uncompatibility?

uncompatibility?

[Pawel Jackowski na Onet]

[-- Attachment #1: Type: text/plain, Size: 318 bytes --]

Hi all!

I've installed TeX based on tl8 which is used paralely with based on tl7.
Attached ZIP contains TeX code and two outputs with logs; tl7 and 8, which
are quite different... There are some warnings in newer one (about doubled
map entries) but shouldn't play a role here. Thanks for any hints!


Regards, Pawe/l

[-- Attachment #2: test.zip --]
[-- Type: application/octet-stream, Size: 18031 bytes --]

Re: uncompatibility?

[Patrick Gundlach]

Hi,

> are quite different... There are some warnings in newer one (about doubled
> map entries) but shouldn't play a role here. Thanks for any hints!

right. You might want to clean up the map files, but that is not
important. 

tl7: ConTeXt  ver: 2002.5.24  fmt: 2003.10.6  int: english  mes: english
tl8: ConTeXt  ver: 2003.10.5  fmt: 2003.10.10  int: english  mes: english

Patrick
-- 
next attempt, please ;-)

Re: Re: uncompatibility?

[Hans Hagen]

At 11:09 11/10/2003, you wrote:
>Hi,
>
> > are quite different... There are some warnings in newer one (about doubled
> > map entries) but shouldn't play a role here. Thanks for any hints!
>
>right. You might want to clean up the map files, but that is not
>important.
>
>tl7: ConTeXt  ver: 2002.5.24  fmt: 2003.10.6  int: english  mes: english
>tl8: ConTeXt  ver: 2003.10.5  fmt: 2003.10.10  int: english  mes: english

actually, cleaning up really matters! I have my own alternative map files 
without the "Internal Font Names" since that will mess up pdf inclusion of 
fonts with similar names but different instances (slanted etc)

Hans  

Re: Re: uncompatibility?

[Pawel Jackowski na Onet]

From: "Pawel Jackowski na Onet" <jackos1@poczta.onet.pl>
Sent: Saturday, October 11, 2003 9:29 AM
> At 11:09 11/10/2003, you wrote:
> >Hi,
> >
> > > are quite different... There are some warnings in newer one (about
doubled
> > > map entries) but shouldn't play a role here. Thanks for any hints!
> >


From: "Patrick Gundlach" <pg@levana.de>
Sent: Saturday, October 11, 2003 11:09 AM
> >right. You might want to clean up the map files, but that is not
> >important.
> >
> >tl7: ConTeXt  ver: 2002.5.24  fmt: 2003.10.6  int: english  mes: english
> >tl8: ConTeXt  ver: 2003.10.5  fmt: 2003.10.10  int: english  mes: english

Ok, the difference is about 17 months in ConTeXt version, while formats are
quite fresh... Can't see the clue :-?

From: "Hans Hagen" <pragma@wxs.nl>
Sent: Saturday, October 11, 2003 11:34 AM
 > actually, cleaning up really matters! I have my own alternative map files
> without the "Internal Font Names" since that will mess up pdf inclusion of
> fonts with similar names but different instances (slanted etc)

I admit that maps in my tl8 installation aren't cleaned up already; I have
to spend some time to make it in the way that suits me best. But the problem
I sent doesn't look to be font-dependent. Every characters are found at this
moment, so I put the problem off.

Lets consider the following code:

\setupitemize[4,packed]
\starttext
\startitemize[n]
 \item one
 \item two
 \item three
\stopitemize
\stoptext

The problem is that in newer installation, default itemize style (meaning
packed, with triangle symbol) IS FORCED instead of style given direclty
(just numbering). The older one do what I would expect; numbered style has a
priority. So after running tl8 I get three items with triangles, but after
running tl7 I get three numbered items. Why it is so? Has anything been
changed in itemizing mechanism? I will accept all changes with pleasure, but
the point is to understand what is going on :-D


Thanks, Pawe/l

Re: uncompatibility?

