« first day (3596 days earlier)      last day (1632 days later) » 

00:00
@PhelypeOleinik It's regarding stripping of braces in PGF keys: github.com/pgf-tikz/pgf/issues/368
@PhelypeOleinik That's a pretty good idea. I'll see whether I can integrate that/
 
6 hours later…
cis
cis
05:42
Can someone tell me if this document is translating for him?
% arara: pdflatex
% arara: pdflatex
% arara: pdflatex

\documentclass[a4paper, landscape]{article}
\pagestyle{empty}
\usepackage[showframe=false,
margin=20mm,
]{geometry}
%\geometry{paperwidth=15cm, paperheight=30cm, margin=2mm}% optional
%\usepackage{mwe} % Dummy Images
\usepackage{textcomp} % \textcopyright

\usepackage{tikz}
\usetikzlibrary{calc}
\pgfkeys{/tikz/savevalue/.code 2 args={\global\edef#1{#2}}}
%\pagecolor{lightgray!22} % see page margins

\usepackage[most]{tcolorbox}
\tcbset{
NoGaps/.style={boxsep=0mm, left=0pt, right=0pt, top=0pt, bottom=0pt, before skip=0pt, after skip=0pt,},
I was told it wouldn't go through. But everything works for me.
cis
cis
06:04
...
...
Not all volunteers at once, please! :()
Hey Can I convert an image in Mathcha to LaTex?
Anonymous
Dear @cis ,

All auxiliary files removed,
I removed the copyrigh sign, used \usepackage{mwe} and removed \usepackage{texcomb}
Anonymous
Still can't compile. I will upload the new log file:
cis
cis
@Maybank I don't use Matcha, but there is
Export --> Export Selection to LaTeX
Anonymous
06:19
@cis @cis Try to use overleaf.com it also produces this error:
Anonymous
Anonymous
yo'
yo'
@barbarabeeton Why warm clothes? For instance today we have 26C high and 17C low, warmer than enough :-)
@VincentMiaEdieVerheyen you are not compiling @cis example above. Beside this: your system is rather old, and debugging a pgf document using some intermediary package versions is up to impossible.
cis
cis
@VincentMiaEdieVerheyen @UlrikeFischer If I test the code at https://cocalc.com/ it runs as desired.

So updtae your TeXLive maybe.
Anonymous
06:33
@cis This code, from this chat, compiles fine for me as well :)
Anonymous
What I receive errors with is when I try to compile the 'New version:' (code most to the bottom) of your answer https://tex.stackexchange.com/questions/560883/automated-ruler-how-to-have-one-set-of-rulers-around-each-picture-2-pictures/560909#560909

with the changes detailed in https://chat.stackexchange.com/transcript/message/55442153#55442153
cis
cis
@VincentMiaEdieVerheyen Sry, I fear I copied the old code. But the 'New Version' runs under cocalc.com too....
% arara: pdflatex
% arara: pdflatex
% arara: pdflatex

\documentclass[a4paper, landscape]{article}
\pagestyle{empty}
\usepackage[showframe=false,
margin=20mm,
]{geometry}
%\geometry{paperwidth=15cm, paperheight=30cm, margin=2mm}% optional
%\usepackage{mwe} % Dummy Images
\usepackage{textcomp} % \textcopyright

\usepackage{tikz}
\usetikzlibrary{calc}
\pgfkeys{/tikz/savevalue/.code 2 args={\global\edef#1{#2}}}
\pgfdeclarelayer{background}
\pgfdeclarelayer{foreground}
\pgfsetlayers{background,main,foreground}
@VincentMiaEdieVerheyen @VincentMiaEdieVerheyen Here
http://latex.informatik.uni-halle.de/latex-online/latex.php
it does not run, but this page uses an older version too:

This is pdfTeX, Version 3.14159265-2.6-1.40.19 (TeX Live 2019/dev/Debian) (preloaded format=pdflatex)
:55444716
tlmgr update --self

tlmgr update --all
;)
Anonymous
@VincentMiaEdieVerheyen you can't update texlive 2019 to texlive 2020 with tlmgr
Anonymous
I see, I am not sure how to get texlive 2020 on Mac. All I ever used for Tex/LaTeX is TeXShop ...
06:48
@VincentMiaEdieVerheyen althouigh they are the generic unix-like commandline instructions the mactex wrapper may have a more mac intuitive interface to that (but I know nothing about macs)
cis
cis
@VincentMiaEdieVerheyen In all cases you can test the code under

https://cocalc.com/doc/latex-editor.html

(without Login!)
https://cocalc.com/doc/latex-editor.html
Anonymous
@cis Yes, I see. It compiles there wonderfully :) though CoCalc seems to be very slow.
Anonymous
I am trying to update TeX Live as per tug.org/texlive/upgrade.html now ... will see if I can manage that :)
Anonymous
Very neat that the scaling of the ruler and the images now go together, without the horizontal spacing!
cis
cis
@VincentMiaEdieVerheyen CoCalc is SageTeX it is not so optimal for TeXing, if you do not need Sagemath-content.
07:03
@VincentMiaEdieVerheyen if this the code cis shows below: if fails with texlive 2018 but works in 2019 (final) for me.
@UlrikeFischer Just checking the code: we need that space?
Anonymous
@DavidCarlisle I feel the explanation at tug.org/texlive/upgrade.html is a bit incomplete for a newbie.
Anonymous
I am stuck at step 4. now.
Anonymous
It details the command

