column balancing question

column balancing question

[Michal Kvasnicka]

Good morning.

One reason for me to switch to ConTeXt were its macros for typesetting 
in columns that are highly superior to any such macro I have ever seen. 
(I typeset almost all my documnts in two columns.) But I am not able to 
supress one strange feature: some pages are one more line longer than 
they should be. It happen randomly -- or at least I was not able to find 
out when it happens. Therefore I'm not able to prepare an minimal file. 
I thougth that setting
\widowpenalty=0
\clubpenalty=0
\brokenpenalty=0
\normalbottom
may help, but it does not.

Can you tell me how to fix this very problem? Well, it does so on all 
versions of ConTeXt I have ever tried and both with teTeX and TeXLive 
distribution.

Many thanks for your attention and help.
Yours
Michal Kvasnicka

P.S. my document setting is like this (some example can be found at 
http://www.econ.muni.cz/~qasar/tmp/delba_prace.pdf):

% output=pdf
\setupbodyfont[ss,10pt]
\setuplayout
[topspace=2cm, height=fit, lines=55,
backspace=15mm, width=fit,
margin=0pt,
header=3\baselineskip, footer=0pt,
grid=yes]
\setuppagenumbering[location=]
\setupheader[style={\em}]
\setupheadertexts[Axel Leijonhufvud: Jednotlivec, trh a průmyslová
dělba práce][pagenumber]
\setupindenting[medium]
\indenting[first]
\setupfootnotes
[location=columns, rule=off,
before={\blank[3*line,small]}, after=]
\setupfootnotedefinition
[location=left,width=fit,
sample={\mathematics{\star\star\star}\ },hang=1]
\setupcolumns[balance=no]

\widowpenalty=0
\clubpenalty=0
\brokenpenalty=0

\setuphead[section]
[number=no, style={\bf},
before={\blank[3*line]}, after={\blank[line]},
indentnext=yes]

Re: column balancing question

[Hans Hagen]

Michal Kvasnicka wrote:

> Good morning.
>
> One reason for me to switch to ConTeXt were its macros for typesetting 
> in columns that are highly superior to any such macro I have ever 
> seen. (I typeset almost all my documnts in two columns.) But I am not 
> able to supress one strange feature: some pages are one more line 
> longer than they should be. It happen randomly -- or at least I was 
> not able to find out when it happens. Therefore I'm not able to 
> prepare an minimal file. I thougth that setting

Hard to see without example. You may try this (i have this in my local expermental setup) 

\def\getmulticolumnlines
  {\mcscratchdimen-\columntextoffset
   \multiply\mcscratchdimen \plustwo
   \advance\mcscratchdimen \columntextheight
   \ifdim\precolumnboxheight>\zeropoint
      \advance\mcscratchdimen -\precolumnboxheight
   \fi
   \settotalinsertionheight
   \advance\mcscratchdimen -\totalinsertionheight
   \getnoflines\mcscratchdimen
   % added 30/7/2004
   \ifnum\layoutlines>\zerocount \ifnum\noflines>\layoutlines
     \noflines\layoutlines
   \fi \fi
   \nofcolumnlines\noflines}

Another option is to switch to column sets (see columns.pdf), 

Hans 



-----------------------------------------------------------------
                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
-----------------------------------------------------------------

Re: column balancing question

[Michal Kvasnicka]

Good evening.

Many thanks for your help. Your solution works well with the newest 
ConTeXt and one particular document I now typeset. I have to check it 
still with more documents. Do you want to gen a report from these tests?

If you want I can send you many examples, but not `minimal files', since 
I don't know when the problem occures. As I typeset our church bulletin, 
I've got that result pretty often---approximately on one third or one 
fourth of all pages.

I can't switch to columnset all my documents. Columnsets are absolutely 
great (one of the ConTeXt miracle---I thought it is not possible to do 
it in TeX because of its breaking routines), but up-to-now columnsets 
cannot be ballanced (or can they?), and my church bulletin is typset in 
such a way that each article can start anywhere on the page, and the 
columns must be ballanced before it. I can't explain it better because 
of my English, but its pages look like this:

    xxxxx   xxxxx
    xxxxx   xxxxx
    xxxxx   xxxxx
    xxxxx   xxxxx

    Title of article
    xxxxx   xxxxx
    xxxxx   xxxxx

Therefore I would lobby for the support of the old \startcolumns ... 
\stopcolumns environment. :-)

As for the Willi Egger's question, I don't use `fit' options very often. 
The problem I've mentioned in my last mail occures even when the `width' 
is set to some dimen and the `length' of page is set to some number of 
lines.

