What is this openout_any=p problem we see so many of? Here is another one tex.stackexchange.com/q/564626/3929

@daleif where did you see another one?

@daleif The other ones were openin

@JosephWright oh, my bad then. What is the issue with those? I don't seem to have had any such issues.

@daleif if you set openin_any=p lualatex doesn't work at all (it will not input article.cls for example) it only inputs files from the current directory. pdflatex not quite as bad input works but \pdffilesize does not which breaks the new expandable tests for if file exists

looks as if again using --output-directory was the culprit.

@UlrikeFischer --make-configuring-the-workflow-as-hard-as-possible

@DavidCarlisle ;-)

@DavidCarlisle so some kind of security setting?

@UlrikeFischer I still do not understand those people who insist on redirecting stuff.

@daleif well yes that's all it is (p for paranoid in this case) but it is supposed to be (don't input dot files and files given by absolute path) not (as interpreted by luatex) don't input dot files or any files not under the current directory.

@daleif They want a wordprocessor-like result: just the input .tex and the .pdf in their working space. It's understandable.

@daleif well I do understand them. It is sometimes difficult to find the relevant files, if there lots of log,out,aux,toc, etc lurking around. But if I wanted a clean folder, I would use the l3build setup: copy everything in a temporary folder and then switch to the folder and compile there.

$ ls -l *.aux | wc -l
754

I have no sympathy ^^^ :-)

Oh .log even more interesting, perhaps I should clear out more after tests...

$ ls -l *.log | wc -l
ls: cannot compare file names ‘zzb.log’ and ‘zzbegr\374\337ung3.log’: Invalid or incomplete multibyte or wide character
4496

quack

@DavidCarlisle I had a collegue where I found 2500+ files in a single folder. Most was versions of chapters etc of this book project. He hadn't thought about using version control for his project. Long time ago I had another collegue who named all his files zzz and then the folder told you what this was.

@daleif oh my

@PauloCereda ls says I have 20717 files in /tmp which is where I usually run mwe from here :-)

@UlrikeFischer I don't think it is that bad unless you have many latex projects (each with its own full latex doc) in the same folder. Why not a folder pr project.

@DavidCarlisle ooh :)

@daleif well I have seen documents which created five or six index, and acronyms, and glossaries, and so on. That creates quite a number of auxiliary files.

@UlrikeFischer I agree also that having the intermediate processing files in a subdirectory will be cleaner. I find myself running latexmk -c more often than not. Maybe we could teach latexmk to move then in and out a subdir... hmmm... or prefixing all of them with a dot in front after compilation and remove the dots before...

@wtsnjp hmm, let me read that again (although won't be able to do much before the evening) Actually as I drafted that bit it may be better if someone else comments @JosephWright @UlrikeFischer it is hard to proof read your own code

how do you do new lines in \node? As in \begin{tikzpicture} \node<1->[anchor=west, overlay] at (0, 2.7) {\tcbox{hello\ how are you?}}; \end{tikzpicture}

@daleif you worked together with @DavidCarlisle?

@Anush (DejaVu?) you have to add the align option to the node...

@DavidCarlisle well one could add in the definition [1] denotes the number of arguments (here one) and #1 denotes the first argument, but as \newcommand\kw[0] actually works, I don't think that one need to discuss the case (and start a discussion about missing spaces ...).

@Anush but you probably need the equivalent for \tcbox there

@Rmano have a look at ltx2any

@Rmano thank you

@samcarter_is_at_topanswers.xyz oh, nice. Didn't know, will look at it!

@UlrikeFischer @wtsnjp would Ulrike's suggestion look better to you? just adding in the definition [1] denotes the number of arguments (here one) and #1 denotes the first argument

@DavidCarlisle Yes, I think that small note works fine for lesson-06 :)

@wtsnjp I'll sort something out this evening... How is the Japanese translation going? Have you done many lessons yet?

@Rmano ltx2any does the compilation in a tmp or subfolder and then brings back the pdf to your main folder. Everything stays nice and clean and avoids all the problems of build folders

@samcarter_is_at_topanswers.xyz I didn't try it yet --- but a question, is the tmp folder "stable"? Re-doing the compilation every time would be quite time-consuming. I normally have latexmk running in continuos mode so I have live preview each time I save...

@DavidCarlisle I'm 'slowly' working on that. I've just completed lesson-01 to lesson-06 (and no more-XX has been done). I have noticed additional notes are needed a bit more than I expected, even in the main contents, for the Japanese version. This is because we are using a different engine from the first place (e.g., I have to explain that just using babel is not really enough for Japanese documents).

But for now, I just postpone thinking much about such things and rather just translating as it is in the English version.