tlmgr path remove

But it does not clearly identify which path to remove... Not sure how to find out.
@VincentMiaEdieVerheyen did you install vanilla texlive 2019 from tug, or macTeX initially?
07:11
@JosephWright well the printed output was wrong, it had lost the space, and so I got test failures but I'm not sure if it is needed to write to the pdf.
Anonymous
@DavidCarlisle I would be inclined to say macTeX.
@VincentMiaEdieVerheyen that step is optional and means type literally type tlmgr path remove it will just undo what it did before.
@VincentMiaEdieVerheyen then don't follow those instructions, do whatever the mactex instructions are.
@UlrikeFischer Ah, right: just looking it seems odd, as LuaTeX ends up with two spaces: I can't see why we need that second one
@UlrikeFischer Could you check github.com/pgf-tikz/pgf/issues/900#issuecomment-686898357 - I'm doing so now
@UlrikeFischer So the lack of space doesn't break the PDF? In that case, we should probably drop it and update our tests
@JosephWright yes I'm just trying to create the file for a test.
@JosephWright which one do you want to drop? the ~ or the \c_space_tl?
@UlrikeFischer The ~: we need the \c_space_tl (with expansion) for pdfTeX, so may as well keep it for LuaTeX
07:17
@JosephWright I need to check if I have a test, testing the writing in the pdf.
@JosephWright did you see my dvipdfmx report?
@UlrikeFischer Yes, that would be good: I think a CTAN update can wait as only we are likely using this code ATM
@UlrikeFischer Yes: hopefully we can get that working and also the colour business
@JosephWright I wonder what to do about the version problem. I mean you can't test from latex if the specials are there or not. Would it make sense if we suggest to add a file which would declare the (x)dvipdfmx version? Or do you have another idea how to query this?
07:58
@barbarabeeton Microsoft woes
@DavidCarlisle oh no
Anonymous
Done that now. First uninstalled: I skipped the last step though https://tug.org/mactex/uninstalling.html 'Uninstalling Ghostscript' ... don't know if that is a problem.

Now installed 2020
Anonymous
Yeeeaaay @cis It compiles now without a problem :)
08:15
@VincentMiaEdieVerheyen Happy New Year
Anonymous
@cis Your code should be adapted in its comments to:
Anonymous
\coordinate[] (LLM) at (LeftRasterbox.south east);% Lower Left Middle
\coordinate[] (ULM) at (LeftRasterbox.north east);% Upper Left Middle
\coordinate[] (LRM) at ([xshift=\HelpBoxRule]RightRasterbox.south west);% Lower Right Middle
\coordinate[] (URM) at ([xshift=\HelpBoxRule]RightRasterbox.north west);% Upper Right Middle
Anonymous
It currently reads:
Anonymous
\coordinate[] (LLM) at (LeftRasterbox.south east);% Upper Left Middle
\coordinate[] (ULM) at (LeftRasterbox.north east);% Upper Left Middle
\coordinate[] (LRM) at ([xshift=\HelpBoxRule]RightRasterbox.south west);% Upper Left Middle
\coordinate[] (URM) at ([xshift=\HelpBoxRule]RightRasterbox.north west);% Upper Left Middle
@PauloCereda just noticed texdoc scratch3 (a proper programming language)
08:16
@DavidCarlisle ooh Français
@DavidCarlisle coin coin
@PauloCereda Petit déjeuner
@DavidCarlisle oh non
/quacks in despair in French
@UlrikeFischer you could get dvipdfmx --version allowed in restricted shell escape.... (although perhaps a file would be quicker?)
@UlrikeFischer For xdvipdfmx we can use XeTeX version as a proxy
Anonymous
@cis The X-rulers are working perfectly now!
Anonymous
08:22
\xRuler{LL}{LLM}{below}{-4mm}{-2mm}
\ifnum\ShowHelps=1
\begin{pgfonlayer}{foreground}
\draw[Help, transform canvas={yshift=15mm}] (LL) -- +(\w,0) node[Help, pos=0.67, above=2pt] {Left Rasterbox width w= \w \\ No. x divisions w/u= \NoXdiv};
\end{pgfonlayer}\fi

\xRuler{LRM}{LR}{below}{-4mm}{-2mm}
\ifnum\ShowHelps=1
\begin{pgfonlayer}{foreground}
\draw[Help, transform canvas={yshift=15mm}] (LRM) -- +(\w,0) node[Help, pos=0.67, above=2pt] {Right Rasterbox width w= \w \\ No. x divisions w/u= \NoXdiv};
Anonymous
Thank you ♥
@UlrikeFischer @JosephWright or you don't test but have new and old back end files and rely on the distributions to ensure that the backend file matches the distributed driver
@JosephWright yes, but we can't be sure that they update "in sync".
@DavidCarlisle you missed another teh ;-)
Anonymous
@cis For the y-rulers to be around each image on each side, there will need to be a minimal amount of spacing between the images.
@UlrikeFischer nope
cis
cis
08:25
@VincentMiaEdieVerheyen Glad to hear.
@VincentMiaEdieVerheyen If necessary, you could do the same procedure for width with height.
Is there a way to avoid spaces after commands in error messages? If I do stuff like \msg_new:nnn { mypackage } { begin_document_not_found } { There~was~no~\exp_not:N\begin{document}.}, then I obviously get “There was not \begin {document}”, which looks wrong in my eyes. I tried writing \textbackslash begin{document} instead, but it just printed \textbackslash
@Gaussler use \string not \noexpand (translated to l3 :-)
@DavidCarlisle Nice, so that would be \token_to_str:N. Thanks, that worked! :-)
Anonymous
@cis If I try:
Anonymous
% Image Ruler y
\newcommand\yRuler[5]{
\path let \p1=($(#1)-(#2)$) in
\pgfextra{ \pgfmathsetlengthmacro{\RasterboxesWidth}{veclen(\x1,\y1)} }
[savevalue={\h}{\RasterboxesHeight}];
\pgfmathsetmacro\NoYdiv{\h/\u}  % Number of x dividions

\draw[Ruler] (#1) coordinate(Y) -- (#2);
\foreach \n in {0,1,...,\NoYdiv}{%%
\draw[Ruler] ([yshift=\n*\u]Y) -- +(#4,0) node[#3, inner sep=1pt]{\ifnum\n=1 \n\,cm \else \n\fi};
}%%
\foreach \n in {0.1,0.2,...,\NoXYiv}{%%
\draw[] ([yshift=\n*\u]Y) -- +(#5,0);
}%%
}
08:34
@DavidCarlisle the new pdftex.def contains two comments:
%%% this definition needs moving from graphicx.sty to graphics.sty
%%% this definition will be added to graphics.sty
Should the commands have been removed?
Anonymous
Then I get the following error:
Anonymous
Anonymous
OK, I notice some discrepancies in that snippet:D will try to update first
Anonymous
It's working now :) :)
@UlrikeFischer no graphics is only updated in dev
08:45
@DavidCarlisle ah, right.
@UlrikeFischer Sure, but to a first approximation the only people who update the binaries are developers
@DavidCarlisle Was my thinking
@JosephWright that wasn't what I mean: xetex and dvipdfmx development isn't done by the same persons. I don't think that we can be sure that a change in the second is reflected in a version change in xetex.
@UlrikeFischer no but if you wait until a new dvipdfmx is in texlive and miktex then ship a new dvipdfmx.def then the only people who have mis-matched files will know what they did.
@UlrikeFischer What @DavidCarlisle said
@UlrikeFischer Bottom line is there are no conditionals in dvipdfmx, so we can't test other than by proxy
@JosephWright of course @UlrikeFischer has 50 live git branches checked out, 26 of them with local texinput settings and she never knows where she is:-)
09:01
@UlrikeFischer I think realistically, we can't make any changes until TL'21 pre-testing: that should have updated dvipdfmx out-of-the-box
@JosephWright well yes, but if I write tests or new code, I need a switch from somewhere, so that I can either add a message "sorry doesn't work" or the working code. I guess I will add a file and adjust the search pathes for the two binaries.
@UlrikeFischer For testing, don't you just need the errorlevel from dvipdfmx exit?
Anonymous
09:19
@cis Please consider the following code, where I have tried to use your code to obtain 2 y-rulers per image:
Anonymous
% arara: pdflatex
% arara: pdflatex
% arara: pdflatex

\documentclass[a4paper, landscape]{article}
\pagestyle{empty}
\usepackage[showframe=false,
margin=20mm,
]{geometry}
%\geometry{paperwidth=15cm, paperheight=30cm, margin=2mm}% optional
%\usepackage{mwe} % Dummy Images
\usepackage{textcomp} % \textcopyright

\usepackage{tikz}
\usetikzlibrary{calc}
\pgfkeys{/tikz/savevalue/.code 2 args={\global\edef#1{#2}}}
\pgfdeclarelayer{background}
\pgfdeclarelayer{foreground}
\pgfsetlayers{background,main,foreground}
Anonymous
Do you think the ends of all of the YELLOW help lines are placed correctly?
Anonymous
I am also curious to understand why the

Left Rasterbox width
is smaller than the
Right Rasterbox width
@JosephWright but you don't get error if you use an unknown special. It is simply ignored. (I didn't check yet if you get errors if you use the extended color syntax). But I didn't mean only automated test, but also while extending the code. The code would be clearer if there where a switch "\ifnewdvipdfmx{...}{....}"
cis
cis
09:35
@VincentMiaEdieVerheyen Yes. It is to much work, to correct pgflinewidth every time.
@VincentMiaEdieVerheyen Can't say, why example-image.jpg and example-image-a.jpg have slightly different widths. If using 2 original images, e.g. mrHdl and cmHgI it should be correct.
Anonymous
@cis Please see this code, where I also included the different widths and heights of the images. They are the same width and height? But still, left rasterbox width and the rasterbox width are not the same somehow ...
Anonymous
% arara: pdflatex
% arara: pdflatex
% arara: pdflatex

\documentclass[a4paper, landscape]{article}
\pagestyle{empty}
\usepackage[showframe=false,
margin=20mm,
]{geometry}
%\geometry{paperwidth=15cm, paperheight=30cm, margin=2mm}% optional
%\usepackage{mwe} % Dummy Images
\usepackage{textcomp} % \textcopyright

\usepackage{tikz}
\usetikzlibrary{calc}
\pgfkeys{/tikz/savevalue/.code 2 args={\global\edef#1{#2}}}
\pgfdeclarelayer{background}
\pgfdeclarelayer{foreground}
\pgfsetlayers{background,main,foreground}
Anonymous
how can I show the inequalities one at a time in the second math line here:
\documentclass{beamer}
\newcommand{\var}[1]{\operatorname{var}{\!#1}}
\begin{document}
\begin{frame}
\visible<+->{
Intro text}
\begin{align*}
\visible<.->{\Pr(Y_r >0) &= \Pr(Y_r \geq 1) \leq \frac{\mathbb{E}(Y_r)}{1} = \frac{d}{2^r}}
\visible<+->{
\intertext{More text,}
\Pr(Y_r=0) &\leq \Pr(|Y_r-\mathbb{E}(Y_r)| \geq d/2^r) \leq \frac{\var(Y_r)}{(d/2^r)^2} \leq \frac{2^r}{d}
}
\end{align*}
\end{frame}
\end{document}
under "More text"
09:50
@samcarter_is_at_topanswers.xyz ^^ another reason for GL: ^^ it shows a emoji near my name. :D
cis
cis
As I said: that lines not so proper as the values. Since one would have to add or subtract \pgflinewidth or 0.5\pgflinewidth or ... here.

