June 2017 - ntg-context - NTG Mailing Lists

Re: [NTG-context] Log message "Unknown filter 'DCTDecode'"
by Christoph Reller 21 Jun '17

21 Jun '17

On Mon, 19 Jun 2017 12:21:00 +0200 luigi scarso <luigi.scarso(a)gmail.com > wrote: > With poppler 0.55, > I am able to reproduce it with > > $ convert -size 100x100 radial-gradient: rgradient.jpg; convert > rgradient.jpg rgradient.pdf > $ cat test.tex > > \starttext > \externalfigure[rgradient.pdf] > \stoptext > > Syntax Error (623): Unknown filter 'DCTDecode' > > I'm investigating > > -- > luigi > Thank you Luigi, This error message is gone now. Sorry for the confusion I have created by not having reported a well defined way to produce this message. Cheers, Christoph

1 0

question about STIX fonts (versions one and two)
by Pablo Rodriguez 19 Jun '17

19 Jun '17

Dear list, I have the following sample: \setupbodyfont[stixtwo] \starttext \input zapf \stoptext Even if \setupbodyfont[stix] is used, I don’t get any text. The fonts seem to be missing. I have checked this with a new ConTeXt Suite install and the same results. Are the STIX fonts distributed with the ConTeXt Suite? Many thanks for your help, Pablo -- http://www.ousia.tk

2 2

Log message "Unknown filter 'DCTDecode'"
by Christoph Reller 19 Jun '17

19 Jun '17

Hi, In recent versions of ConTeXt/luatex, I observe the following log entry: Syntax Error (77727): Unknown filter 'DCTDecode' This happens if the document includes JPEG images, e.g: \starttext \externalfigure[hacker.jpg] \stoptext The output PDF looks, however, fine. Does this message have any significance? Cheers, Christoph

3 6

Re: [NTG-context] typesetting serial letters in Context
by r.ermers＠hccnet.nl 18 Jun '17

18 Jun '17

Dear Context’ers, I am trying to use the letter module in mkiv for typesetting serial letters from an xml database. Thanks, Thomas for looking into this. I appreciate it very much. Yet I don’t know either whether I actually need the letter module, but it seemed reasonable to try this first. Today I tried to insert some simple luatex code which I found on another forum. The results are promising: a. The address appears. b. The salutation + formalname also appear. Yet there are some problems: 1. the address elements appear in one line, not underneath as desired, 2. When I add text such as Knuth or references to the database, Context stops. See the code below. My question at this point is: how can I proceed? Do I actually need the letter module? Kind regards, Robert \usemodule[letter] \starttext \startbuffer[participants] <?xml version="1.0" encoding="UTF-8"?> <document> <contacts> <contact class='participant' label='Hendriks'> <formalname>Hendriks</formalname> <informalname>Karel</informalname> <prefix>mr.</prefix> <initials>K.</initials> <courselevel>levelA</courselevel> <coursedate>1 june 2017</coursedate> <city>Arnhem</city> <address> Mr. K. Hendriks Grotestraat 5 1234 Arnhem </address> </contact> <contact class='participant' label='Janssen'> <formalname>Janssen</formalname> <informalname>Piet</informalname> <initials>P.</initials> <prefix>mr.</prefix> <courselevel>levelA</courselevel> <coursedate>1 june 2017</coursedate> <city>Nijmegen</city> <address> Mr. P. Janssen Kortestraat 8 1234 Nijmegen </address> </contact> </contacts> </document> \stopbuffer \startxmlsetups xml:setups \xmlsetsetup{#1}{contacts|contact|address|prefix}{xml:*} \stopxmlsetups \xmlregistersetup{xml:setups} \startxmlsetups xml:contacts \xmlflush{#1} \stopxmlsetups \startxmlsetups xml:p \xmlflush{#1} \par \stopxmlsetups \startxmlsetups xml:contact \defineletterelement[layer][address][example]% {\correspondenceparameter{toname}\\ \correspondenceparameter{toaddress} \par} \setupletterlayer[address][alternative=example] \startletter [opening={{Dear \xmltext{#1}{/prefix} \xmltext{#1}{/formalname},}}, closing={Kind regards,}, signature={Sender}, toaddress={{\xmltext{#1}{/address}}}, subject={Interesting offer}] %\input knuth %{\xmltext{#1}{/informalname}} %\xmltext{#1}{/address/p} \stopletter \stopxmlsetups \startluacode userdata = userdata or { } userdata.letter = { } \stopluacode \xmlprocessbuffer{main}{participants}{} \stoptext > Op 18 jun. 2017, om 17:33 heeft Thomas A. Schmitz <thomas.schmitz(a)uni-bonn.de> het volgende geschreven: > > On 06/15/2017 11:58 PM, r.ermers(a)hccnet.nl wrote: >> Thusfar I have been trying to use the letter module with the xmltext and xmlflush commands. What can be done to make this work? >> Thanks! > > I'm sorry, I don't use the letter module and don't have it installed. As far as I can see, your way of using the xml commands is correct. There appears to be some sort of incompatibility between the letter module and xml, but I guess only Wolfgang (the author of the module) can really say what's going on. If you want to typeset simple letters, maybe you can bypass the module? > > Thomas

