TeX, LaTeX and Friends - 2020-06-05

« first day (3505 days earlier) ← previous day next day → last day (1420 days later) »

12:07 AM

@DavidCarlisle You are right, but for different packages with many options to conditionally apply, setting a global switch is the best way for me to not go through the whole preamble

1 hour later…

1:24 AM

@Rmano thank you so much for the big help. So the most important thing is the line \path [use as bounding box] (-0.25,-0.45) rectangle (2.25,0.45);. This is the key

However, those numbers are not automatic. If I want to use right of or left of, then that code will not be more useful -- I need to manually calculate what are the top, left, bottom and right margins and add them a plus of space (that's why you say 2.25 instead of 2.0 and 0.45 instead of 0.0)

Do you want me to post a question on the main site asking for this automation?

1 hour later…

2:56 AM

If you don't mind, in geometry package, after setting the paper dimensions, I need to change only one dimension starting from the second page till the end of the document. is it possible without having to write all the other package options inside \newgeometry?

For this minimal example, \usepackage[top = 10pt, bottom = 20pt]{geometry} is used for the first page. For the second page and the rest, I just need to edit the top to be top=5pt. Is there a way to make it without having to write the bottom option as well in \newgeometry{top = 5pt, bottom = 20pt}?

4 hours later…

6:58 AM

@Diaa I'd say 99% of the time that people think they need \newgeometry they don't. if you want the first page to start 5pt higher then you don't need to reset the margins, just start the document with \vspace*{-5pt}

7:19 AM

@DavidCarlisle yes!

@manooooh the idea is that you do a run without the use as bounding box and with the \showboundingbox command. When you have finished, just look at the bounding box number in every slide and put the biggest ones in the use as bounding box path.

Anyway, a quantic cat have a kind of automatic solution here: tex.stackexchange.com/a/542681/38080 @manooooh

4 hours later…

11:32 AM

@JosephWright tikz will pose its own color problems. In this here only the second correctly set the spot color:

\documentclass{article}
\usepackage{colorspace}
\usepackage{tikz}
\definespotcolor{mygreen}{Green 355 C}{1,0,1,0}
\begin{document}\pagecolor{red!50} %to make it more visible
\begin{tikzpicture}
\fill[color=mygreen](2,0)rectangle (3,1); %fails
\fill[mygreen](3,0)rectangle (4,1);
\end{tikzpicture}

\begin{tikzpicture}
  \fill[color=blue](0,0)rectangle (1,1);
  \fill[blue](2,0)rectangle (3,1);
\end{tikzpicture}
\end{document}

@UlrikeFischer Oh, I know: we'll need to agree changes in TikZ and other places to get everything to work togethr

@JosephWright ooh agree in disagreeing. :)

@UlrikeFischer First we get an interface we are happy with, then we add hooks, then we deal with third-party issues

@UlrikeFischer So it's some set-up/ordering thing: works if you swap the two green lines

@JosephWright yes, good plan. I looked a bit at possible issues - they are a number of incompabilities. And colorspace for example fails if xcolor is loaded with cmyk. So sorting this is quite needed.

@UlrikeFischer Sure, but my plan is basically 'get it sorted in l3color and tell people to use that'

11:37 AM

@JosephWright yes then some missing initialization is done. But it shouldn't be needed ;-(

@JosephWright ;-)

@UlrikeFischer You were thinking of a different plan?

@UlrikeFischer The l3build changes working for you?

@JosephWright this you change something today? I hadn't yet the time to check.

@JosephWright did we change the e-type in xparse?

@UlrikeFischer I meant last night's changes

@UlrikeFischer Nope; must be a change in the document, some incorrect catcode set up

12:24 PM

@DavidCarlisle I see. So, would you please suggest how to apply \vspace*{-4cm} at the beginning of every page starting from the second page till the rest of the document?

@Diaa do it the other way round: add \vspace*{4cm} only on the first page.

@UlrikeFischer I have a header on the first page taller than the running headers afterwards

So I fixed the headheight to accommodate the taller header, then I need at every page from the second page to be pushed up

@Diaa that doesn't matter, you can hide its depth and let is stick out into the page. Setup the geometry so that it is correct from page 2 on. Btw: That is how e.g. letter classes handle the large address block on the first page.