[Patrick Gundlach]

Hi,

>> >tl7: ConTeXt  ver: 2002.5.24  fmt: 2003.10.6  int: english  mes: english
>> >tl8: ConTeXt  ver: 2003.10.5  fmt: 2003.10.10  int: english  mes: english
>
> Ok, the difference is about 17 months in ConTeXt version, while formats are
> quite fresh... Can't see the clue :-?

Sorry for not being in verbose mode. You say you compare tl7 with tl8
but you are using two different (very different) context versions. I
bet that tl7 will genrate the same output as tl8 if you use the 2003.10.5
context. Or that tl8 generates the same output with 2002.5.24 as tl7. 

I was confused with the fact that you mention the tl version. But the
most important one is the context version. Can't give you a better
answer today.

Patrick

Re: Re: uncompatibility?

[Pawel Jackowski na Onet]

From: "Patrick Gundlach" <pg@levana.de>
Sent: Saturday, October 11, 2003 6:48 PM


> Hi,
>
> >> >tl7: ConTeXt  ver: 2002.5.24  fmt: 2003.10.6  int: english  mes:
english
> >> >tl8: ConTeXt  ver: 2003.10.5  fmt: 2003.10.10  int: english  mes:
english
> >
> > Ok, the difference is about 17 months in ConTeXt version, while formats
are
> > quite fresh... Can't see the clue :-?
>
> Sorry for not being in verbose mode. You say you compare tl7 with tl8
> but you are using two different (very different) context versions. I
> bet that tl7 will genrate the same output as tl8 if you use the 2003.10.5
> context. Or that tl8 generates the same output with 2002.5.24 as tl7.

I belive so!

> I was confused with the fact that you mention the tl version. But the
> most important one is the context version. Can't give you a better
> answer today.

I should describe it more precisely. I use installations (2 at the moment)
based on TL distributions. I tried to upgrade ConTeXt many times using
pragma archives, but unfortunately couldn't do that transparently, meaning
in the way that all my current job will be compiled without problems. That
is why I have to do this 17-months step now. Anyway, I'm aware, that ConTeXt
is developing quite fast and I probably shouldn't expect, that 2002.524 and
2003.10.5 versions gives the same output. Could you give me a recipe to spy
changes? I'm sure they are documented, but don't know how to get them.



Thanks, Pawe/l

Re: uncompatibility?

[Patrick Gundlach]

Hi,

> I should describe it more precisely. I use installations (2 at the moment)
> based on TL distributions. 

With the supplied ConTeXt versions?

> Anyway, I'm aware, that
> ConTeXt is developing quite fast and I probably shouldn't expect,
> that 2002.524 and 2003.10.5 versions gives the same output

IMO you should be allowed to expect that your old documents still
work. But that is not the way it is.

> . Could you give me a recipe to spy changes? I'm sure they are
> documented, but don't know how to get them.

I don't know of any other source then the source itself. Taco once
did maintain a list of changes, but this is looong ago.

Patrick
-- 
You are your own rainbow!

Re: Re: uncompatibility?

[Hans Hagen]

At 11:59 12/10/2003, you wrote:
>Hi,
>
> > I should describe it more precisely. I use installations (2 at the moment)
> > based on TL distributions.
>
>With the supplied ConTeXt versions?
>
> > Anyway, I'm aware, that
> > ConTeXt is developing quite fast and I probably shouldn't expect,
> > that 2002.524 and 2003.10.5 versions gives the same output
>
>IMO you should be allowed to expect that your old documents still
>work. But that is not the way it is.

it depends, there are some areas where small changes are made, for instance 
in thinsg related to grid snapping and special float placement; i try to 
guard compatibility so if you run into anoying incompatibilities, let me know

Hans

Re: Re: uncompatibility?

[Hans Hagen]

At 08:26 12/10/2003, you wrote:

>2003.10.5 versions gives the same output. Could you give me a recipe to spy
>changes? I'm sure they are documented, but don't know how to get them.

most changes concern (yet) undocumented features, bug fixes, extensions 
discussed on this list (user requests) ; drastic changes / extensions are 
often announces (or accompanied by a special manual)

Hans  

Re: Re: uncompatibility?