Many greetings and thanks again.
Your sincerely
Michal Kvasnicka

> Hard to see without example. You may try this (i have this in my local 
> expermental setup)
> \def\getmulticolumnlines
>  {\mcscratchdimen-\columntextoffset
>   \multiply\mcscratchdimen \plustwo
>   \advance\mcscratchdimen \columntextheight
>   \ifdim\precolumnboxheight>\zeropoint
>      \advance\mcscratchdimen -\precolumnboxheight
>   \fi
>   \settotalinsertionheight
>   \advance\mcscratchdimen -\totalinsertionheight
>   \getnoflines\mcscratchdimen
>   % added 30/7/2004
>   \ifnum\layoutlines>\zerocount \ifnum\noflines>\layoutlines
>     \noflines\layoutlines
>   \fi \fi
>   \nofcolumnlines\noflines}
>
> Another option is to switch to column sets (see columns.pdf),
> Hans
>
>
> -----------------------------------------------------------------
>                                          Hans Hagen | PRAGMA ADE
>              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
>     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
>                                             | www.pragma-pod.nl
> -----------------------------------------------------------------
>
> _______________________________________________
> ntg-context mailing list
> ntg-context@ntg.nl
> http://www.ntg.nl/mailman/listinfo/ntg-context

Re: column balancing question

[Hans Hagen]

Michal Kvasnicka wrote:

> Good evening.
>
> Many thanks for your help. Your solution works well with the newest 
> ConTeXt and one particular document I now typeset. I have to check it 
> still with more documents. Do you want to gen a report from these tests?
>
> If you want I can send you many examples, but not `minimal files', 
> since I don't know when the problem occures. As I typeset our church 
> bulletin, I've got that result pretty often---approximately on one 
> third or one fourth of all pages.

just test this on many docs and let me know if it fails

(one source of problems with i'm currently trying to catch is rounding 
errors  in comparisons)

> I can't switch to columnset all my documents. Columnsets are 
> absolutely great (one of the ConTeXt miracle---I thought it is not 
> possible to do it in TeX because of its breaking routines), but 
> up-to-now columnsets cannot be ballanced (or can they?), and my church 
> bulletin is typset in such a way that each article can start anywhere 
> on the page, and the columns must be ballanced before it. I can't 
> explain it better because of my English, but its pages look like this:
>
>    xxxxx   xxxxx
>    xxxxx   xxxxx
>    xxxxx   xxxxx
>    xxxxx   xxxxx
>
>    Title of article
>    xxxxx   xxxxx
>    xxxxx   xxxxx
>
> Therefore I would lobby for the support of the old \startcolumns ... 
> \stopcolumns environment. :-)

don't worry, i'm not driopping the old one, actually, in due time i'll provide more specialized ones and interfaces to write your own -) 

balancing in columnsets is a tricky business, the following should work 

(kind of manual balacing, complete control over how long columns on pages are; probably a bit fragile) 

% output=pdftex interface=en

\setuplayout[grid=yes] \showgrid \showstruts % gaat nogal fout zonder grid

\setuptolerance[verytolerant]

\definecolumnset [test-1] [n=2]
\definecolumnset [test-2] [n=3]

\starttext

\startcolumnset [test-1] \dorecurse {1}{\input tufte \par} \stopcolumnset
\startcolumnset [test-2] \dorecurse {2}{\input ward  \par} \stopcolumnset

\page

\setupcolumnset [test-1] [balance=yes]
\setupcolumnset [test-2] [balance=yes]

\setupcolumnsetlines[test-1][1][1] [8]
\setupcolumnsetlines[test-1][1][2][10]

\startcolumnset [test-1] \dorecurse {1}{\input tufte \par} \stopcolumnset
\startcolumnset [test-2] \dorecurse {2}{\input ward  \par} \stopcolumnset

\page

\setupcolumnsetlines[test-1][1][1] [8]
\setupcolumnsetlines[test-1][1][2][10]

\setupcolumnsetlines[test-2][1][1][19]
\setupcolumnsetlines[test-2][1][2][18]
\setupcolumnsetlines[test-2][1][3][19]

\startcolumnset [test-1] \dorecurse {1}{\input tufte \par} \stopcolumnset
\startcolumnset [test-2] \dorecurse {2}{\input ward  \par} \stopcolumnset

\page

\setupcolumnsetlines[test-1][1][1] [8]
\setupcolumnsetlines[test-1][1][2][10]

\setupcolumnsetlines[test-2][1][1][0]
\setupcolumnsetlines[test-2][1][2][-2]
\setupcolumnsetlines[test-2][1][3][0]