@UlrikeFischer Good point. Final question please, is there a way to make the header text be top-aligned with its allocated space?

In other words, can the header text be printed from top to bottom? since in exam, the header is written from bottom to top, which makes the text go over the top border. However, I need it to go over the bottom border instead

12:41 PM

@Diaa sure. You can place anything anywhere.

\documentclass{article}
\usepackage{fancyhdr}
\pagestyle{fancy}

\lhead{\raisebox{0pt}[0pt][0pt]{\begin{tabular}[t]{l}a\\b\\c\\d\\e\\f\\g\\h\\i\\j\end{tabular}}}
\begin{document}

some text blub
\end{document}

12:52 PM

I am sorry, but I think I can't understand how to apply what you have suggested to the following

\documentclass{exam}
\usepackage{graphicx}
\usepackage[xparse,breakable,raster,skins]{tcolorbox}
\tcbuselibrary{fitting}

\def\headerContents{%
\tcboxfit[height=4cm]{heading}
}

\pagestyle{headandfoot}
\firstpageheader{}{\headerContents}{}

\usepackage[
includehead,
top = 2mm,
headheight = 4cm,
headsep = 2mm,
showframe
]{geometry}

\begin{document}
text
\end{document}

I need the header to overflow the bottom border not the top one

user image

1:23 PM

@Diaa \tcboxfit[height=4cm,box align=top,enlarge bottom by=-4cm]{heading}

1 hour later…

2:35 PM

Question:
I heard a phrase: "Never stop running!"

Does that mean: "Never give up!"

@cis not really a phrase I'd recognise as an idiom. Google suggests a Nike advert where that phrase is used (quite literal meaning) and this book amazon.co.uk/dp/B07MTTRJQK/…

2:50 PM

@DavidCarlisle Ah ok.

2 hours later…

Pablo González L

5:10 PM

@Skillmonlikestopanswers.xyz kpfonts-otf :)

@DavidCarlisle@UlrikeFischer pdfcrop update?

@PabloGonzálezL did you test it??

Pablo González L

Yes:)

@UlrikeFischer There is a part of the code that acts with xetex, I am not sure what it does...it is named in the README, but, I don't know if it is necessary ...

@PauloCereda Did you take a look at my dilemma with "arararc.yaml"?

@PabloGonzálezL you shouldn't have mentioned it. With xetex the pdf version string is broken. I will have to look at it.

Pablo González L

@UlrikeFischer I blame myself :(

6:19 PM

@PabloGonzálezL well the good news was, it was not my fault ...

1 hour later…

7:31 PM

Hello all!

I am drawing some graph and I am having some arrow collisions:

\begin{center}
\begin{tikzpicture}[>=stealth',thick]
\node at (0,0) (1) {$1$};
\node at (3,0) (2) {$2$};
\node at (1.5,3) (3) {$3$};

\draw (1) edge[bend left,->] (2);
\draw (2) edge[bend left,->] (1);
\draw (1) edge[bend left,->] (3);
\draw (3) edge[bend left,->] (1);
\draw (2) edge[bend left,->] (3);
\draw (3) edge[bend left,->] (2);
\draw (1) edge[loop above] (1);
\draw (2) edge[loop above] (2);
\draw (3) edge[loop above] (3);
\end{tikzpicture}
\end{center}

user image

What do you recommend me to avoid collisions? Thank you!!

Perhaps a proper inner sep?

@manooooh Change the bend angle (e.g. bend left=5), or the in/out-angles ((1) edge[out=50,in=260] (3)) for example.

@manooooh Or use specific anchors ((3.260) edge[bend left,->] (1.50))

@TorbjørnT. that is a good idea

@TorbjørnT. now I am trying to draw a background with a circle, that includes this drawing

@manooooh \draw (1.5,1.2) circle[radius=3cm];?

7:50 PM

@TorbjørnT. yes, thank you!!

8:00 PM

@JosephWright I'm try to create the pdfcrop upload. l3build insist to include 4 pdf files which are in the github for tests. I can't find a way to exclude them.

Konbanwa!

