Widows and orphans in the Bibliography

[meho_r]

Hi

I recently noticed that I get widows and orphans in the bibliography
and it seems that \widowpenalty and \clubpenalty do not have any
effect on this. Document class seems to be irrelevant. I use memoir in
the example in attachment (take a look at pages 9 and 11 for an
example of this behaviour), but if you try, e.g., book document class,
you'll notice the same (e.g. take a look at pages 3 or 5).

I'll provide a complete example, but since I don't know how to attach
it directly here, I've uploaded the .zip file (named test.zip)
containing .tex and .bib files on MediaFire: http://is.gd/7xpV2-

Note: I use pdflatex for compiling the document, TeXLive2009 without
any updates yet. I also use apacite package for producing output
according to APA style requirements.

[zappathustra]

\widowpenalty is an indication, not an order, to TeX not to break
there.
Avoiding widows requires more than a penalty, unless you set it to
10,000 or more.
Giving enough stretch between bibliography items will help preventing
underfull pages, but then with unpredictable space.
There might be a macro to set the distance between items, but if you
don't find it, setting "\parskip=0pt plus <some stretch>"
in the bibliography may work, where <some stretch> is more than 1pt
(default stretch).

Paul

[meho_r]

Actually, it seems to be working, setting \bibitemsep to 0pt plus 1pt
(an example) works fine. Interesting that adding such a small amount
sorted things out. I'll test it to see if it works for full
bibliography.

Thanks :-)

[Lars Madsen]

meho_r wrote:
> On Feb 2, 3:17 pm, zappathustra <zappathus...@free.Fr> wrote:

just giving it a little play often solves a lot of bad breaks.

same thing can be done with the spacing before chapters in the TOC.

--

/daleif (remove RTFSIGNATURE from email address)

LaTeX FAQ: http://www.tex.ac.uk/faq
LaTeX book: http://www.imf.au.dk/system/latex/bog/ (in Danish)
Remember to post minimal examples, see URL below
http://www.tex.ac.uk/cgi-bin/texfaq2html?label=minxampl
http://www.minimalbeispiel.de/mini-en.html

[zappathustra]

The 1pt stretch is enough because there are probably at least 13 items
per page,
hence sufficient inter-item space to stretch and fill the space left
by a line
(12pt, I suppose). I'd not taken this into account, hence my useless
indication
to increase the stretchable part.

Paul

[Donald Arseneau]

On Feb 3, 2:41 am, zappathustra <zappathus...@free.Fr> wrote:
> The 1pt stretch is enough because there are probably at least 13 items
> per page,
> hence sufficient inter-item space to stretch and fill the space left
> by a line
> (12pt, I suppose). I'd not taken this into account, hence my useless
> indication
> to increase the stretchable part.

Another perspective is that the penalty did not work to discourage a
page break because there was no vertical flexibility on the page. By
adding a modest vertical flexibility, you have allowed the penalties
to work properly.

When there is no flexibility, penalties have only two effective
values:
finite (like zero) and infinite (>9999).

[meho_r]

But shouldn't items in bibliography behave like paragraphs in normal
text where \widowpenalty=10000 prevents widows from appearing (where
no stretchable space between paragraphs is allowed, and with
\raggedbottom activated)? Or you're saying that one absolutely cannot
force TeX into preventing widows if there is no vertical flexibility?

[zappathustra]

\widowpenalty=10000 indeed prevents widows, and TeX will break before,
even without \raggedbottom (although \raggedbottom avoids underfull
pages).

TeX will always prefer an underfull page (i.e. a page whose stretch
isn't enough) to an overfull page (a page that is too long).
\widowpenalty=10000 means "you can't break here"; TeX can't break
after the widow, since the page would be overfull (unless you have
shrinkable vertical space, of course), so it breaks before the
penultimate line; if there's vertical stretch, it is used and indeed
overused to fill the space (i.e. if there's 6pt of stretch and TeX
needs 12pt, it'll do as if there was 12pt and issue an "underfull box"
message - generally suffixed with "while output is active"); if
there's no vertical stretch, then the space won't be filled, in effect
creating a ragged bottom, and you'll get the same message. As Donald
put it, without stretch a penalty is meaningful only if infinite
(unless you use it to communicate with the output routine, but that's
another matter).

Paul

[Donald Arseneau]

No, I said what I said, no more no less. Finite penalties (like 4000,
hint hint) have no effect unless there is some flexibility.

I will say now that your document does not have \widowpenalty=10000
during the bibliography: check the definition of \thebibliography.

Donald Arseneau

[meho_r]

> No, I said what I said, no more no less.  Finite penalties (like 4000,
> hint hint) have no effect unless there is some flexibility.
>
> I will say now that your document does not have \widowpenalty=10000
> during the bibliography: check the definition of \thebibliography.
>
> Donald Arseneau

I tested the file again, it seems that the bibliography (printed with
\begin{btSect}...\end{btSect} of bibtopic package) does follow global
definition of \widowpenalty which I set to 10000 (I tried to disable
the penalty and set it to lower values just to check if the
bibliography reacts). Since \widowpenalty is set to infinite, it
should prevent widows no matter is there stretchable vertical space or
not, right? It is possible, however, that bibtopic interfere in the
process, but I can't tell.

Well, the important part is that Paul's suggestion worked. Thank you
both for your comments and explanations :-)

Best regards,

M.