TeX, LaTeX and Friends

8:34 AM

Hi everyone !

Tom Mac Farlin explains in this article how to install WordPress to develop plugins. Is it possible to install MikTeX or TeXlive in the same mode so that one can compile documents either with the distributed version or with other versions of older packages or those under development?

tommcfarlin.com/subversion-for-wordpress-development

Ulrike Fischer

9:08 AM

@MarcelKrüger that seems to work, but one still doesn't see the used fonts in the log - very irritating ;-). The installation description is a bit unclear about the kpse. At first I thought is should go into the bin folder too, but that seems not to work, it has to be in the tex folder beside the other files. It would be useful to have a build.lua to update the files in the tex folder with l3build install.

Paulo Cereda

10:48 AM

quack

Paulo Cereda

1:21 PM

@MarcelKrüger ooh wacky engines

@MarcelKrüger we might need an arara rule for that

Marcel Krüger

4:56 PM

@UlrikeFischer Logging used files will be added ... at some later point. I added build.lua. The main motivation for having kpse.? not in the bin folder was to avoid storing stuff there as much as possible to make updates easier (assuming that the directory is symlinked and not copied. That could be changed though if you think that that would be simpler.

Marcel Krüger

5:36 PM

@PauloCereda Does arara support searching for a file using kpathsea which is then used as parameter?

Alan Munn

7:31 PM

I tried to fix the image markup here tex.stackexchange.com/q/553151/2693 but it didn't work. What did I do wrong?

samcarter_is_at_topanswers.xyz

7:49 PM

@AlanMunn Did you try with an empty line before [1]: ...?

Alan Munn

@samcarter_is_at_topanswers.xyz No. But moewe fixed it.

Ulrike Fischer

9:09 PM

@MarcelKrüger no it is fine, it is easier for now, I only thought that the description is a bit unclear here.

Marcel Krüger

@UlrikeFischer I tried to make it a bit clearer.

Ulrike Fischer

9:25 PM

@MarcelKrüger I just tried the l3build, it moves the files to tex/luametalatex. Is that intended? (previously you suggested /tex/luatex/luametalatex or /tex/lualatex/luametalatex imho)

Marcel Krüger

@UlrikeFischer The l3build should move the files to tex/luametatex/luametalatex actually.

Ulrike Fischer

@MarcelKrüger yes, that what I meant. If that is okay, I will have to adjust the texmf.cnf.

Marcel Krüger

@UlrikeFischer That's intentional. I moved them to a different texmf/tex subtrees some time ago because most of the code isn't LuaTeX compatible, so it seemed wrong to have it there. (I never updated the README though,, only the install script)

Ulrike Fischer

@MarcelKrüger the build.lua doesn't install the .ini file.

Marcel Krüger

9:41 PM

@UlrikeFischer Damn. But the latest upload should have a working version now.

Ulrike Fischer

@MarcelKrüger ;-). I can't rebuild the format ;-(. After I realized that I had to adapt the fmtutil.cnf, I now get errors because latex can't find ltexpl.ltx. All this development branches get a bit complicated ...

Marcel Krüger

@UlrikeFischer Did you pull the latest version? The error comes from old build.lua versions not installing luametalatex-ltexpl-hook.tex.

Ulrike Fischer

@MarcelKrüger ah, I thought it were a problem in latex. Now it worked.

Marcel Krüger

It is related to a hack involving ltexpl.ltx: The first time it's loaded, I actually loads luametalatex-ltexpl-hook.tex which then loads the real ltexpl.ltx. That was necessary because I needed a hook after \newtoks is defined but before ltexpl is loaded.

Ulrike Fischer

@MarcelKrüger I tried the pdfresource code and it doesn't like it ;-(

! token call, execute: [\directlua]:1: attempt to call a nil value (field 'setpagesattributes')

I guess most of the pdf backend will be different?

Marcel Krüger

9:52 PM

@UlrikeFischer Not really, I tried to stay more or less compatible, but mostly focused on the TeX side for now. So the pdf Lua library is basically missing right now. Do you have a list of the most important functions or is everything needed?

Ulrike Fischer

@MarcelKrüger I only used pdf.setpageattributes and pdf.setpageresources (and one could avoid it at the cost of an additional compilation as it is done with pdflatex).

Alan Munn

@MarcelKrüger What is it and why might someone want to use it?

Ulrike Fischer

@MarcelKrüger I can't create an uncompressed pdf, it doesn't react to \pdf_uncompress:

Marcel Krüger

@UlrikeFischer That's odd. It works for me.

Ulrike Fischer

@MarcelKrüger I tried the following, and I'm using the dev formats:

\RequirePackage{l3pdf}
\ExplSyntaxOn
\pdf_uncompress:
\ExplSyntaxOff
\documentclass{article}
\begin{document}
abc
\end{document}

Marcel Krüger

10:05 PM

@UlrikeFischer I tried it and get an uncompressed file. Weird. Can you send me the PDF you get?

