On 05/28/2015 04:59 PM, Otared Kavian wrote:

Hi Pablo,

I tested your sample with

ConTeXt ver: 2014.05.21 22:04 MKIV beta fmt: 2015.3.31 int: english/english

(which is from TeXLIve 2014) and also with

ConTeXt ver: 2015.05.21 13:39 MKIV beta fmt: 2015.5.21 int: english/english

and I get exactly the same result, with the last two line of page 1 pushed to the second page.

However with mmkii, thta is

ConTeXt ver: 2013.06.07 17:34 MKII fmt: 2015.5.21 int: english/english

the penalties act correctly and page one has all the lines expected, while page 2 begins with a new section.

Actually I hav enoticed that recent versions of mkiv create strange page breaks in some of documents containing maths, but up to now I have not succeeded to reproduce this bad behavior in a short (and minimal) example.

Many thanks for your testing, Otared. It seems that different MkIV betas may give different results. I noticed it by compiling the same source with different MkIV betas. Many thanks for your help, Pablo -- http://www.ousia.tk

On 05/28/2015 07:10 PM, Hans Hagen wrote:

On 5/27/2015 10:38 PM, Pablo Rodriguez wrote:

...

[...] Compiling it with beta from 2015.03.25 22:13, page break is fine.

But with latest beta from 2015.05.24 12:42 (or even with beta from 2015.05.15 11:42 includes a wrong page break. The setups for widow and orphan lines move the two last lines to the next page.

Could anyone confirm this bug?

i wonder if it's a bug or just tex's decision ... you can use

\showmakeup[penalty]

and see what penalties get injected

are the linebreaks identical, i.e. same hyphenation?

Many thanks for your reply, Hans. Hyphenation is the same in both documents.

btw, brokenpenalty 1 gives a different result

Page break is wrong when any of the three values (brokenpenalty, widowpenalty or clubpenalty) is 0 or above 9999. I replacing \subsection with \section gives the right page break: \showframe\showgrid \setuplayout[setups=*lessstrict] \startsetups[*lessstrict] \setup[reset] \widowpenalty=10000 \clubpenalty=10000 \brokenpenalty=0 \stopsetups \setupbodyfont[pagella, 12.6pt] \starttext \chapter{Quotes} \section{Zapf} \dorecurse{4}{\input zapf\par} \section{Knuth} \input knuth \stoptext Why does \subsection cause a different page break than \section? Many thanks for your help, Pablo -- http://www.ousia.tk

Hi Pablo, Indeed the document I was talking about has been typeset with various versions of mkiv, but now I observe that sometimes the pagebreaks are made in a very strange way. Since I can’t make a minimal example, I put on the following link an excerpt of the PDF document where the bad pagebreaks can be seen: https://dl.dropboxusercontent.com/u/4316076/bad-pagebreak.pdf (Please see pages 19 and 32). It seems that this happens, sometimes, but as far as I can say from my experiments the bad pagebreak happens always before a « proclaim » which is defined as follows: \defineenumeration[proclaim] [text=, before={\blank\startbackground[frame=off,leftframe=on, rulethickness=2pt,framecolor=darkred, background=color, backgroundcolor=screen]}, after={\stopbackground\blank}, style=slanted, title=yes, titleleft=, titleright={.}, alternative=hanging, width=fit, right={~}, way=bysection, prefix=yes, prefixsegments={chapter:section}] Actually I found a workaround for this by changing the above definition into: \defineenumeration[proclaim] [text=, before={\blank\startmarginrule[rulethickness=1pt,rulecolor=darkred]}, after={\stopmarginrule}, style=slanted, title=yes, titleleft=, titleright={.}, alternative=hanging, width=fit, right={~}, way=bysection, prefix=yes, prefixsegments={chapter:section}] However it would be better to understand why the first definition does not work properly, in a rather random way. Best regards: OK

On 01 Jun 2015, at 00:16, Hans Hagen wrote:

On 5/31/2015 1:56 PM, Pablo Rodriguez wrote:

...

On 05/28/2015 07:10 PM, Hans Hagen wrote:

...

On 5/27/2015 10:38 PM, Pablo Rodriguez wrote:

...

[...] Compiling it with beta from 2015.03.25 22:13, page break is fine.

But with latest beta from 2015.05.24 12:42 (or even with beta from 2015.05.15 11:42 includes a wrong page break. The setups for widow and orphan lines move the two last lines to the next page.

Could anyone confirm this bug?

i wonder if it's a bug or just tex's decision ... you can use

\showmakeup[penalty]

and see what penalties get injected

are the linebreaks identical, i.e. same hyphenation?

Many thanks for your reply, Hans.

Hyphenation is the same in both documents.

...

btw, brokenpenalty 1 gives a different result

Page break is wrong when any of the three values (brokenpenalty, widowpenalty or clubpenalty) is 0 or above 9999.

I replacing \subsection with \section gives the right page break:

\showframe\showgrid \setuplayout[setups=*lessstrict] \startsetups[*lessstrict] \setup[reset] \widowpenalty=10000 \clubpenalty=10000 \brokenpenalty=0 \stopsetups \setupbodyfont[pagella, 12.6pt] \starttext \chapter{Quotes} \section{Zapf} \dorecurse{4}{\input zapf\par} \section{Knuth} \input knuth \stoptext

Why does \subsection cause a different page break than \section?

because there are penalties related to (sub)sections in a way that keeps them together but at the same time still permits breaks in case of too many successive ones .. and penalties determine the way the page is broken

Hans

----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl ----------------------------------------------------------------- ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context webpage : http://www.pragma-ade.nl / http://tex.aanhet.net archive : http://foundry.supelec.fr/projects/contextrev/ wiki : http://contextgarden.net ___________________________________________________________________________________

On 01 Jun 2015, at 09:28, Pablo Rodriguez wrote:

[…] Hi Otared,

having only three theorems my guess will be probably wrong, but I’ll try it.

The theorem that doesn’t cause a wrong pagebreak is preceeded by a single line paragraph.

In order to confirm whether this is the pattern, I would remove some words from the last paragraph on page 19, so that there has only one line. And I would add some words to the pagraph before theorem 2.3.2 on page 31, so this pargraph contains two lines.

Could you check this?

Pablo -- http://www.ousia.tk

Hi Pablo, Actually I have done what you suggest by adding or removing text, but at the end when the document changes all of a sudden the bad pagebreaks appear or disappear. The second definition of « proclaim » I sent you suppresses the bad pagebreaks, but in some other situations still I have other bad pagebreaks… Until I have not made a rela minimal example the problem cannot be understood and permanently solved. Best regards: OK