@Rmano I think with tmp it tries to reuse if all the files are still there, but I'm not sure (you can also specify the folder yourself to be on the safe side)

github.com/reitzig/ltx2any

@wtsnjp yes we don't currently have any non latin script translations, there are Japanese, Chinese and Marathi translations "in progress" but I would not be surprised if they all have some "issues" to get good text/example balance in practice. Hopefully it works out OK in the end...

(disadvantage of ltx2any compared to latexmk is that it always does one compilation, even if nothing changed)

@wtsnjp for the babel I don't know. Perhaps you can say "if you are writing a document in a European Language, do this [translation of current example description, but with the existing German text], else [some new text specific to Japanese, or a forward reference to the language-xx lessons or...]

@DavidCarlisle Well, we (I consulted with a few Japanese TeX users including Hisanobu) decided to primarily focusing on Japanese documents with pLaTeX in the main lessons. Thus, English texts (and with other Europian languages) are treated as secondary. This makes the situation further complicated...

You might think we should treat English texts in the main course, but our thought is that is meaningless. Since Japanese people who want to create English documents are already good at reading English, they don't really need Japanese translation. We assume that the expected readers of the Japanese version are people who want to write Japanese documents with LaTeX.

So yes, "getting good text/example balance in practice" is exactly the thing we are aiming for.

@wtsnjp I agree that for most of it you should switch to Japanese being the main focus but I wonder if for just the babel lesson that means you can not say much sensible and there you should keep the existing German text, just as in the English version the text is German (supplied by @UlrikeFischer so it might even be correct German:-)

@DavidCarlisle Yes, so I left some examples intentionally in English, including the babel example for instance.

@wtsnjp OK I leave it with you, I can barely read the German, so I am not going to argue too much about how you phrase the Japanese:-)

@DavidCarlisle Thanks. For now, I am almost working alone, but I am planning to ask other Japanese speakers to check the translation because I can't proofread perefectly my own code ;-)

@wtsnjp :-) but at least you don't have the problem of getting the letters "the" in the right order. @UlrikeFischer claims I have sometimes written "teh"

@DavidCarlisle Inverting letters? Since you only type \zzzz the problem doesn't show.

@egreg zz\zz

@daleif I had a colleague who claimed his computer had "run out of space". He was storing all his files on the desktop and ran out of screen.

look as if I broke shadings in xdvipdfmx ;-(

@UlrikeFischer upholding tradition

@DavidCarlisle someone has to ;-). But this inch + hoffset business is really a pain. Everywhere people are adding and removing it, and at the end you no longer know where you are ...

@AlanMunn easy fix, increase the resolution

@UlrikeFischer Well yes, I think the issue is a few people tried to 'fix' things rather then leave well alone, and soon it gets very complex

New personal record: [ 1/196, ??:??/??:??] update: acmart [37k] (55004 -> 56343) ... done

@Skillmonlikestopanswers.xyz wow. Let's start a palindrome chase for that, too!

@Skillmonlikestopanswers.xyz How many weeks did you not update?

@samcarter_is_at_topanswers.xyz that was my PC at work, I had two weeks of vacation, was sick for a week, am back for a week now and had much stress the week before the vacation, so at least a month, I'd say.

quack

could someone on linux try luaotfload-tool --version?

[paulo@cambridge ~] $ luaotfload-tool --version

luaotfload-tool:
        Luaotfload font management and diagnostic utility.
        License: GNU GPL v2.0.
        Report problems to <https://github.com/latex3/luaotfload/issues>

luaotfload-tool version: "3.15"
Revision: nil
Lua interpreter: stock; version "Lua 5.3"
Luatex version: 1.12
Platform: type=unix name=linux
    +  machine: x86_64
    + nodename: cambridge
    +  release: 5.8.11-200.fc32.x86_64
    +  sysname: Linux
    +  version: #1 SMP Wed Sep 23 13:51:28 UTC 2020

@UlrikeFischer ^^

so it works. Thanks.

@UlrikeFischer On zref again, looking at \zref@labelbylist vs \zref@labelbyprops, it feels to me like they are N- and n-type variants for a comma-list argument: make sense?

    \usepackage[
    	backend=biber,
    	style=alphabetic,
    	citestyle=alphabetic,
    	sortlocale=en_US,
    	url=true,
    	doi=true,
    	eprint=false
    ]{biblatex}
    \usepackage{bibentry}

    ....
    \bibentry{RS:12} % prints nothing!

from the documentation I get the feeling that it's only intended for bibtex

@ComFreek No, the standard styles provide a \fullcite command that gives you the same effect,

wonderful

works

ooh presidential debate tonight

unfortunately, i cannot get Biblatex to pick up the bib files

This is a recurring problem. Why is doing bibliography so brittle?