[Hans Hagen]

At 17:36 11/10/2003, you wrote:

>Lets consider the following code:
>
>\setupitemize[4,packed]
>\starttext
>\startitemize[n]
>  \item one
>  \item two
>  \item three
>\stopitemize
>\stoptext
>
>The problem is that in newer installation, default itemize style (meaning
>packed, with triangle symbol) IS FORCED instead of style given direclty
>(just numbering). The older one do what I would expect; numbered style has a

hm, i'll look into it; it's probably related to the fact that last year 
itemize became definable and retaining downward compatibility for the 
\setupitemize command was tricky

Hans   

Re: Re: uncompatibility?

[Hans Hagen]

[-- Attachment #1: Type: text/plain, Size: 3125 bytes --]

This one is supposed to work with the latest version.


At 17:36 11/10/2003, you wrote:

>From: "Pawel Jackowski na Onet" <jackos1@poczta.onet.pl>
>Sent: Saturday, October 11, 2003 9:29 AM
> > At 11:09 11/10/2003, you wrote:
> > >Hi,
> > >
> > > > are quite different... There are some warnings in newer one (about
>doubled
> > > > map entries) but shouldn't play a role here. Thanks for any hints!
> > >
>
>
>From: "Patrick Gundlach" <pg@levana.de>
>Sent: Saturday, October 11, 2003 11:09 AM
> > >right. You might want to clean up the map files, but that is not
> > >important.
> > >
> > >tl7: ConTeXt  ver: 2002.5.24  fmt: 2003.10.6  int: english  mes: english
> > >tl8: ConTeXt  ver: 2003.10.5  fmt: 2003.10.10  int: english  mes: english
>
>Ok, the difference is about 17 months in ConTeXt version, while formats are
>quite fresh... Can't see the clue :-?
>
>From: "Hans Hagen" <pragma@wxs.nl>
>Sent: Saturday, October 11, 2003 11:34 AM
>  > actually, cleaning up really matters! I have my own alternative map files
> > without the "Internal Font Names" since that will mess up pdf inclusion of
> > fonts with similar names but different instances (slanted etc)
>
>I admit that maps in my tl8 installation aren't cleaned up already; I have
>to spend some time to make it in the way that suits me best. But the problem
>I sent doesn't look to be font-dependent. Every characters are found at this
>moment, so I put the problem off.
>
>Lets consider the following code:
>
>\setupitemize[4,packed]
>\starttext
>\startitemize[n]
>  \item one
>  \item two
>  \item three
>\stopitemize
>\stoptext
>
>The problem is that in newer installation, default itemize style (meaning
>packed, with triangle symbol) IS FORCED instead of style given direclty
>(just numbering). The older one do what I would expect; numbered style has a
>priority. So after running tl8 I get three items with triangles, but after
>running tl7 I get three numbered items. Why it is so? Has anything been
>changed in itemizing mechanism? I will accept all changes with pleasure, but
>the point is to understand what is going on :-D
>
>
>Thanks, Pawe/l
>
>
>_______________________________________________
>ntg-context mailing list
>ntg-context@ntg.nl
>http://www.ntg.nl/mailman/listinfo/ntg-context
>
>================================================================
>Deze e-mail is door E-mail VirusScanner van Planet Internet gecontroleerd 
>op virussen.
>Op http://www.planet.nl/evs staat een verwijzing naar de actuele lijst 
>waar op wordt gecontroleerd.

-------------------------------------------------------------------------
                                   Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                       Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
  tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------
                        information: http://www.pragma-ade.com/roadmap.pdf
                     documentation: http://www.pragma-ade.com/showcase.pdf
-------------------------------------------------------------------------

[-- Attachment #2: pvtest.tex --]
[-- Type: text/plain, Size: 5940 bytes --]