\startcolumnset [test-1] \dorecurse {1}{\input tufte \par} \stopcolumnset
\startcolumnset [test-2] \dorecurse {6}{\input ward  \par} \stopcolumnset

\page

\setupcolumnsetlines[test-1][1][1] [8]
\setupcolumnsetlines[test-1][1][2][10]

\setupcolumnsetstart[test-2][1][1][20]
\setupcolumnsetstart[test-2][1][2][20]
\setupcolumnsetstart[test-2][1][3][20]

\startcolumnset [test-1] \dorecurse {1}{\input tufte \par} \stopcolumnset
\startcolumnset [test-2] \dorecurse {2}{\input ward  \par} \stopcolumnset

\page

\setupcolumnsetlines[test-1][1][1] [8]
\setupcolumnsetlines[test-1][1][2][10]

\setupcolumnsetlines[test-2][1][1][-20]
\setupcolumnsetlines[test-2][1][2][-20]
\setupcolumnsetlines[test-2][1][3][-20]

\startcolumnset [test-1] \dorecurse {1}{\input tufte \par} \stopcolumnset
\startcolumnset [test-2] \dorecurse {2}{\input ward  \par} \stopcolumnset

\stoptext

Hans 


-----------------------------------------------------------------
                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
-----------------------------------------------------------------

Re: column balancing question

[Willi Egger]

Good evening,

Though Hans has sent already a piece of code ...
I saw that you are using height=fit, and width=fit. Did you ever try to 
compile your file without these options? -- Some time ago I experienced 
troubles with such options, although my layout was different and the 
document was not setup with columns.

Kind regards

Willi

Michal Kvasnicka wrote:
> Good morning.
> 
> One reason for me to switch to ConTeXt were its macros for typesetting 
> in columns that are highly superior to any such macro I have ever seen. 
> (I typeset almost all my documnts in two columns.) But I am not able to 
> supress one strange feature: some pages are one more line longer than 
> they should be. It happen randomly -- or at least I was not able to find 
> out when it happens. Therefore I'm not able to prepare an minimal file. 
> I thougth that setting
> \widowpenalty=0
> \clubpenalty=0
> \brokenpenalty=0
> \normalbottom
> may help, but it does not.
> 
> Can you tell me how to fix this very problem? Well, it does so on all 
> versions of ConTeXt I have ever tried and both with teTeX and TeXLive 
> distribution.
> 
> Many thanks for your attention and help.
> Yours
> Michal Kvasnicka
> 
> P.S. my document setting is like this (some example can be found at 
> http://www.econ.muni.cz/~qasar/tmp/delba_prace.pdf):
> 
> % output=pdf
> \setupbodyfont[ss,10pt]
> \setuplayout
> [topspace=2cm, height=fit, lines=55,
> backspace=15mm, width=fit,
> margin=0pt,
> header=3\baselineskip, footer=0pt,
> grid=yes]
> \setuppagenumbering[location=]
> \setupheader[style={\em}]
> \setupheadertexts[Axel Leijonhufvud: Jednotlivec, trh a průmyslová
> dělba práce][pagenumber]
> \setupindenting[medium]
> \indenting[first]
> \setupfootnotes
> [location=columns, rule=off,
> before={\blank[3*line,small]}, after=]
> \setupfootnotedefinition
> [location=left,width=fit,
> sample={\mathematics{\star\star\star}\ },hang=1]
> \setupcolumns[balance=no]
> 
> \widowpenalty=0
> \clubpenalty=0
> \brokenpenalty=0
> 
> \setuphead[section]
> [number=no, style={\bf},
> before={\blank[3*line]}, after={\blank[line]},
> indentnext=yes]
> 
> _______________________________________________
> ntg-context mailing list
> ntg-context@ntg.nl
> http://www.ntg.nl/mailman/listinfo/ntg-context

map problems

[Boris Tschirschwitz]

Hi.

A year ago I wrote a set of ConTeXt-set documents on gentoo linux with
teTeX without any problems.
Now, trying to update them on Mac OS X with Gerben Wieda's teTeX
distribution, I get these messages

specials       : fdf,tpd loaded
)
pdftex         : using map file: original-context-symbol.map
pdftex         : using map file: pl0-ams-cmr.map
pdftex         : using map file: original-vogel-symbol.map
)
*

Seaching Google, I found some references to this problem, but all in
connection with installing new fonts.
Since Gentoo managed to set up everything automatically, I wonder if there
is a simple fix for that.

Thanks,
Boris.