1 0

Choosing font range
by Hans Åberg 17 Jun '17

17 Jun '17

Is it possible have say Pagella in math, and also in \tt except for the ASCII range, where a monospace font is used? (Both for largely untranslated Unicode points.)

3 13

STIX vs XITS
by Hans Åberg 17 Jun '17

17 Jun '17

It says that XITS is an OpenType version of STIX on [1], but the latter now supports it. Does that mean that one can now use the latter directly? 1. http://wiki.contextgarden.net/Math_fonts

2 2

Selecting font for \starttyping
by Hans Åberg 17 Jun '17

17 Jun '17

How does one choose font for \starttyping? It is apparently different from that of \tt.

2 1

compressing register ranges
by Florian Grammel 17 Jun '17

17 Jun '17

Using [compress=all] for registers yields somewhat puzzling results (1.0.4 (TeX Live 2017)). I would expect the register of the following code to become one single range "1--7", though context produces "1--2, 3--7". Is this the expected behaviour? And if so: is there another switch to get the ranges to be collapsed completely? Also: There wouldn't be a shorthand to define explicit ranges? LaTeX has \index{Demorange|(} to \index{Demorange|)} which is actually very convenient. Best regards, Florian. \defineregister [demo] [demoes] \starttext 1 \demo{Demorange} \page[yes] 2 \demo{Demorange} \page[yes] 3 \startregister[demo][demorange]{Demorange} \page[yes] 4 \demo{Demorange} \page[yes] 5 \page[yes] 6 \stopregister[demo][demorange] \page[yes] 7 \demo{Demorange} \page[yes] \placeregister [demo] [compress=all] \stoptext ____________________________________________ Florian Grammel Copenhagen, Denmark

2 2

Re: [NTG-context] Bug regarding placement and size of primes
by Mathias Schickel 16 Jun '17

