Font and index trouble (was Re: mathpple-a-like for ConTeXt?)

Font and index trouble (was Re: mathpple-a-like for ConTeXt?)

[Magnus Lie Hetland]

Hans Hagen <pragma@wxs.nl>:
>
> At 11:56 PM 2/20/2003 +0100, Magnus Lie Hetland wrote:
> 
> >I guess I'm just too green with ConTeXt, but I can't get this to
> >work... Must i call \usetypescriptfile[somethingorother] first?
> >(Actually, I get "undefined control sequence"...)
> 
> What version do you run?

Well, texexec --version returns 1.9. I'm using the version that comes
with Gentoo Linux.

I may well be missing something basic here -- I simply slapped your
\usetypescript commands at the beginning of my ConTeXt file (before
\starttext), and I get the following error messages:

! Undefined control sequence.
<recently read> \usetypescript

l.1 \usetypescript
                  [berry][ec]
! Undefined control sequence.
l.2 \usetypescript
                  [palatino][ec]
bodyfont       : unknown variant palatino

The commands were:

\usetypescript[berry][ec]
\usetypescript[palatino][ec]
\setupbodyfont[palatino]

Another thing that surprised me (not font-related) is that the \index
command gobbles up the following space, even when I use braces; i.e.

  Foobar, babar\index{indexentry} krangas

is typeset as

  Foobar, babarkrangas

Is this normal behaviour? Must I manually insert a space there?

> (let's move this discussion to the context list since it's not
> related to pdftex)

OK.

> Hans 

-- 
Magnus Lie Hetland               "Nothing shocks me. I'm a scientist." 
http://hetland.org                                   -- Indiana Jones

Re: Font and index trouble (was Re: mathpple-a-like for ConTeXt?)

[Hans Hagen]

At 09:50 PM 3/17/2003 +0100, you wrote:
>[snip]
>
>By the way: In 3.1, it still seems that the space after \index{...} is
>eaten. Is this the way it should?

yes, since the index should precede the word it refers to and not get 
decoupled (i.e. a page break in between); an \index command will even hook 
into the next paragraph if needed

Hans


-------------------------------------------------------------------------
                                   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
-------------------------------------------------------------------------

Re: mathpple-a-like for ConTeXt?

[Patrick Gundlach]

Magnus Lie Hetland <magnus=WUZL2GlSCMBAfugRpC6u6w@public.gmane.org> writes:

Hello,


> Can the tfm file be generated from these using texfont, or is uplr8t
> something completely different from uplr8a (as might seem to be the
> case)? In that case, is the uplr8t files (and related files) available
> somewhere?

you can use an alternative to the berry typescript. Get adobekb.tex
from http://levana.de/context and use 

\setupencoding[default=8r] 
\usetypescript[adobekb][\defaultencoding]
\usetypescript[palatino][\defaultencoding]
\setupbodyfont[palatino]

There is no need to install new fonts.

HTH,

Patrick

PS: let us move this discussion to the context mailinglist, since it
    is not pdftex related

Re: Re: mathpple-a-like for ConTeXt?

[Jens-Uwe Morawski]

On Tue, 18 Mar 2003 16:11:52 +0100
Patrick Gundlach <pg@levana.de> wrote:

> you can use an alternative to the berry typescript. Get adobekb.tex
> from http://levana.de/context and use 
> 
> \setupencoding[default=8r]

what is the advantage of using 8r instead of ec or texnansi?

Jens

8r vs. ec vs. texnansi

[Patrick Gundlach]

Jens-Uwe Morawski <morawski@gmx.net> writes:

Hello Jens and others,

>> \setupencoding[default=8r]
>
> what is the advantage of using 8r instead of ec or texnansi?

It is a matter of taste and the characters you need. There is no big
difference for my german and english texts.

texnansi does not work with the preinstalled (psnfss) fonts from
teTeX. So this leaves my preferred coding system to 8r or ec.

Sometimes I need glyphs that are encoded in 8r but not in ec (euro sign
for example). I never use glyphs that are encoded in ec but not in
8r. 