Re: map problems

[Boris Tschirschwitz]

I apologise for that email.

Of course there is no problem with ConTeXt, as it also didn't indicate
any. It turned out to be a problem with HPFS+'s sloppyness with upper and
lower case letters.

Cheers,
Boris.

On Wed, 8 Sep 2004, Boris Tschirschwitz wrote:

> Hi.
>
> A year ago I wrote a set of ConTeXt-set documents on gentoo linux with
> teTeX without any problems.
> Now, trying to update them on Mac OS X with Gerben Wieda's teTeX
> distribution, I get these messages
>
> specials       : fdf,tpd loaded
> )
> pdftex         : using map file: original-context-symbol.map
> pdftex         : using map file: pl0-ams-cmr.map
> pdftex         : using map file: original-vogel-symbol.map
> )
> *
>
> Seaching Google, I found some references to this problem, but all in
> connection with installing new fonts.
> Since Gentoo managed to set up everything automatically, I wonder if there
> is a simple fix for that.
>
> Thanks,
> Boris.
>
>

alpha version

[Martin Kolařík]

Hi Hans,

I have:

----
This is pdfeTeXk, Version 3.141592-1.11a-2.1 (Web2c 7.5.2) 
 %&-line parsing enabled.
 (d:/TeX/texmf/web2c/cp8bit.tcx)
-----

in Win32.

I made current alpha version efmt, it was OK, but after trying to typeset some text (I did not tested anyone else) pdfetex failed with general protection failure (a memory cannot be read).

The document is UTF8 XML, the failure occured (see >>HERE<<):

-----
specials       : fdf loaded
)
specials       : fdf,tpd loaded
encoding       : coding default
encoding       : coding il2

>>HERE<<

)
color          : system rgb is global activated
-----

I did not try yet to experiment with removing some code from the last included (my) file (inside context of this file the afilure occurred), maybe I will find more information.

Have a nice day,

Martin

alpha version

[Hans Hagen]

Hi,

For those interested in sorting registers ...

... the alpha release has sort-* files in it.

For the moment you need to load:

  \readfile{sort-ini}{}{}

After that

mainlanguage[sl]

\starttext
    test \index {aa1}  test \index {ab1}         test \index {aa2}
    test \index {ab2}  test \index {aa10}        test \index {aa8}
    test \index {aa9}  test \index {aa11}        test \index {aa10}
    test \index {cccc} test \index {\ccaron ccc} test \index {\cacute ccc}
    test \index {caaa} test \index {\ccaron aaa} test \index {\cacute aaa}
    \placeindex
\stoptext

will adapt itself to the language. Sort orders are defined in sort-lan.tex. 

You need to run 'newtexexec' (ruby reimplementation of texexec), which uses the texutil module/class instead of the program (faster). 

The reimplementation of texutil permits user plugins and such (more about that later).

I need more info on sort order about for instance czech and german and ...  

Hans 

-----------------------------------------------------------------
                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
-----------------------------------------------------------------

Re: alpha version

[Vit Zyka]

Hans Hagen wrote:
> Hi,
> 
> For those interested in sorting registers ...

OK, I am very interesting. Now I set aside some time to help with Cz part.

> ... the alpha release has sort-* files in it.

I looked at it. Seems that basic sorting rules can be set by macro, more 
difficult by Rubu module, right? Let start with macro.

I derrived test file from your example. The only difference is that I 
replace \ccaron by č in IL2 and put \enableregime[IL2] to very 
beginning. Seems such char is not recognized by sorting rules. What is 
wrong?

I see there are 4 macros. I guess their meaning:

\exportsortrule {zacron}{z+1} ... replaces two lines:
   \exportsortexpansion {ccaron}  {cz} ... char to string replacement
   \exportsortdivision  {cz}  {ccaron} ... reverse

\exportsortreduction{ch}{c} ... second pass replacement ???
   if I want 'ch' be sorted between 'h' and 'i' so I do
   \exportsortreduction{ch}{h+1} ??