If the numerical values ​​should be the same, you have to use example-image both times or example-image-a both times.

The decisive factor is whether everything is fine with \def\ShowHelps{0}.
@Anush try
\documentclass{beamer}
\newcommand{\var}[1]{\operatorname{var}{\!#1}}
\begin{document}
\begin{frame}
\visible<+->{
Intro text}
\begin{align*}
\visible<.->{\Pr(Y_r >0) &= \Pr(Y_r \geq 1) \leq \frac{\mathbb{E}(Y_r)}{1} = \frac{d}{2^r}}
\visible<+->{
\intertext{More text,}
\visible<+->{\Pr(Y_r=0)} & \visible<+->{\leq \Pr(|Y_r-\mathbb{E}(Y_r)| \geq d/2^r)} \visible<+->{\leq \frac{\var(Y_r)}{(d/2^r)^2}} \visible<+->{\leq \frac{2^r}{d}}
}
\end{align*}
\end{frame}
\end{document}
@PauloCereda ohhh
thank you
@PauloCereda too many ducks, there was a time...
@DavidCarlisle ooh
10:00
@PauloCereda recognise that bloke?
Who's that guy?!
@DavidCarlisle vaguely. :)
@Anush You're welcome!
@PauloCereda wayback machine never forgets
@DavidCarlisle ooh a machine
@samcarter_is_at_topanswers.xyz asking a related question on topanswers that might be trivial... I don't know
10:02
@DavidCarlisle From which year is this? I vaguely remember a bee duck, but that is also a couple of years ago
@samcarter_is_at_topanswers.xyz 8 years ago, I believe. :)
@samcarter_is_at_topanswers.xyz ooh the bee duck
@Anush I'll take a look
thanks
@samcarter_is_at_topanswers.xyz 2014
Anonymous
10:04
@cis But the values are different. I am using both times example-image-a now and Left yields 301.12 and the Right Rasterbox Width yields 304.82
Anonymous
% arara: pdflatex
% arara: pdflatex
% arara: pdflatex

\documentclass[a4paper, landscape]{article}
\pagestyle{empty}
\usepackage[showframe=false,
margin=20mm,
]{geometry}
%\geometry{paperwidth=15cm, paperheight=30cm, margin=2mm}% optional
%\usepackage{mwe} % Dummy Images
\usepackage{textcomp} % \textcopyright

\usepackage{tikz}
\usetikzlibrary{calc}
\pgfkeys{/tikz/savevalue/.code 2 args={\global\edef#1{#2}}}
\pgfdeclarelayer{background}
\pgfdeclarelayer{foreground}
\pgfsetlayers{background,main,foreground}
The Land Before Time is a 1988 animated adventure drama film directed and produced by Don Bluth and executive produced by Steven Spielberg, George Lucas, Kathleen Kennedy, and Frank Marshall. The film stars the voices of Gabriel Damon, Candace Hutson, Judith Barsi and Will Ryan with narration provided by Pat Hingle. Produced by the American companies Amblin Entertainment and Lucasfilm, and the American-Irish Sullivan Bluth Ltd., it features dinosaurs living in the prehistoric times. The plot concerns a young "Longneck" (Apatosaurus) named Littlefoot, who is orphaned when his mother is killed by...
@VincentMiaEdieVerheyen 3.3pt difference so I'd put money on a missing % at end of line somewhere.
@DavidCarlisle ooh you are slowing turning into egreg
@PauloCereda OH NO
8
10:11
@DavidCarlisle you didn't see the new auxhook testfile yet ;-)
@PauloCereda If you squint a bit, one can almost see the first green pixel in the avatar
@samcarter_is_at_topanswers.xyz oooooh
@UlrikeFischer ooh @PauloCereda will be so pleased
@DavidCarlisle ooh
@UlrikeFischer Better than the breakfast/dinner file from yesterday(?)
10:14
@samcarter_is_at_topanswers.xyz oh no
@samcarter_is_at_topanswers.xyz not very different I would guess
@DavidCarlisle OH NO
:55446244 > \duckB=macro:
->breakfast.
blame @UlrikeFischer
cis
cis
@VincentMiaEdieVerheyen
It looks like tcolorbox measures the rasterboxwidth like

- Measure left from the left boxrule to the end of the left content (1 time boxrule).

- measures right from the middle boxrule to the end of the right boxrule (2 times boxrule)

But this becomes obsolete if the box rule is set to 0pt, as it should be in the result.
I already noted that, it seems to be a peculiarity of tbcitemize.
But, as I said, shouldn't be wild.
@VincentMiaEdieVerheyen I think my theory is true, because if I set \def\HelpBoxRule{11mm} the differenze is much larger:
@VincentMiaEdieVerheyen Sry, bad done: Set \def\HelpBoxRule{20pt} and you will see: right 40pt more....
...
....
....
I set up

"The Big tcolorbox Theory" :()
10:34
@samcarter_is_at_topanswers.xyz regarding the bracket, stix has it:
\documentclass{article}
\usepackage{stix}
\begin{document}