Patrick

Re: 8r vs. ec vs. texnansi

[Maarten Sneep]

On Tuesday 18 March 2003 23:39, Patrick Gundlach wrote:
> > \setupencoding[default=8r]
> >
> > what is the advantage of using 8r instead of ec or texnansi?
>
> It is a matter of taste and the characters you need. There is no big
> difference for my german and english texts.
(snip)
> Sometimes I need glyphs that are encoded in 8r but not in ec (euro sign
> for example). I never use glyphs that are encoded in ec but not in
> 8r.

But the disadvantage of 8r is that in most languages the line-breaking 
algorithm performs badly. Take the dutch "verifi\"{e}ren". If the hyphenation 
patterns are set up correctly, it should break like "ve-ri-fi-e-ren" (hope I 
got this right ;-), and you should loose the accent.

you can verify this by using \showhyphens{verifi\"{e}ren} in tex. In the log, 
the possible word breaks will be shown.

In the default 8r it doesn't break correcly, near the accent. In ec it does 
break well. I don't know about texnansi.

I'm sure others can give better examples, and explain why the wordbreaking 
doesn't work correctly.

Kind regards,

Maarten Sneep

Re: 8r vs. ec vs. texnansi

[Patrick Gundlach]

Maarten Sneep <sneep@nat.vu.nl> writes:

Hello Maarten,

> But the disadvantage of 8r is that in most languages the line-breaking 
> algorithm performs badly. Take the dutch "verifi\"{e}ren". 

so the 'funny' character is ediaeresis ? It is defined in 8r as well
as in ec, so the hyphenation does not make any differences here. (In
my test, I could not get rid of the accent, but the 8r and ec
hyphenation were identical.)

You are right though, this is why I wrote "in my german and english
texts".

> In the default 8r it doesn't break correcly, near the accent. In ec it does 
> break well. 

I could not confirm it here.

> I don't know about texnansi.

It should be exactly like ec/8r.

> I'm sure others can give better examples, and explain why the
> wordbreaking doesn't work correctly.

Because of the \accent macro???


Patrick

Re: Re: 8r vs. ec vs. texnansi

[Hans Hagen]

At 10:54 AM 3/19/2003 +0100, you wrote:

> > I'm sure others can give better examples, and explain why the
> > wordbreaking doesn't work correctly.
>
>Because of the \accent macro???

it's indeed this primitive or any kern/skip/penalty/box that stops the 
hyphenation process

so, your patterns should match the encoding (some languages have multiple 
patterns, for instahce czech has patterns for il2 and ec, and context can 
load multiple patterns per language and switch accordingly)

Hans
-------------------------------------------------------------------------
                                   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
-------------------------------------------------------------------------

Re: 8r vs. ec vs. texnansi

[Jens-Uwe Morawski]

On Tue, 18 Mar 2003 23:39:15 +0100
Patrick Gundlach <pg@levana.de> wrote:

> Jens-Uwe Morawski <morawski@gmx.net> writes:
> 
> 
> >> \setupencoding[default=8r]
> >
> > what is the advantage of using 8r instead of ec or texnansi?
> 
> It is a matter of taste and the characters you need. There is no big
> difference for my german and english texts.
 
> Sometimes I need glyphs that are encoded in 8r but not in ec (euro sign
> for example). I never use glyphs that are encoded in ec but not in
> 8r. 

some time ago Hans and me discussed how to make TS1 symbols available.
I've never tried it but it sounds interesting since it is an unified
way to access the symbols. For example for Palatino


add to the typescripts

\definefontsynonym [PalatinoSymbols] [<TS1-Font>]
\definefontsynonym [PalatinoSymbolsBold] [<TS1-Bold-Font>]

\definefontsynonym [SerifSymbols] [PalatinoSymbols]
\definefontsynonym [SerifSymbolsBold] [PalatinoSymbolsBold]

and define the symbols in a symbol-typescript

\definesymbol[textregistered] [{\symbol[\currentencoding-registered]}]
\definesymbol[ec-registered][\getglygh{SerifSymbols}{\char123}]

this could add the capability that the LaTeX textcomp package provides

Jens

Re: 8r vs. ec vs. texnansi

[Patrick Gundlach]

Jens-Uwe Morawski <morawski@gmx.net> writes:

Hello Jens and others,

> some time ago Hans and me discussed how to make TS1 symbols available.
> I've never tried it but it sounds interesting since it is an unified
> way to access the symbols. For example for Palatino
>
>
> add to the typescripts
>
> \definefontsynonym [PalatinoSymbols] [<TS1-Font>]
> \definefontsynonym [PalatinoSymbolsBold] [<TS1-Bold-Font>]
>
> \definefontsynonym [SerifSymbols] [PalatinoSymbols]
> \definefontsynonym [SerifSymbolsBold] [PalatinoSymbolsBold]
>
> and define the symbols in a symbol-typescript
>
> \definesymbol[textregistered] [{\symbol[\currentencoding-registered]}]
> \definesymbol[ec-registered][\getglygh{SerifSymbols}{\char123}]
>
> this could add the capability that the LaTeX textcomp package provides

At a quick glance this would fail for example for the euro
sign/ec. (Not there in ec). Why not access the existing virtual fonts
or generate new ones if needed? (phvr8c.vf for example - let us not
fall back to the urw/adobe discussion yet :) 

Patrick

Re: Re: 8r vs. ec vs. texnansi

[Jens-Uwe Morawski]

On Wed, 19 Mar 2003 14:58:48 +0100
Patrick Gundlach <pg@levana.de> wrote:

> Jens-Uwe Morawski <morawski@gmx.net> writes:
> 
> Hello Jens and others,
> 
> > some time ago Hans and me discussed how to make TS1 symbols available.
> > I've never tried it but it sounds interesting since it is an unified
> > way to access the symbols. For example for Palatino
> >
> >
> > add to the typescripts
> >
> > \definefontsynonym [PalatinoSymbols] [<TS1-Font>]
> > \definefontsynonym [PalatinoSymbolsBold] [<TS1-Bold-Font>]
> >
> > \definefontsynonym [SerifSymbols] [PalatinoSymbols]
> > \definefontsynonym [SerifSymbolsBold] [PalatinoSymbolsBold]
> >
> > and define the symbols in a symbol-typescript
> >
> > \definesymbol[textregistered] [{\symbol[\currentencoding-registered]}]
> > \definesymbol[ec-registered][\getglygh{SerifSymbols}{\char123}]
> >
> > this could add the capability that the LaTeX textcomp package provides
> 
> At a quick glance this would fail for example for the euro
> sign/ec. (Not there in ec). Why not access the existing virtual fonts
> or generate new ones if needed? (phvr8c.vf for example - let us not
> fall back to the urw/adobe discussion yet :) 

your glance was too quick. :) The chain is:

starting with texteuro symbol -->
if in ec-encoding it becomes ec-texteuro symbol -->
it takes char 0xBF from the current ...Symbol... font
 (the symbol macro should be able to resolve Serif.., Sans..., Serif...Bold etc
  automatically) -->
the current ...Symbol... font is *NOT* a EC-encoded font but a TS1 (aka 8c)
encoded one that is associated via typescript with the T1 encoded font
(as i have already written above: <TS1-Font>)

This would easily allow for other text encodings to define the usual
symbols, for example for 8r one can declare either

\definesymbol[8r-texteuro][\getglygh{Serif}{\char128}]

or better
\definefontsynonym [SerifSymbols] [Serif]
\definesymbol[8r-texteuro][\getglygh{SerifSymbols}{\char128}]

since in 8r or texnansi there are no special symbol fonts. Thus, when
you switch the encoding in your document from ec to 8r to texnansi
then the euro is from the users view always the symbol texteuro, but
for ec taken from the associated 8c-font and for 8r or texnansi
taken directly from the text-font.

So, we should look if the macro-names from textcomp package conflict
with macro-names already defined in ConTeXT. If there is no conflict
we can define all text-symbols according to 'textcomp' in some
typescripts:

\starttypescript[textsymbols]
  ...
  \definesymbol[texteuro] [{\symbol[\currentencoding-texteuro]}]
  ...
\stoptypescript

\starttypescript[textsymbols][ec]
  ...
  \definesymbol[ec-texteuro] [\getglygh{Textsymbol}{\char191}]
  ...
\stoptypescript

\starttypescript[textsymbols][8r]
  ...
  \definesymbol[8r-texteuro] [\getglygh{Textsymbol}{\char128}]
  ...
\stoptypescript


the font typescripts then have to declare what 'Textsymbol' is:

\starttypescript [serif] [Adobe Times] [name]
  \definefontsynonym [Serif]            [Times-Roman]
  \definefontsynonym [SerifBold]        [Times-Bold]
  \definefontsynonym [SerifItalic]      [Times-Italic]
  \definefontsynonym [SerifSlanted]     [Times-Italic]
  \definefontsynonym [SerifBoldItalic]  [Times-BoldItalic]
  \definefontsynonym [SerifBoldSlanted] [Times-BoldItalic]
  \definefontsynonym [SerifCaps]        [Times-Roman]
  \definefontsynonym [TextsymbolSerif]            [TextsymbolTimes-Roman]
  \definefontsynonym [TextsymbolSerifBold]        [TextsymbolTimes-Bold]
  \definefontsynonym [TextsymbolSerifItalic]      [TextsymbolTimes-Italic]
  \definefontsynonym [TextsymbolSerifSlanted]     [TextsymbolTimes-Italic]
  \definefontsynonym [TextsymbolSerifBoldItalic]  [TextsymbolTimes-BoldItalic]
  \definefontsynonym [TextsymbolSerifBoldSlanted] [TextsymbolTimes-BoldItalic]
  \definefontsynonym [TextsymbolSerifCaps]        [TextsymbolTimes-Roman]
\stoptypescript

\starttypescript [serif] [Adobe Times] [ec]
  \definefontsynonym [Times-Roman]                [ptmr8t]  [encoding=ec]
  \definefontsynonym [Times-Bold]                 [ptmb8t]  [encoding=ec]
  \definefontsynonym [Times-Italic]               [ptmri8t] [encoding=ec]
  \definefontsynonym [Times-BoldItalic]           [ptmbi8t] [encoding=ec]
  \definefontsynonym [TextsymbolTimes-Roman]      [ptmr8c]
  \definefontsynonym [TextsymbolTimes-Bold]       [ptmb8c]
  \definefontsynonym [TextsymbolTimes-Italic]     [ptmri8c]
  \definefontsynonym [TextsymbolTimes-BoldItalic] [ptmbi8c]
\stoptypescript
% the encoding=?? is only needed for mapping input; since we access symbols
% directly (\char...) the encoding for 'Textsymbol' fonts should not be important


\starttypescript [serif] [Adobe Times] [8r]
  \definefontsynonym [Times-Roman]                [ptmr8r]  [encoding=8r]
  \definefontsynonym [Times-Bold]                 [ptmb8r]  [encoding=8r]
  \definefontsynonym [Times-Italic]               [ptmri8r] [encoding=8r]
  \definefontsynonym [Times-BoldItalic]           [ptmbi8r] [encoding=8r]
  \definefontsynonym [TextsymbolTimes-Roman]      [ptmr8r]  [encoding=8r]
  \definefontsynonym [TextsymbolTimes-Bold]       [ptmb8r]  [encoding=8r]
  \definefontsynonym [TextsymbolTimes-Italic]     [ptmri8r] [encoding=8r]
  \definefontsynonym [TextsymbolTimes-BoldItalic] [ptmbi8r] [encoding=8r]
\stoptypescript


Jens

Re: Re: mathpple-a-like for ConTeXt?

[Magnus Lie Hetland]

Patrick Gundlach <pg@levana.de>:
[snip]
> you can use an alternative to the berry typescript. Get adobekb.tex
> from http://levana.de/context and use 
> 
> \setupencoding[default=8r] 
> \usetypescript[adobekb][\defaultencoding]
> \usetypescript[palatino][\defaultencoding]
> \setupbodyfont[palatino]

Thanks. I tried this, and got an error -- it couldn't find the proper
font file. So I removed the 8r encoding, and now it no longer
complains that it can't find the font, but it silently uses Computer
Modern _anyway_. :(

I've downloaded adobekb.tex to the current directory and used the
following code:

%\setupencoding[default=8r]
\usetypescript[adobekb]%[\defaultencoding]
\usetypescript[palatino]%[\defaultencoding]
\setupbodyfont[palatino]

\starttext

This is a test

\stoptext

> There is no need to install new fonts.

Well... There must be _something_ wrong with my setup (which is
default teTeX 2.0 in Gentoo).