8:28 PM

@Rmano ohhh that was excellent. I use \showboundingbox when I finished the drawing, then look what numbers show, copy them, comment out that line and finally add the \path with these numbers. Excellent, it is easy to implement. Thank you so much!!

@DavidCarlisle I prepared pdfcrop for ctan. Could you take a quite look if the new readme.md is okay?

9:15 PM

@UlrikeFischer ok let me pull here....

@UlrikeFischer we could add this bit from the perl file, so the readme is a bit less minimal

PDFCROP takes a PDF file as input, calculates the BoundingBox
for each page by the help of ghostscript and generates a output
PDF file with removed margins.

@DavidCarlisle added

@UlrikeFischer thanks looks good to go:-)

ghostscript is a pain ;-) I just got a pdfcrop failure because in the git bash it looks for gs and found an old gs version in the lilypond folder and this then complained because if found newer resources somewhere else (probably texlive folder).

@DavidCarlisle let's look if ctan accept ...

10:09 PM

@DavidCarlisle -- "an output file".

@manooooh you're welcome, happy to be useful!

Pablo González L

10:23 PM

@UlrikeFischer That's what happens when you use the git bash console on windows (@blame cygwin/mingw)

@PabloGonzálezL it is not only the git bash - I had a similar problem a fews days ago in cmd because gswin64c had a different version than the one in texlive and had to install a new version there too. I don't know why ghostscript looks for resources everywhere.

Pablo González L

@UlrikeFischer That's because it tries to search the variable defined in PATH and then in the dynamic library by means of the windows registry...

10:39 PM

@PabloGonzálezL I don't mean pdfcrop, but ghostscript itself . Every few month it fails somewhere because the gsinit file (or how ever it is called) has "version 9.27 instead of 9.50"

Pablo González L

@UlrikeFischer For the same reason, when you search PATH, it tries the full absolute path, every vex you update changes the numbering and the path is lost, unless you use mklink

@PabloGonzálezL but this is simply madness, it should look for resources first on a path relative to the binary.

Pablo González L

@UlrikeFischer It's been like that since win95 :), I assume the gs maintainers aren't big fans of windows

It is assumed that if you are a windows user you will try to use gs*.dll and not the executable itself (no idea why this assumption)

11:05 PM

@barbarabeeton blame @UlrikeFischer :-)

@DavidCarlisle -- How uncharacteristic. (But do fix it.)

@barbarabeeton I corrected it, but the file was already on the way to ctan, so it will have to wait for the next bug report.

@UlrikeFischer just realised the ctan zip has README and README.md I'm not sure ctan will like that we should probably have renamed README to README.md, sorry

@DavidCarlisle the readme is the documentation.

@UlrikeFischer yes I know but the .md extension is dropped when you view the rendered markdown so they'll both end up as README I think

@UlrikeFischer If Petra complains, blame @barbarabeeton

11:17 PM

@DavidCarlisle hm. let's wait what happens. But if we should rename it the README should probably get a name like pdfcrop-doc.txt or so.

@UlrikeFischer yes, probably.

@UlrikeFischer 'teh' :-)

@DavidCarlisle ;-)

@UlrikeFischer I just fixed "a output" in README as well

If I have to reupload I can perhaps try to improve the system detection. In the git bash pdfcrop believes to be on unix and doesn't find the windows ghostscripts.

@UlrikeFischer in a cygwin setup that's what you want, as you want to use the cygwin gs as the windows one doesn't understand the file structure

11:27 PM

@DavidCarlisle hm. And how could one distinguish this? It detects miktex if an environment variable is set, so I could check for a user set variable.

@UlrikeFischer is it using its own perl or a windows perl?

@DavidCarlisle hm. it has no own perl, but it probably will use the texlive one doesn't it?

@UlrikeFischer im confused. The first line of pdfcrop is #!/usr/bin/env perl what does which perl report in git bash?

@UlrikeFischer but its late here and even later there you are allowed to sleep

@DavidCarlisle Yes, will look tomorrow, good night!

night

« first day (3505 days earlier) ← previous day next day → last day (1420 days later) »

Transcript for

Jun '205

$TeX - LaTeX$ TeX, LaTeX and Friends

\begin{chat}

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

$TeX - LaTeX$

site design / logo © 2024 Stack Exchange Inc; legal

mobile