\unprotect

\def\redostartitemgroup[#1][#2]%
  {\setfalse\inlinelistitem % new, no indent (leftskip)
   \setfalse\concatnextitem % new, concat
   \ifhmode
     \ifconditional\autoconcatnextitem % new, concat
       \ifdim\lastskip=\itemsignal     % new, concat
         \settrue\concatnextitem       % new, concat
       \fi                             % new, concat
     \fi                               % new, concat
     \iftextitems\else\doifnotinset\v!tekst{#1}\par\fi % suboptimal
   \fi
   \begingroup
   \ifnum\itemlevel=\plusone % NIEUW
     \doadaptleftskip{\getitemparameter1\c!marge}%
   \fi
   \dosetraggedcommand{\getitemparameter\itemlevel\c!uitlijnen}\raggedcommand
   \doifsomething{\getitemparameter\itemlevel\c!inspringen}
     {% is \expanded needed?
      \expanded{\setupindenting[\getitemparameter\itemlevel\c!inspringen]}}%
   \doifinset\v!kolommen{#1}%
     {\ifbinnenkolommen\else\ifnum\itemcolumndepth=\zerocount
        \globallet\itemcolumndepth\itemlevel
        \getitemparameter\itemlevel\c!voor
        \processfirstactioninset
          [#1]
          [  \v!een=>\!!counta1\relax,
            \v!twee=>\!!counta2\relax,
            \v!drie=>\!!counta3\relax,
            \v!vier=>\!!counta4\relax,
            \v!vijf=>\!!counta5\relax,
         \s!unknown=>\@EA\!!counta\getitemparameter\itemlevel\c!n]%
        % new
        \edef\columneditemleftskip{\the\leftskip}%
        \def\postprocesscolumnbox##1%
          {\scratchdimen\columneditemleftskip
           \divide\scratchdimen \nofcolumns
           \hbox{\hskip\columneditemleftskip\hbox{\box##1}}}%
        \scratchdimen-\columneditemleftskip
        \multiply\scratchdimen \nofcolumns
        \advance\scratchdimen \columneditemleftskip
        \advance\scratchdimen \hsize
        \edef\columntextwidth{\the\scratchdimen}%
        \leftskip\zeropoint
        % so far
        \startkolommen
          [\c!n=\!!counta, % netter \??op\itemlevel\c!n
           \c!hoogte=,
           \c!lijn=\v!uit,
           \c!balanceren=\v!ja,
           \c!uitlijnen=\v!nee]%
      \fi\fi}%
   \doifinsetelse\v!intro{#1}\itemintrotrue\itemintrofalse
   \doglobal\increment\noflists
   \let\currentlist\noflists
   \newcounter\noflistelements
   \headitemfalse
   \subitemfalse
   \symbolitemfalse
   \let\marsymbol\relax
   \globallet\somdestination\empty
   \let\symsymbol\empty
   \the\itemgroupcommands
   \setitemlevel{#1}%
  %\getitemparameter\itemlevel\empty
   \let\listitem\empty % ** start value
   \doifelsenothing{#1} % iffirstargument
     {\edef\@@opsymbool{\getitemparameter\itemlevel\c!symbool}%
      \letgvalueempty{\@@globalitemsymbol\itemlevel}%
      \global\letitemparameter\itemlevel\v!verder\empty
    % \setitemmark\@@opsymbool % ** default value
      \dosetupitemgroupvariable[\itemlevel][#2]}
     {\dosetupitemgroupconstant[\itemlevel][#1]%
      \dosetupitemgroupvariable[\itemlevel][#2]%
      \doifinsetelse\v!verder{#1}% \noexpand, else problems in non-etex with chinese
        {\edef\@@opsymbool{\noexpand\getvalue{\@@globalitemsymbol\itemlevel}}%
         \getitemparameter\itemlevel\v!verder}
        {\edef\@@opsymbool{\noexpand\getitemparameter{\itemlevel}{\c!symbool}}%
         \global\setitemparameter\itemlevel\v!verder
           {\dosetupitemgroupconstant[\itemlevel][#1]%
            \dosetupitemgroupvariable[\itemlevel][#2]}}%
      \def\docommando##1% \setitemmark resets \docommando
        {\doifnot{##1}{0}{\setitemmark{##1}}}%
    % \processcommalist[#1,\@@opsymbool]\docommando
      \processcommalist[#1]\docommando}% ** preset sequence or provided sequence
   \ifx\listitem\empty
     \setitemmark\@@opsymbool % ** default value
     \ifx\listitem\empty
       \edef\currentitemsymbol{\itemlevel}% ** fall back
     \fi
   \fi
   \ifautoitemintro\ifnum\prevgraf<3
     \itemintrotrue
   \fi\fi
   \ifparagraphitems
     \ifnum\itemlevel>\plusone
       \letitemparameter\itemlevel\c!tussen\empty
     \fi
   \else\ifpackeditems
     \letitemparameter\itemlevel\c!tussen\empty
   \fi\fi
   \calculatelistwidth\itemlevel{\dimen0}%
   \ifdim\dimen0>\zeropoint\relax
     \ifconditional\inlinelistitem\else
       \advance\leftskip \dimen0\relax
     \fi
   \fi}