$ \lbrackubar abc \rbrackubar$
\end{document}
@UlrikeFischer nice! So maybe some font master can setup the DeclareUnicodeChar to use it even if using CM in the rest of the formula. @samcarter_is_at_topanswers.xyz I am still a n00b with font switching...
@Rmano there are lots of questions on the main site about "how to use a single symbol from a math font". I have to go now so I can't search for one.
@UlrikeFischer Yes, I'll do --- I have to check a "bimodal synchro" class stuff now for Monday so I am a bit in a rush too. Will try to learn later! Thanks!
yo'
yo'
@PauloCereda best pun ever in this room!
10:57
@UlrikeFischer I'm slowly working on DeviceN - it's pretty intricate! (Unlike colorspace, I'd like to be able to deal with the alternative space, etc., automatically)
@UlrikeFischer I'm likely to leave a few 'for the future' gaps - I suspect we may need to offer 'do it yourself' methods for the most complex color spaces
11:18
@JosephWright yes, we should document what is needed to declare a color space and then wait for some eager package writer ;-).
cis
cis
@VincentMiaEdieVerheyen Basically you want it that way somehow
\def\hbr{\HelpBoxRule}
\coordinate[shift={(\hbr,\hbr)}] (LL) at (LeftRasterbox.south west);% Lower Left
\coordinate[shift={(-\hbr,\hbr)}] (LR) at (RightRasterbox.south east);% Lower Right
\coordinate[shift={(\hbr,-\hbr)}] (UL) at (LeftRasterbox.north west);% Upper Left
\coordinate[shift={(-\hbr,-\hbr)}] (UR) at (RightRasterbox.north east);% Upper Right
\coordinate[shift={(0,\hbr)}] (LLM) at (LeftRasterbox.south east);% Upper Left Middle
\coordinate[shift={(0,-\hbr)}] (ULM) at (LeftRasterbox.north east);% Upper Left Middle
@UlrikeFischer Thanks! I added an answer
cis
cis
@VincentMiaEdieVerheyen Then it is also true with the auxiliary view. But in my opinion one could have lived without this correction, which the colored help lines at the end are set to 0pt anyway.
% arara: pdflatex
% arara: pdflatex
% arara: pdflatex
% arara: ghostscript: {  resolution: 166  , device: png16m}

\documentclass[a4paper, landscape]{article}
\pagestyle{empty}
\usepackage[showframe=false,
margin=20mm,
]{geometry}
%\geometry{paperwidth=15cm, paperheight=30cm, margin=2mm}% optional
%\usepackage{mwe} % Dummy Images
\usepackage{textcomp} % \textcopyright

\usepackage{tikz}
\usetikzlibrary{calc}
\pgfkeys{/tikz/savevalue/.code 2 args={\global\edef#1{#2}}}
\pgfdeclarelayer{background}
Anonymous
11:33
@cis You nailed it! 🙏.
Anonymous
11:54
What remains a bit strange now, but it is not an immediate problem to me at the moment (because many of my objects are sort of squarish in dimensions, such that I will take pictures of equal height), is that, if you take 2 images, of very different heights -> then the Y-length of the Y-rulers remain the same, i.e. adapted to the image of largest height.
Anonymous
@cis What would be more important for me at the moment is the possibility to fiddle with the separation between the 2 images, such that the Y-rulers in the middle become legible.
12:11
I am using basic BibTeX, no special packages (no natbib, biblatex or anything of the sort). Where is it controlled how many author names to allow before just using et al.? Is this usually hardcoded in the .bst file? Or is there an option for it?
I'm open to editing the bst, I just need to figure out what to change ...
cis
cis
@VincentMiaEdieVerheyen They are readable: tex.stackexchange.com/a/560909/46023
12:34
@UlrikeFischer Indeed: as it is, this is looking pretty mind-bending - almost certainly we will want to have some 'I know what I'm doing' interface at some stage
12:58
@yo' <3
13:09
@DavidCarlisle I'm surprised that you can't close a longtable question. Don't you have diamond badge?
2
Anonymous
@cis They are readable as long as the image behind is is not too black in that area, but what I mean is: the rulers overlap with the image next to it ... It would be better if there was enough horizontal whitespace between the images, such that the y-ruler of image 1 stays clear and seperated from the image and the y-ruler of image 2.
Anonymous
I think only 1 variable would be needed to achieve that.
Anonymous
@cis Please note in your answer at https://tex.stackexchange.com/a/560909/67761 that the comments are not correct currently:

\coordinate[shift={(0,\hbr)}] (LLM) at (LeftRasterbox.south east);% Upper Left Middle
\coordinate[shift={(0,-\hbr)}] (ULM) at (LeftRasterbox.north east);% Upper Left Middle
\coordinate[shift={(\hbr,\hbr)}] (LRM) at ([xshift=0]RightRasterbox.south west);% Upper Left Middle
\coordinate[shift={(\hbr,-\hbr)}] (URM) at ([xshift=0]RightRasterbox.north west);% Upper Left Middle
@UlrikeFischer I think that only works for some things like duplicates
Anonymous
13:19
@DavidCarlisle oh, I always thought one has to be extra careful with close vote with a tag badge.
Note to self: where possible avoid reading your students' LaTeX source, especially if they use Overleaf.
4
@AlanMunn Note to self: where possible avoid your students
4
Note to self: read note to self
4
13:41
@PauloCereda meta.codidact.com/418 (refresh a couple of times to see the different messages :) )
yo'
yo'
@AlanMunn /diminishing in guilt
@yo' :) the non-stop mode promotes really bad coding it seems. 80 errors with an apparently usable output.
Anonymous
@cis Perhaps? ->
Anonymous
RasterStyle/.style={
raster equal skip=75pt,
yo'
yo'
@AlanMunn we're coming there (slowly, in incremental steps, but we are).
13:50
@yo' Also some of the error fix suggestions (I think these are generate by Overleaf) are decidedly bad advice. To fix missing $ it says to use $ for inline and $$ for display, for example.
yo'
yo'
@AlanMunn we're aware of some. Any specific example in mind? (Note that there's an internal struggle to or not to add more of these. You can guess on which side I'm staying :-) )
@samcarter_is_at_topanswers.xyz ooh
@yo' see edit
@AlanMunn don't make start talking about arara. :)
@PauloCereda I'm just thankful you're a duck. An arara in here would be unbelievably annoying. :)
13:53
@AlanMunn ooh louro quer bolacha :D
418
stat error