-- 
Magnus Lie Hetland               "Nothing shocks me. I'm a scientist." 
http://hetland.org                                   -- Indiana Jones

adobekb.tex

[Patrick Gundlach]

Hello,

just for those who have the same problems.

>> you can use an alternative to the berry typescript. Get adobekb.tex
>> from http://levana.de/context and use 
>> 
>> \setupencoding[default=8r] 
>> \usetypescript[adobekb][\defaultencoding]
>> \usetypescript[palatino][\defaultencoding]
>> \setupbodyfont[palatino]

I forgot to say that you need to load the file with

\usetypescriptfile[adobekb]

Sorry about that.

Patrick

Re: adobekb.tex

[Guo Yang]

Hello Patrick,

>>>\setupencoding[default=8r] 
>>>\usetypescript[adobekb][\defaultencoding]
>>>\usetypescript[palatino][\defaultencoding]
>>>\setupbodyfont[palatino]
>>>      
>>>
Sorry if this question is very silly. Where does this 'palatino' come 
from? I tried 'helvetica', 'helvet' and 'utopia' but none of them 
worked, while 'times' worked well.

Regards,
Guo

Re: adobekb.tex

[Patrick Gundlach]

Guo Yang <gxy3139@njit.edu> writes:

Hello Guo,

> Sorry if this question is very silly. 

I don't think it is.

> Where does this 'palatino' come from? I tried 'helvetica', 'helvet'
> and 'utopia' but none of them worked, while 'times' worked well.

It comes from the file "type-exa.tex"  that is shipped with ConTeXt:


\starttypescript [palatino] [texnansi,ec,8r]
\definetypeface [palatino] [rm] [serif] [palatino]        [default] [encoding=\typescripttwo]
\definetypeface [palatino] [mm] [math]  [palatino]        [default] [encoding=\typescripttwo]
\definetypeface [palatino] [tt] [mono]  [computer-modern] [default] [rscale=1.075]
\stoptypescript

there are also defined:

\starttypescript [modern] [ec,il2,pl0,default]
\starttypescript [postscript] [texnansi,ec,8r]
\starttypescript [times] [texnansi,ec,8r]
\starttypescript [lucida] [texnansi,ec,8r]

and some others, not directly related.


Patrick

Re: colors

[Magnus Lie Hetland]

Patrick Gundlach <pg@levana.de>:
>
> 
> Hello,
> 
> for the colors, use \setupcolors[state=start] (plural).

Ah -- right, thanks. It works now. There is a typo in the beginner's
manual (the "excursion") on page 59 -- it says
\setupcolor[state=start]. (On the previous page, the general form is
mentioned as \setupcolors[...], but I simply copied the statement to
my code.)

> If you need, I can send you a small test document.

No problem. I was simply mislead by the documentation ;)

> Pat

-- 
Magnus Lie Hetland               "Nothing shocks me. I'm a scientist." 
http://hetland.org                                   -- Indiana Jones

Re: foonotes

[Magnus Lie Hetland]

[-- Attachment #1: Type: text/plain, Size: 1195 bytes --]

Patrick Gundlach <pg@levana.de>:
>
> Magnus Lie Hetland <magnus@hetland.org> writes:
> 
> >> could you send me the dvi (not pdf) file of the following output:
> >
> > Attached.
> 
> the \tfx and \tfxx don't so anything (they are the same size as \tf). 
> 
> I have absolutely no idea how this can happen. It is far beyound my
> knowledge of ConTeXt. You might think about 
> 
> a) trying typescript [times] instead of [palatino] (but I guess it
> won't do anything better)

Having to use times would certainly be a disappointment... :]

> b) send the dvi file to the list and ask the list (= Hans ;-} )

OK... Attaching dvi file again, and Cc'ing the list. (I hope
attachments are OK when they're as small as this?)

The erroneous version comes from the Solaris installation (which isn't
really mine). I ran the same code on the Linux installation, and there
\tfx and \tfxx worked just fine. (Still a bit confused about such
things as \ss and \sc here -- \ss seems to simply use Palatino, and
\sc doesn't seem to produce small caps...?)

> Patrick

-- 
Magnus Lie Hetland               "Nothing shocks me. I'm a scientist." 
http://hetland.org                                   -- Indiana Jones

[-- Attachment #2: test.dvi --]
[-- Type: application/x-dvi, Size: 2908 bytes --]