\def\setitemmark#1% % en pas op: resets \docommando
  {\doifsymboldefinedelse{#1}
     {\edef\currentitemsymbol{#1}%
      \setxvalue{\@@globalitemsymbol\itemlevel}{\currentitemsymbol}%
      \setgvalue{\@@localitemsymbol \itemlevel}{\unknownitemsymbol}%
      \def\listitem{\symbol[\currentitemsymbol]}%
      \let\@@opsymbool\empty}% \let\docommando\gobbleoneargument}
     {\doifconversiondefinedelse{#1}
        {\edef\currentitemsymbol{#1}%
         \setxvalue{\@@globalitemsymbol\itemlevel}{\currentitemsymbol}%
         \setgvalue{\@@localitemsymbol \itemlevel}%
           {\convertnumber{\currentitemsymbol}{\countervalue{\@@itemcounter\itemlevel}}}%
         \iftextitems
           \doifsomething{\getitemparameter\itemlevel\c!linkertekst}
             {\let\tempsymbol\empty}%
         \fi
         \def\listitem
           {\getitemparameter\itemlevel
              {\iftextitems\c!linkertekst\else\c!links\fi}%
            \getvalue{\@@localitemsymbol\itemlevel}\tempsymbol
            \getitemparameter\itemlevel
              {\iftextitems\c!rechtertekst\else\c!rechts\fi}}%
        \let\@@opsymbool\empty}%\let\docommando\gobbleoneargument}
       {}}}

\protect

\starttext

\startitemize[packed,n]
 \item one   n
 \item two
 \item three
\stopitemize

\setupitemize[packed,4]

\startitemize[n,unpacked]
 \item one n
 \item two
 \item three
\stopitemize

\startitemize[unpacked]
 \item one 4
 \item two
 \item three
\stopitemize

\setupitemize[packed,2]

\startitemize[unpacked]
 \item one 2
 \item two
 \item three
\stopitemize

\setupitemize[packed][symbol=2]

\startitemize[unpacked]
 \item one 2
 \item two
 \item three
\stopitemize

\startitemize[unpacked,n]
 \item one n
 \item two
 \item three
\stopitemize

\stoptext

Re: Re: uncompatibility?

[Pawel Jackowski na Onet]

Problem: a (very) different ConTeXt versions and a different output

From: "Hans Hagen" <pragma@wxs.nl>
To: <ntg-context@ntg.nl>
Sent: Tuesday, October 14, 2003 9:52 PM


> At 17:36 11/10/2003, you wrote:
>
> >Lets consider the following code:
> >
> >\setupitemize[4,packed]
> >\starttext
> >\startitemize[n]
> >  \item one
> >  \item two
> >  \item three
> >\stopitemize
> >\stoptext
> >
> >The problem is that in newer installation, default itemize style (meaning
> >packed, with triangle symbol) IS FORCED instead of style given direclty
> >(just numbering). The older one do what I would expect; numbered style
has a
>
> hm, i'll look into it; it's probably related to the fact that last year
> itemize became definable and retaining downward compatibility for the
> \setupitemize command was tricky

Thanks in advance. The simplest way to customize that code to suit the fresh
context versions is to comment first line, so I can live with it.




Pawe/l