Or is this a problem of latexmk?

@ComFreek In general it's not

@ComFreek Sounds likely: Biber and BibTeX work fine usually

I have been facing such strange issues across computers and set-ups (latexmk, TeXStudio's built-in build system)

Usually, if I can get to work once, then it works

(until I clean cache files)

@ComFreek does your workflow work in the terminal, using typical tool calls instead of build tools?

No, usually I press build in TexStudio

and I use

% !TEX TS-program = latexmk -xelatex -shell-escape -silent -latexoption="-synctex=1" -f %

which is picked up my TeXStudio

But issuing that command on the CLI doesn't help either

@ComFreek @PauloCereda Means doing things one step at a time

oh, never tried that

I don't even know what I would call there :D

@ComFreek I don't use TeXStudio, but latexmk works consistently for me with TeXShop.

I don't use these automation tools, they are naughty and do naughty things.

oh wait

Ooopsie

@ComFreek What exactly do you mean by "pick up the bib files"

I mean that in my document \cite{abc} is typeset as \textbf{[abc]}

@PauloCereda Fake news has made it to chat.

@ComFreek So biber isn't being run automatically even though latexmk should detect it?

possibly

how does latexmk detect it?

i smell a problem resulting from my multi-file setup with \inputs

where arguments to \inputs are macros...

(e.g. \input{\main/config/config})

@ComFreek that's a good question, but this happens sometimes even in a standard setup (a big caveat of latexmk usage in Overleaf)

@ComFreek \input (unlike \include) doesn't affect the aux file so unlikely to affect bib processing

I meant that if latexmk detected I use biber by grepping files recursively for "biber" and \inputs, then that would be a problem.

Hence, my question on how latexmk does it

@ComFreek the normal log file has the message to run biber (or bibtex) so \input unlikley to be the issue but without any example, hard to know:-) (I never use latexmk personally) if I want biber to run, I type biber seems simpler to me:-)

@ComFreek Looks for a .bcf file I suspect (that's how l3build does it)

oh, I hope it looks in the correct output directory

namely ./build

@ComFreek people who use an output directory deserve no sympathy:-)

I get eye cancer the moment my folder contains more than 5 different files

@ComFreek There's your problem in a nutshell

@ComFreek There may be ways to configure latexmk to deal with that, but I've never bothered to try, since I need sympathy. :)

Fortunately, I get enough sympathy from other software engineers such that I can break yours :-)

in .latexmkrc I already have $out_dir = 'build';

all temp files are then successfully put there

@AlanMunn :D

not sure whether latexmk is internally wired correctly to look there for bcf files, too

@DavidCarlisle ooh this is good

@yo' this sounds really strange like a non-deterministic bug

@ComFreek unfortunately it's deterministic (on the same input, it always fails or always succeeds), but nobody is able to determine it :)

@ComFreek it's been a hard day's night :)

oh wait, it's bug not beetle

/ba dum tss

@PauloCereda Appropriate Beatles song responses to this: Tell me why, No reply or You can't do that :)

@AlanMunn oh look at all the wacky people

:)

@UlrikeFischer Hope you like zref2 ;)

@JosephWright just on the way home, will look then.

@UlrikeFischer Have fun

@JosephWright Number of document requests: 248

@DavidCarlisle WOW

@DavidCarlisle Did you like my suggestion to marmot?

@JosephWright 12 users so someone using it a lot

@DavidCarlisle Yup!!

@JosephWright Yes I'd actually started to write something about arxiv being more prominent in some subjects than others, but I got busy and didn't send it

@DavidCarlisle Well quite: for me with my work hat on it's a non-starter, and there are loads of economists, linguists, ... using LaTeX

@DavidCarlisle She might have reasons.

@HaraldHanche-Olsen Fake News

@JosephWright ?

@HaraldHanche-Olsen ooh more counters

@DavidCarlisle ooh more news

@ComFreek I have a strange deja vu today, but have a look at ltx2any - avoids all problems with build directories and keeps your folder clean

Is there a quick and simple way to not show any numbers in the bibliography list when using a plain \begin{thebibliography} (i.e. no bibtex or anything advanced)?

I understand that this may sound pointless (the whole point of thebibliography is to add those numbers) but I'm working with an existing document and I only need to switch this off temproarily. I hope not to have to rewrite the thing.