Ulrike Fischer

@MarcelKrüger on the way

Marcel Krüger

@AlanMunn It is a LaTeX format on top of LuaMetaTeX, the new engine the ConTeXt team is developing for lmtx. (LuaMetaTeX is also sometimes called LuaTeX 2.0) It's supposed to be smaller and more flexible while also providing some new primitives.

Ulrike Fischer

@MarcelKrüger do you see a way to get harfbuzz or will this be lost?

Marcel Krüger

@UlrikeFischer It's not built-in, but we could add it similarly to kpse. That is a bit harder than in LuaTeX because symbols are missing, but by linking in a separate Lua library it still works.

luaharfbuzz might need changes to be Lua 5.4 compatible, I'll have to check that.

Ulrike Fischer

@MarcelKrüger oh does luametatex use lua 5.4?

Marcel Krüger

10:14 PM

@UlrikeFischer Yes, it already started with the alpha versions of Lua 5.4.

@UlrikeFischer The file you sent me is not compressed.

My PDF backend does not use hex-encode strings though and uses a more compact representation in many cases, so maybe it just looks different.

Ulrike Fischer

@MarcelKrüger well in my editor I see only chinese. It looks very compressed to me.

Alan Munn

@MarcelKrüger Thanks

Ulrike Fischer

@MarcelKrüger I tried in the editor too ^^^

Marcel Krüger

@UlrikeFischer Does the editor show you which encoding it uses?

Ulrike Fischer

@MarcelKrüger the editor shows utf-16le and winedt showed "uni" which is imho similar.

Marcel Krüger

10:23 PM

@UlrikeFischer That certainly explains the chinese. Interpreted in latin1, the file becomes

%PDF-1.7
%ð
4 0 obj
<</Length 104>>stream
BT
/F25 9.96264 Tf 0 Tr
148.712318 639.870475 Td
[(

Ulrike Fischer

@MarcelKrüger yes, but why does it guess this encoding? normally (eg. with pdf created with luatex) it works fine.

Marcel Krüger

@UlrikeFischer I don't understand why. There certainly isn't a BOM at the beginning.

Ulrike Fischer

@MarcelKrüger but the pdf marker at the begin is different: %ÌÕÁÔÅØÐÄÆ versus %ðŸ–‹ (at least that is what kdiff is showing me, it opened it fine).

Marcel Krüger

@UlrikeFischer Yes, that might be the issue. Can you try the version I just pushed?

(The weird characters are actually the UTF-8 encoding of U+1F58B LOWER LEFT FOUNTAIN PEN, I originally thought that it would be nice to start the document with a fountain pen emoji)

Ulrike Fischer

@MarcelKrüger the marker changed, I now see %Ã¤Ã¶Ã¼Ã¡Ã Ã³Ã²ÃºÃ¹ in kdiff, but it is still chinese in winedt ;-(.

Marcel Krüger

10:40 PM

@UlrikeFischer Oh, I forgot that my editor still saves as UTF-8 and not latin1. The new push should really fix it now.

Ulrike Fischer

@MarcelKrüger the marker is now %äöüß. But winedt and the editor still insist that it is utf16. I will try to check with a hex editor tomorrow.

manooooh

11:05 PM

Hello all!! This question is probably answered somewhere, but I could not find anything

In beamer class (using LaTeX->PDF), is it possible to automate the change of slides?

i.e. consider this MWE:

\documentclass{beamer}

\begin{document}

\begin{frame}{}
\begin{enumerate}
\item<2-> The United States of America \uncover<3->{(USA),}
\item<4-> commonly known as the United States (U.S. or US) or America,
\item<5-> is a country mostly located in central North America, between Canada and Mexico.
\only<6-7>{
\item<6-> It consists of 50 states, a federal district, five major self-governing territories, and various possessions.
\item<7-> At 3.8 million square miles (9.8 million km2), it is the world's third- or fourth-largest country by total area.

Suppose I forget to add a paragraph between \item<5-> and \only<6-7>, let's say \item<6-> New item.. To make this works, I must manually change the numbers from \only<6-7> to \item<8->, which may cause a fail if I forget to change a \uncover which is out there

My question is, can we automate some numbers? Not \only but commands like \uncover, or \item<...>? Something like:

\documentclass{beamer}

\begin{document}

\begin{frame}{}
\begin{enumerate}
\item<*-> The United States of America \uncover<*->{(USA),}
\item<*-> commonly known as the United States (U.S. or US) or America,
\item<*-> is a country mostly located in central North America, between Canada and Mexico.
\only<6-7>{
\item<*-> It consists of 50 states, a federal district, five major self-governing territories, and various possessions.
\item<*-> At 3.8 million square miles (9.8 million km2), it is the world's third- or fourth-largest country by total area.

The *- states for the regular "number-". We should also have -* and * (the last for only appears on a certain slide). \only can have or not depending what slides do I consider to show only once

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends