Weird godown, dorecurse, colors and page-fitting interaction

Weird godown, dorecurse, colors and page-fitting interaction

[Adam Lindsay]

Hi all,

What's happening here?

\setupcolors[state=start]
\starttext
\startTEXpage
\dorecurse{2}{abc\godown[0pt]}
\stopTEXpage
\startTEXpage
\startcolor[yellow]
\dorecurse{2}{abc\godown[0pt]}
\stopcolor
\stopTEXpage
\stoptext

-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
  Adam T. Lindsay, Computing Dept.     atl@comp.lancs.ac.uk
  Lancaster University, InfoLab21        +44(0)1524/510.514
  Lancaster, LA1 4WA, UK             Fax:+44(0)1524/510.492
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: Weird godown, dorecurse, colors and page-fitting interaction

[Hans Hagen]

Adam Lindsay wrote:

> Hi all,
>
> What's happening here?
>
> \setupcolors[state=start]
> \starttext
> \startTEXpage
> \dorecurse{2}{abc\godown[0pt]}
> \stopTEXpage
> \startTEXpage
> \startcolor[yellow]
> \dorecurse{2}{abc\godown[0pt]}
> \stopcolor
> \stopTEXpage
> \stoptext
>
a long story ... to which ttp wizzard taco might want to add comment

ok, a fix first (missing \fi)

\def\godown[#1]%
  {\relax
   \ifhmode\endgraf\fi
   \ifvmode\nointerlineskip\vskip#1\relax\fi}

what you observe is some interference; the fitting page macros use 
framed and when you set width and height to fit, it will use the 
reshaperr to determine the width; this mechanism fails as soon as nodes 
show up that cannot be \un-whatevered or \last-whatevered; also, some 
mechanisms have as side effect that they force the box to have a width 
of \hsize

question for taco: how difficult would it be to have something:

   \squeezebox\scratchbox

where the result is a box with a width equal to the max(natural width of 
lines)

(that way we could replace the reshaper and get around those interfering 
and blocking nodes)

you may see some of the side effects when you play with:

\starttext

\setupcolors[state=start]

\startTEXpage
\dorecurse{2}{abc\godown[0pt]}
\stopTEXpage

\startTEXpage[foregroundcolor=yellow]
\dorecurse{2}{abc\godown[0pt]}
\stopTEXpage

\startTEXpage
\startcolor[red]\dorecurse{2}{abc\godown[0pt]}\stopcolor
\stopTEXpage

\startTEXpage[width=fixed]
\startcolor[green]\dorecurse{2}{abc\godown[0pt]}\stopcolor
\stopTEXpage

\startTEXpage[width=fixed]
\dorecurse{2}{\startcolor[cyan]abc\stopcolor\godown[0pt]}
\stopTEXpage

\startTEXpage
\dorecurse{2}{\startcolor[magenta]abc\stopcolor\godown[0pt]}
\stopTEXpage

\startTEXpage
\vbox{\startcolor[blue]\dorecurse{2}{abc\godown[0pt]}\stopcolor}
\stopTEXpage

\stoptext

so, there is no real solution unless you use the foreground color or box 
things yourself

Hans

Re: Weird godown, dorecurse, colors and page-fitting interaction

[Taco Hoekwater]

Hans Hagen wrote:
> question for taco: how difficult would it be to have something:
> 
>   \squeezebox\scratchbox
> 

Will think about it. should be doable enough, after all you can
do this by hand as well (using \lastbox etc.).

Taco

Re: Weird godown, dorecurse, colors and page-fitting interaction

[Adam Lindsay]

Hans Hagen wrote:
> a long story ... to which ttp wizzard taco might want to add comment
> 
> ok, a fix first (missing \fi)
> 
> \def\godown[#1]%
>  {\relax
>   \ifhmode\endgraf\fi
>   \ifvmode\nointerlineskip\vskip#1\relax\fi}

okay. Those missing \fi's were troubling.

> what you observe is some interference; the fitting page macros use 
> framed and when you set width and height to fit, it will use the 
> reshaperr to determine the width; this mechanism fails as soon as nodes 
> show up that cannot be \un-whatevered or \last-whatevered; also, some 
> mechanisms have as side effect that they force the box to have a width 
> of \hsize

Okay, I thought it might be something like that.

> you may see some of the side effects when you play with:
  ...
> so, there is no real solution unless you use the foreground color or box 
> things yourself

Hmm. Forgroundcolor is not a solution for me, as I'm trying to mix 
colors (e.g., black + a spot).

However, if I go ahead and \hbox each line and \vbox the set, it looks 
like I have something workable!

Thanks,
adam
-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
  Adam T. Lindsay, Computing Dept.     atl@comp.lancs.ac.uk
  Lancaster University, InfoLab21        +44(0)1524/510.514
  Lancaster, LA1 4WA, UK             Fax:+44(0)1524/510.492
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-