vit

> For the moment you need to load:
> 
>  \readfile{sort-ini}{}{}
> 
> After that
> 
> mainlanguage[sl]
> 
> \starttext
>    test \index {aa1}  test \index {ab1}         test \index {aa2}
>    test \index {ab2}  test \index {aa10}        test \index {aa8}
>    test \index {aa9}  test \index {aa11}        test \index {aa10}
>    test \index {cccc} test \index {\ccaron ccc} test \index {\cacute ccc}
>    test \index {caaa} test \index {\ccaron aaa} test \index {\cacute aaa}
>    \placeindex
> \stoptext
> 
> will adapt itself to the language. Sort orders are defined in sort-lan.tex.
> You need to run 'newtexexec' (ruby reimplementation of texexec), which 
> uses the texutil module/class instead of the program (faster).
> The reimplementation of texutil permits user plugins and such (more 
> about that later).
> 
> I need more info on sort order about for instance czech and german and ... 
> Hans
> -----------------------------------------------------------------
>                                          Hans Hagen | PRAGMA ADE
>              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
>     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
>                                             | www.pragma-pod.nl
> -----------------------------------------------------------------
> 
> _______________________________________________
> ntg-context mailing list
> ntg-context@ntg.nl
> http://www.ntg.nl/mailman/listinfo/ntg-context
> 

-- 
=======================================================
Ing. Vít Zýka, Ph.D.                         TYPOkvítek

database publishing              databazove publikovani
data maintaining and typesetting in typographic quality
priprava dat a jejich sazba v typograficke kvalite

tel.: (+420) 777 198 189     www: http://typokvitek.com
=======================================================

Re: alpha version

[Vit Zyka]

[-- Attachment #1: Type: text/plain, Size: 3217 bytes --]

Hi, I just add the test code for sorting. It demonstrates different 
behaviour for
a) \ccaron
b) \v c
c) č
I expected both b) and c) cases expand to \ccaron under il2 regime.
Or there is some miss in my configuration?

vit

Vit Zyka wrote:
> Hans Hagen wrote:
> 
>> Hi,
>>
>> For those interested in sorting registers ...
> 
> 
> OK, I am very interesting. Now I set aside some time to help with Cz part.
> 
>> ... the alpha release has sort-* files in it.
> 
> 
> I looked at it. Seems that basic sorting rules can be set by macro, more 
> difficult by Rubu module, right? Let start with macro.
> 
> I derrived test file from your example. The only difference is that I 
> replace \ccaron by č in IL2 and put \enableregime[IL2] to very 
> beginning. Seems such char is not recognized by sorting rules. What is 
> wrong?
> 
> I see there are 4 macros. I guess their meaning:
> 
> \exportsortrule {zacron}{z+1} ... replaces two lines:
>   \exportsortexpansion {ccaron}  {cz} ... char to string replacement
>   \exportsortdivision  {cz}  {ccaron} ... reverse
> 
> \exportsortreduction{ch}{c} ... second pass replacement ???
>   if I want 'ch' be sorted between 'h' and 'i' so I do
>   \exportsortreduction{ch}{h+1} ??
> 
> vit
> 
>> For the moment you need to load:
>>
>>  \readfile{sort-ini}{}{}
>>
>> After that
>>
>> mainlanguage[sl]
>>
>> \starttext
>>    test \index {aa1}  test \index {ab1}         test \index {aa2}
>>    test \index {ab2}  test \index {aa10}        test \index {aa8}
>>    test \index {aa9}  test \index {aa11}        test \index {aa10}
>>    test \index {cccc} test \index {\ccaron ccc} test \index {\cacute ccc}
>>    test \index {caaa} test \index {\ccaron aaa} test \index {\cacute aaa}
>>    \placeindex
>> \stoptext
>>
>> will adapt itself to the language. Sort orders are defined in 
>> sort-lan.tex.
>> You need to run 'newtexexec' (ruby reimplementation of texexec), which 
>> uses the texutil module/class instead of the program (faster).
>> The reimplementation of texutil permits user plugins and such (more 
>> about that later).
>>
>> I need more info on sort order about for instance czech and german and 
>> ... Hans
>> -----------------------------------------------------------------
>>                                          Hans Hagen | PRAGMA ADE
>>              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
>>     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
>>                                             | www.pragma-pod.nl
>> -----------------------------------------------------------------
>>
>> _______________________________________________
>> ntg-context mailing list
>> ntg-context@ntg.nl
>> http://www.ntg.nl/mailman/listinfo/ntg-context
>>
> 

-- 
=======================================================
Ing. Vít Zýka, Ph.D.                         TYPOkvítek

database publishing              databazove publikovani
data maintaining and typesetting in typographic quality
priprava dat a jejich sazba v typograficke kvalite

tel.: (+420) 777 198 189     www: http://typokvitek.com
=======================================================

[-- Attachment #2: sorting-cz.zip --]
[-- Type: application/zip, Size: 724 bytes --]

[-- Attachment #3: Type: text/plain, Size: 139 bytes --]

_______________________________________________
ntg-context mailing list
ntg-context@ntg.nl
http://www.ntg.nl/mailman/listinfo/ntg-context