The Spanish Inquisition raised an unexpected error. Cannot continue without comfy-chair-interrogation.
@PauloCereda My favourite:
@samcarter_is_at_topanswers.xyz ^^
>The Spanish Inquisition raised an unexpected error. Cannot continue without comfy-chair-interrogation.
@samcarter_is_at_topanswers.xyz HA!
@samcarter_is_at_topanswers.xyz <3
yo'
yo'
@AlanMunn ouch.
13:54
@PauloCereda :)
yo'
yo'
@AlanMunn thanks, noted.
@yo' Yeah, that one's pretty bad. Also, I suspect that somewhere in the docs there's a suggestion to use \newline. This is almost certainly wrong.
@PauloCereda That was like playing stone, paper, spock
@samcarter_is_at_topanswers.xyz ooh
yo'
yo'
@PauloCereda it's not my shift today :-) but wait a second.
13:59
@yo' ooh
cis
cis
14:10
@VincentMiaEdieVerheyen What do I know, try it out. I was of the opinion that you want everything to cement together.
@UlrikeFischer who's putting stuff in everyjob and messing up the l3 banner?
$ lualatex small2e
This is LuaHBTeX, Version 1.12.0 (TeX Live 2020)
 restricted system commands enabled.
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/small2e.tex
LaTeX2e <2020-02-02> patch level 5


 L3 programming layer <2020-09-03>
$ pdflatex small2e
This is pdfTeX, Version 3.14159265-2.6-1.40.21 (TeX Live 2020) (preloaded format=pdflatex)
 restricted \write18 enabled.
entering extended mode
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/small2e.tex
LaTeX2e <2020-02-02> patch level 5
L3 programming layer <2020-09-03>
@UlrikeFischer @MarcelKrüger can I say "I really never want to know this" to this information? :-)
@DavidCarlisle you mean the empty lines? I wondered about them too, did we have luaotfload messages there previously?
Lua module: luaotfload-main 2020-09-02 3.15 luaotfload entry point
Lua module: luaotfload-init 2020-09-02 3.15 luaotfload submodule / initializatio
n
Lua module: lualibs 2020-08-31 2.72 ConTeXt Lua standard libraries.
Lua module: lualibs-extended 2020-08-31 2.72 ConTeXt Lua libraries -- extended c
ollection.
Lua module: luaotfload-log 2020-09-02 3.15 luaotfload submodule / logging
Lua module: luaotfload-parsers 2020-09-02 3.15 luaotfload submodule / parsers
Lua module: luaotfload-configuration 2020-09-02 3.15 luaotfload submodule / conf
@UlrikeFischer empty lines + a final space before L3, they are from the Log I think ^^^^
@UlrikeFischer we should at least get the L3 programming layer banner out first
yo'
yo'
@PauloCereda still not sure what you needed to test :)
@DavidCarlisle less lines per file than this one:
(c:/Users/Nililand2020/Documents/TeXroots/gitinstall/tex/latex-dev/tools/longta
ble.sty
Package: longtable 2020/01/07 v4.13 Multi-page Table package (DPC)
\LTleft=\skip49
\LTright=\skip50
\LTpre=\skip51
\LTpost=\skip52
\LTchunksize=\count184
\LTcapwidth=\dimen139
\LT@head=\box48
\LT@firsthead=\box49
\LT@foot=\box50
\LT@lastfoot=\box51
\LT@cols=\count185
\LT@rows=\count186
\c@LT@tables=\count187
\c@LT@chunks=\count188
\LT@p@ftn=\toks15
)
14:15
@UlrikeFischer but they are short enough not to auto-wrap and put a line in the terminal output (but allocation logging that by default is pretty useless as well
@DavidCarlisle you could hunt the spurious spaces ;-)
@UlrikeFischer the real problem is the line wrap, that's what triggers terminal output (I think)
@DavidCarlisle not the first line ;-) But I guess we could shorten the messages so that they stop after the version number.
@DavidCarlisle Don't worry, blank lines are not significant ;)
@UlrikeFischer came about as I was wondering about asking the OP to post the log from my MWE here tex.stackexchange.com/questions/560480/lualatex-0-95-needed/…
14:24
@yo' <3
@DavidCarlisle well if we want to shorten the messages we need either to change ltluatex as we are using luatexbase.provides_module for this, or we need a special function, or we should change all "description" key into something else that luatexbase doesn't use.
@UlrikeFischer hmm
@DavidCarlisle that is the output if I add a -- in the right place in ltluatex:
Lua module: luaotfload-main 2020-09-02 3.15
Lua module: luaotfload-init 2020-09-02 3.15
Lua module: lualibs 2020-08-31 2.72
Lua module: lualibs-extended 2020-08-31 2.72
Lua module: luaotfload-log 2020-09-02 3.15
But it doesn't resolve the problem with the empty lines: there are messages about the cache and other which are quite long because of the pathes.
@UlrikeFischer I was just going to say that:-) can we simply arrange that the L3 banner comes first?
14:34
@DavidCarlisle Have to re-arrange where it loads (again)
@UlrikeFischer you just need to re-order the everyjob hook code.....
@DavidCarlisle Or that, yes
@JosephWright or just decide that no one looks at the terminal output anyway, and leave it?
@DavidCarlisle Well yes
what was the setting to make the lines longer?
14:36
@UlrikeFischer I see a plan has arisen for the dvipdfmx version business
@DavidCarlisle What I said earlier: blank lines don't count for proper tools :)
@JosephWright although it's a bit embarrassing to ask someone to post the log from the 1 line MWE \documentclass{novel}\begin{document}zzz\end{document} in default mode with no debug tracing, when it produces over 2000 lines of log file :(
@DavidCarlisle Ah, history
@JosephWright and then we have @AlanMunn wondering why users never look at the log.
@JosephWright extractbb ;-) nice trick.
@JosephWright 2000 not 200
14:38
@DavidCarlisle I mean 'there is history to putting lots of lines in the log, whereas today one would likely choose to have it very short'
@UlrikeFischer Yup: just checked it does give different values on differnet systems
we could have a \jobname.warn and send warnings there rather than the log (in some future life)
@DavidCarlisle :)
@DavidCarlisle log? what is this log thing you speak of?
4
@AlanMunn it's a wooden thingy
@DavidCarlisle well I really want the date and the version number, that is useful (as it is useful to see which files/packages are loaded and their version). But the description can go, and we could format some of the messages better, and perhaps silence the callback messages <-- @MarcelKrüger
14:42
@DavidCarlisle I can cut 700 lines with \ExplSyntaxOn \msg_redirect_class:nn { info } { none } \ExplSyntaxOff
@DavidCarlisle Some of it is clearly stray debug data, for example
The property list \polyglossia@langsetup contains the pairs (without outer
braces):
>  {latex/script}  =>  {latin}
>  {latex/lcscript}  =>  {latin}
>  {latex/scripttag}  =>  {}
>  {latex/language}  =>  {Latex}
>  {latex/langtag}  =>  {ENG}
>  {latex/bcp47}  =>  {}
>  {latex/hyphennames}  =>  {english}
>  {latex/direction}  =>  {LR}
>  {latex/hyphenmins}  =>  {2,3}
>  {latex/frenchspacing}  =>  {false}
>  {latex/indentfirst}  =>  {false}
>  {latex/fontsetup}  =>  {true}
>  {latex/envname}  =>  {latex}
@UlrikeFischer There is quite a lot there, I think it could be shorter
@JosephWright That would most likely mean loading expl3 before ltluatex which I will try to avoid in any way possible because it leads to issue when the Lua allocators are not available in expl3. (Of course, we could separate that from loading luaotfload which is not needed by any other code)
@MarcelKrüger Like @DavidCarlisle says, all we really need to do is make sure the \everyjob entries are re-ordered
@MarcelKrüger yes I agree we don't want to change the loading order after you have just got it moved up, but if we can fix the banner it would be good I think
@PauloCereda My button to open the log file (second from the bottom):
@JosephWright yes polyglossia got quite wordy too.
14:46
@UlrikeFischer Very much so
@UlrikeFischer Things like xcolor add stuff too, etc.: like I said, 'history' (people liked the idea of having this stuff)
@samcarter_is_at_topanswers.xyz ooh <3
@PauloCereda you can probably guess what the top button does
@JosephWright well I sometimes like it too, but it would be nice if one could adapt the debug level here. In tagpdf I added "log-levels" and you can change them.
14:49
@UlrikeFischer Maybe we could make it much shorter by only reporting the version once and afterwards only reporting mismatches. (This doesn't help with the paths though because we want to see full paths there) You are right, we really should suppress the callback messages. Maybe add a (team-only? only in everyjob?) silent version of add_to_callback?
@UlrikeFischer Yes, I mean by default things should be a lot less noisy: @DavidCarlisle is right, 2k lines for a simple non-debug run is over-the-top; of course, part is due to file loading lines, plus register allocation
@MarcelKrüger sound like a plan. We know the callbacks. And there could be a conf-option to get it more verbose again. Regarding the pathes: we could format the messages better so that pathes always start a new line. Then they will break less often.
Weird libertinus / listings interaction?
\documentclass{article}
\usepackage{libertinus}
\usepackage{listings}
\lstset{basicstyle=\ttfamily,columns=flexible,language=TeX}
\begin{document}
\begin{lstlisting}[caption={A simple document}]
\begin{forest}
    [DP [D\\the ] [NP [N\\man ]]]
\end{forest}
\end{lstlisting}
\end{document}
Without libertinus no bold (which is the correct output).
@MarcelKrüger and the fontloader message doesn't need to show the file name twice, only the path is enough.
@JosephWright should we discuss KOMA? ;-)
14:54
@AlanMunn Which engine do you use? I get strange "
@samcarter_is_at_topanswers.xyz LuaTeX
@AlanMunn The bold seems to be from the language style
@samcarter_is_at_topanswers.xyz Right, but it shouldn't be there, and it's somehow connected to loading libertinus.
@AlanMunn does that have bold monospace? (in most families since you have \ttfamily bold wil substitute back to medium
@UlrikeFischer ???
15:00
@JosephWright regarding wording messages ;-)
@AlanMunn Why shouldn't it be there if you use tex highlighting? (with latex highlight it looks better)
@AlanMunn I get bold also without it, only less visible as it is a different font.
@UlrikeFischer Hmm, you're right. Good eye!
@samcarter_is_at_topanswers.xyz Ah, maybe that's my problem. But I don't want the to be highlighted. So I'd prefer no highlighting at all.
@AlanMunn Don't use the language key then :)
@AlanMunn libertinus has no bold font, and so he fakens it: BoldFeatures = {RawFeature={embolden=3}}, and imho it is a bit too much.
15:07
@UlrikeFischer That makes sense. Thanks. This solves it for me. I'll just leave out the language key entirely.
@UlrikeFischer I take it you'll sort out a parser for that version data
@AlanMunn If the is the only problem, you could delete it:
% !TeX TS-program = lualatex