16 Jun '17

> Am 12.06.2017 um 17:10 schrieb Hans Hagen <pragma(a)wxs.nl>: > > On 6/12/2017 4:54 PM, Mathias Schickel wrote: >>> the alternative is that i waste time again and again on getting something to work for all fonts and all combinations which leads to a mess that i then need to maintain without any gain >> At first: Thank you very much for your work! >>> so, i decided to follow a route that works for most fonts / cases and not some messy hybrid solution (and i don't want a different setup or bag of tricks for each font that is out there as that is a long term no-go) >> But I have to contradict you here: It has nothing to do with special or hybrid fonts. The problem persists at present (using the LuaTeX version that is shipped with the Garden version) in /all/ fonts. The primes are not rendered as they should in /any/ font I tried out. >> As Mikael told me, you optimised the code for LuaTeX 1.05 which seems to be terminated September this year. But until then the output of primes will be not correct. I have no possibility to switch back to the current (non-beta) version of ConTeXt, because this version contains the bugs with primes you have already resolved. > > there will probably be a 1.0.5 one of these days (also for other reasons) but Mojca has to configure the garden compile farm > Okay, that would be great. But could you tell me how I can get the files and compile them? Mikael did write me instructions that work for Linux, but I am on MacOS. My problem would be solved if I could get the source and compile LuaTeX 1.05 on my own on MacOS. :-) >> This means that ConTeXt is not usable for me until LuaTeX 1.05 will be released or until I am able to compile the source of LuaTeX 1.05 on my own. Some of the latest betas (the version one or two weeks ago) did solve all problems regarding primes. The problem I am referring to rose up in the most recent betas. As Mikael told me, this is because you introduced new code to work with LuaTeX 1.05. This is very nice, but because this version is not yet released, this decision is questionable, since this makes ConTeXt unusable until this release (if you need primes). >> So I appreciate your hard work, but I would be nice to have a version that runs with the current LuaTeX version of the beta of ConTeXt (or alternatively it would be nice if instructions are provided to compile the source of LuaTeX for every OS that can be used simply). > > then just use old versions and then accept the bugs and limitations in those versions How can I get the beta of last week or the week before? In those betas you have not yet included the new code (for use with LuaTeX 1.05) and everything worked really fine. The --context=current garden version has several bugs regarding primes you have solved in the mentioned betas. So it would make no sense at all to use this version. I would need one of the older betas or the most recent one with LuaTeX 1.05. > >> And (to repeat myself) the problem has nothing to do with special fonts of hybrid version or whatever. You get the problem using standard Latin Modern or any font else. Simply try > latin modern has all kind of issues ... and it's inconsistent with cambria, other gyre fonts etc ... in a decade of luatex and open type math fonts i captured a lot of font issues in runtiem fixes but that will always give side effects > > so, i can fix somethign for you with lm and then someone else complains about lucida etc etc > I am not using Latin Modern. I only mentioned that because I wanted to stress that I am not referring to exotic fonts. (I totally agree on you regarding Latin Modern; I think Cambria should be the reference since this font is the font that established the opentype math standard.) In fact I am using Minion Math (which is a really stunning font) or Stix 2. There had been no problems in those fonts with primes in the betas of the last weeks. And as Mikeal suggests the problems are resolved if I could use LuaTeX 1.05. So in fact the only thing I am complaining about is that everything worked one or two betas ago until it is „broken“ in the newest beta since this one seems to be based on code that requires LuaTeX 1.05 (which makes sense of course since this will be the upcoming release) that is not shipped alongside the beta. So I am not able to understand why code is provided in the beta that requires a LuaTeX version that is not shipped alongside (and especially in view of the fact that we had a perfectly working beta regarding primes last week; this means I think it would have been a great idea to wait including the newest code for primes that requires LuaTeX 1.05 until LuaTeX 1.05 is shipped with the beta too). :-) > as i explained before, primes are a mess conceptually .. not a real character and not a proper superscript candidate either so when one corrects for one aspect one has to fight the other > > also, we have unicode and therefore multiple primes are one character > > i simply have no time (nor motivation) to spend days on some temporary solution that will be complained about also and discarded a while later anyway I perfectly understand that. And every work regarding primes is done, as I think. But one way to make us of it would be to use the beta of last week (or maybe the week before) or the most recent one together with LuaTeX 1.05. But I have no idea how to make use of that ways. :-) So I am simply asking: How can I use older betas or how can I get the source of LuaTeX 1.05 on a Mac and compile it? > Hans Mathias > > ----------------------------------------------------------------- > Hans Hagen | PRAGMA ADE > Ridderstraat 27 | 8061 GH Hasselt | The Netherlands > tel: 038 477 53 69 | www.pragma-ade.nl <http://www.pragma-ade.nl/> | www.pragma-pod.nl <http://www.pragma-pod.nl/> > -----------------------------------------------------------------

4 8

A solved problem
by Eric Détrez 16 Jun '17

16 Jun '17

Hello Just a message to explain a problem I had and the answer. The trouble : the big parenthesis were not big, $($ and $\bigl($ gave the same shape The fix : I defined my own environment in which I defined very big font sizes. \definefontsize[e] \definebodyfontenvironment [default] [e=4] One of them was \definefontsize[g] I changed it to \definefontsize[gg] and all went right again The explanation (?) : maybe \big was read \bi (bold italic) with size g ???? I had this problem for 2 years, I'm happy to have figured it now. -- Eric Détrez Informatique Lycée Faidherbe, Lille http://eric-detrez.fr

2 1