@Szabolcs the label is the optional argument to \bibitem so you can use [] or [zzz] or whatever

@Szabolcs or probably \begin{thebibliography}{99}\renewcommand\theenumiv{}

@samcarter_is_at_topanswers.xyz ooh another good reference

@DavidCarlisle @samcarter_is_at_topanswers.xyz for context: gitlab.com/islandoftex/arara/-/issues/46

@PauloCereda impressed with my Portuguese? github.com/learnlatex/learnlatex.github.io/commit/…

@DavidCarlisle very impressed!

@PauloCereda maybe you will finally understand \newcommand now!

@DavidCarlisle It works with maçãs and laranjas, but how about chocolate?

@DavidCarlisle twitter.com/nixcraft/status/1310987404627439618

@PauloCereda and not forgetting panquecas de pato

@DavidCarlisle oh no

@DavidCarlisle Unfortunately, that didn't work. I bit the bullet and started to rewrite it to use biblatex instead.

@DavidCarlisle I will visit this city someday: en.wikipedia.org/wiki/Pato_Branco

Now this solution doesn't work with biblatex :((( tex.stackexchange.com/a/26347/6302

@Szabolcs oh no

Someone help mr. Mathematica bloke

Found the same question ^

@samcarter_is_at_topanswers.xyz regarding ltx2any, I've found this repository: github.com/reitzig/ltx2any/tree/master, last update was in 2018 (same for dev branch)

Seems dead?

This is unexpectedly cumbersome. A quick fix was to use backend=bibtex (i.e. not biber) with biblatex. Didn't notice bad effects on my simple reference list, but I'm sure there are some in more complex cases ...

@Ulrike Thanks! It wouldn't be much work to do that for the few affected references, but it would break compatibility with graphical .bib editors such as JabRef (I assume; didn't try, but here it says it's not compatible: github.com/JabRef/jabref/issues/4558 )

@JosephWright ! Argument of \loadrefdata has an extra }. ;-(

If you're aware of a better GUI editor, I'm interested. I'm getting tired of JabRef's bugs (every release brings more...)

But for now backend=bibtex is good enough :-) Thanks again!

Hey there everyone

Stupid question: is it possible to easily change all the enumerate labels in all lists in the document? I can't seem to find anything

Better put: what's the easiest way to do something like that?

@JosephWright I think you meant \tl_to_str:n { \loadrefdata } not \tl_to_str:N \loadrefdata

My workday ends and many of you are already experiencing early evening.

@ComFreek er I have code last updated 1990s 2018 isn't that long ago:-)

@OttavioBartenor \renewcommand\theenumi{\Alph{enumi}} perhaps? (probably better to use enumitem package settings

@ComFreek ... and how does the update date matter if something works?

@OttavioBartenor Which documentclass are you using?

@DavidCarlisle Nice! Works fine on \documentclass{book}

@PauloCereda a ltx2any rule for arara would solve the problem :)

(I'm wondering a bit where the sudden wave of "I don't like auxiliary files" comes from. There have been multiple questions about this on stack overflow during the last week)

@samcarter_is_at_topanswers.xyz nobody ever says "I don't want my cross references to resolve"

@JosephWright usage is definitely picking up, you tug video perhaps?

Tue Sep 29 20:51:02 UTC 2020
  Number of distinct users:      14
  Number of document requests:  314
   ... Number from learnlatex:  311
  Number of documents returned: 219
  Number of unique user agents:  11
Mon Sep 28 23:58:28 UTC 2020
  Number of distinct users:       7
  Number of document requests:  171
   ... Number from learnlatex:  171
  Number of documents returned: 148
  Number of unique user agents:   7
Sun Sep 27 23:55:58 UTC 2020
  Number of distinct users:      10
  Number of document requests:   90

@DavidCarlisle Maybe they are just too shy? Clearly this seems to be the wish of a lot of people :)

@samcarter_is_at_topanswers.xyz outdated code is a sure sign of open bugs. I never use a library that's older than 6 months

also new features won't be implemented, and only few tools are feature-complete

not even the GitHub issues are being responded to

@DavidCarlisle crazy :o

@ComFreek you do know the update frequency for tex-the program????

(for context: I am an npm JS library kind of guy)

@DavidCarlisle roughly, but the core tex is incomparable to build wrappers.

they are on whole different levels

but funnily enough, one would think that microcode in CPUs would also only get rarely updated

but in fact, it gets regularly updated

@ComFreek the tex world world is far more change averse than hardware. The vast majority of the code in tex the program and the latex format comes from the 1980s and early 90s.

@PauloCereda you just can't trust arara

WELP

This is a new one

I've managed to bread \end{document}

OH, fixed it

That was strange

Hi guys, is this a good place to ask beginner level questions? I posted on StackExchange but my question has not received much traction there