\documentclass{article}
\usepackage{libertinus}
\usepackage{listings}
\lstset{basicstyle=\ttfamily,columns=flexible,language=[LaTeX]{TeX},deletetexcs={the}}
\begin{document}
\begin{lstlisting}[caption={A simple document}]
\begin{forest}
    [DP [D\\the ] [NP [N\\man ]]]
\end{forest}
\end{lstlisting}
\end{document}
@PhelypeOleinik and even the context user complain now ...
@UlrikeFischer Yup
@UlrikeFischer Looking at that right now
@JosephWright probably some "is it latex test" around like I did for the page color code?
15:19
@samcarter_is_at_topanswers.xyz For this particular purpose the keyword hightlighting isn't helpful. And if it gets \\the wrong, then it will probably get other things wrong in a semi random fashion.
@UlrikeFischer I told @JosephWright: that version-checking will be the most hated error of all time :-)
@JosephWright I forgot to define the loaded date in generic mode. I'll fix it in a bit
@PhelypeOleinik Ooops
@JosephWright Drat, I have to leave now. I'll fix it tonight
16:21
@JosephWright I checked, you can't remove the ~ in the luatex code, you end with wrong object references in the pdf: << /blub 50 R >>
@UlrikeFischer Ah, right: we should add a note to the source
16:56
@JosephWright I know remembered what is the problem with extractbb --version and similar: you can't redirect it to a file, you have to use pipes tex.stackexchange.com/q/306987/2388, and on miktex it could be problematic. Well at least for now texlive should be okay. Is there an expl3 equivalent of \openin \mypipe="|extractbb --version"? I tried \ior_open:Nn\g_tmpa_ior{|extractbb~--version} but it complains about a missing file ...
17:51
@yo' -- Depends on when he visits. Remember -- he thinks 20C is cold. (Have fun.)
@barbarabeeton mild cold. :)
18:02
@UlrikeFischer \sys_get_shell:nnN
 
1 hour later…
19:14
> Blatherskite (noun): A person who talks a lot of nonsense.
Hmmmmmmmmm
20:03
@DavidCarlisle -- Thanks. Am researching. Will be away from laptop for a while but back later.
@barbarabeeton thanks, seems a harder question than I expected-)
20:37
@JosephWright expl3 is so nice ;-)
\documentclass{article}

\begin{document}
\ExplSyntaxOn
\sys_get_shell:nnN{extractbb~--version}{}\l_tmpa_str
\exp_args:Nno\regex_extract_once:nnN{\d+}{\l_tmpa_str}\l_tmpa_seq
\int_compare:nNnTF {\seq_item:Nn\l_tmpa_seq{1}}<{20200801}{Old~dvipdmx}{New~dvipdfmx}

\ExplSyntaxOff
\end{document}
@UlrikeFischer although \regex_extract_once:nnN will be 1000 times slower than a space delimited macro:-)
Anonymous
I do not understand why you put the 0 there in

(0,\hbr) and (0,-\hbr)

For LLM and ULM
instead of the version on the left.

If the x-coordinate of LL is \hbr, then why would the x-coordinate of LLM be 0?
@DavidCarlisle but it looks much tidier so. And there is no space in the argument, it starts like this: ThisisextractbbVersion20200315Copyright(C)2008-2020byJin-HwanChoand
Anonymous
@UlrikeFischer well who dropped the spaces:-)
@UlrikeFischer \def\zz#1n#2C#3\relax{#2}
20:46
@DavidCarlisle I just wanted to complain about the numbers ;-)
@DavidCarlisle Tidiness is in the eye of the beholder. :)
@AlanMunn but speed is an objective measure. (as is number of expansions)
@DavidCarlisle One unintended consequence of expl3 will be the loss of ability (of people) to do this sort of thing I suspect.
So programming in TeX will become analogous to programming in assembly language.
@AlanMunn some people think it's that already
@DavidCarlisle True
20:54
@AlanMunn although have worked with people who really work in assembly, and that's a lot weirder
@DavidCarlisle Yes, it's a whole different way of thinking. Although compared to other languages people know, macro languages are also a whole other way of thinking.
@AlanMunn out performing C compilers in matrix multiplications is a specialist art:-)
@DavidCarlisle Yes, when the calculations have to scale up, tiny differences really matter.
@DavidCarlisle better so:
\documentclass{article}

\begin{document}
\ExplSyntaxOn
\sys_get_shell:nnN{extractbb~--version}{}\l_tmpa_str
\str_new:N\g_uf_dvipdfmx_version_str
\cs_new:Npn \uf_extract_version:w #1n#2C#3\q_stop
  {\str_gset:Nn\g_uf_dvipdfmx_version_str{#2}}
\exp_last_unbraced:No \uf_extract_version:w \l_tmpa_str\q_stop

\int_compare:nNnTF {\g_uf_dvipdfmx_version_str}<{20200801}{Old~dvipdmx}{New~dvipdfmx}

\ExplSyntaxOff
\end{document}
@UlrikeFischer I think so although perhaps the cost shelling out to call the program and inputing the resulting string back in is probably the slowest part. Only done once anyway so it's not going to be an observable difference:-)
21:00
@DavidCarlisle yes, the call is the slowest imho. But one could perhaps do it when creating the format, then it wouldn't matter much.
It works in miktex too, it seems to have enabled pipes.
Good evening everybody into chat. Good chat...
@egreg Hi, I hope that you are well. Please, do not angry with me, I am very happy if sometimes look my questions in Math.SE. if it is possible. math.stackexchange.com/users/705338/sebastiano. Always all the best for all.

« first day (3596 days earlier)      last day (1632 days later) »