }
> oddFooterMarkup = \markup \column { }
> }
>
> { \footnoteGrob #'NoteHead #'(1 . 2) "note" "footnote" c'' }
>
> Nicolas
>
AHH!! Attack of the footnotes!!
I get the sense that this bug may have to do with the way that footers and/or
markups work in general, not just the footnotes. More precisely, the footnotes
get attached to the footer via Page_layout_problem::add_footnotes_to_footer on
line 98 of page-layout-problem.cc . This is only ever called twice in the code
- once for the dummy page used for configs (follow line 589 of page-breaking.cc
to its logical conclusion) and once on line 553 of page-breaking.cc for the
real deal. When I use pretty prints in the code to compare the buggy output w/
the good output, the pretty prints in the footnote stream never change.
So, my guess is that the blank footer is printing multiple times, and thus, the
non-blank footnote is tagging along with it.
What needs to be scrubbed, then, is this behavior.
As the behavior is correct for \markup { \column { } } and for \markup \column
{ foo }, I think that it may be a parser issue &/or a markup interpretation
issue &/or a footer creation issue.
At any rate, if there is someone who is conversant in markups and/or footers,
I'd appreciate their taking this on.
The quick workaround, as already suggested by Nicolas, is simply to put \markup
{ \column { } } anytime someone wants to use a blank column as the footer
(which, admittedly, is esoteric, but as I am all about esoteric, I support
someone's desire to do this).
Cheers,
MS
_______________________________________________
bug-lilypond mailing list
bug-lilypond@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-lilypond
