@DavidCarlisle oh, that was so simple! And it makes sense

@JosephWright uhm... in their documentation on p.50 they say "Complement of rgb color" where the new colors are calculated as "E-(original color)", where "E=(1,1,1)". I do not understand why they pick a particular color "E" instead of a general one. Is the same formula that share @DavidCarlisle?

@AlexG wow, you have explained very well, thanks!! However, I do not understand how \frame works. I am using \frame{<content>} but it says "LaTeX Error: Something's wrong--perhaps a missing \item" (where do I have to put an \item???). However, if I implement your solution using \frame{\parbox{\dimexpr\paperwidth-2\fboxrule-2\fboxsep\relax}{<content>}} it works like a charm!, so I did not know if I have done it correctly or not (but it works! haha)

Hello. is it true that \include{} produces a pagebreak at the end? I mean, using \include{}\end{document} produces a new page with the command I defined in \AtEndDocument{}.

It doesn't happen with \input{}.

@Sigur yes, see here: tex.stackexchange.com/a/250/101651

Someone who asked and answered in the answer section, is it possible to start a bounty and give it to him? It would be a not-allowed-autogift, because so anyone offers 500 rep and attracts the attention of the public without anyone helping

Well, at least in math.SE I could see the "start a bounty" button in a question of mine and offer some rep (idk if I can receive that) lol...

@manooooh You cannot give a bounty to yourself

@CarLaTeX uf, thanks to the administration and thanks to you!

@manooooh You're welcome!

Is it possible to do the following big table and extol it in LaTeX?: civil.frba.utn.edu.ar/2011/archivos/Plan%2095-A.pdf

@DavidCarlisle how can we do it automatically? I have \definecolor{original}{RGB}{x,y,z}, but I cannot do \definecolor{new}{RGB}{255-x,255-y,255-z} or \colorlet{new}{255-original.x,255-original.y,255-original.z}

I also cannot do \color{-original} because I cannot put that piece of code in a fill parameter of tikzpicture (it throws me errors)

I think we need to extract the label of \color, but I do not know how. I also tried with \definecolor{new}{RGB}{\color{-original}} but it does not work

@Raaja Can you please explain to me what Dave commented in my answer? I am feeling quite curious.

@JouleV ^^^^ maybe? ¯\_(ツ)_/¯

Hi @manooooh!

Hello!!

@manooooh Well I don't think my answer has something to do with that apt :)

@JouleV hahaha, I am not a smart duck, @marmot is a smarmot, so I guess he can answer that question, too :)

@manooooh But our smarmot won't be online in a couple of hours.

Yup you are right, he is hibernating right now u.u

@manooooh In supersymmetry, the superpartners get an s. The superpartner of an electron is a selectron, and the superpartner of a top is a stop. Since supersymmetry is a symmetry of nature, also sducks must exist.

@marmot mind blown.

@marmot Oopsie doopsie

So you haven't hibernated yet :)

@marmot OH MY...!!

:D :D

@marmot There are five users but only four ducks.

@JouleV a chat where there are only animals is strange (?)

@HenriMenke Thank you for accepting my feature request about inner sep. I'm very sad you refused the one for duck footprint, but I know you're not mean :)

@CarLaTeX I'm happy to accept a pull request for such features but I can't invest time into this myself.

@HenriMenke I know. You're already doing a big work!

@JouleV In fact, I poked Dave to let him know that your approach to answer the particular question is better than mine. Hence, I proposed to accept yours as the answer :D

@JouleV And, he did not comment ;D

@Raaja The two answers are equal imho. Each follows a different approach, I don't even use pgfplots.

@manooooh yes it's same as 1,1,1 is white which is 255,255,255 in RGB if you are not using xcolor you could simply do the subtraction I suggested directly eg from [RGB]{100,20,3} do [RGB]{\the\numexpr 255-100\relax,\the\numexpr 255-20\relax,\the\numexpr 255-3\relax}

@Sigur before and after (\clearpage)

@Sigur \input is a very different beast to \include

@Skillmon plain tex? \csname newif\endcsname ...

perspective :D

@JouleV

@manooooh \frame draws a tight frame around its content, without increasing the overall size of the resulting box. Maybe it is less robust than \fbox. Thus, go with the latter if it works for you.

What's the matter with non edited threads (having several upvoted answer) being bumped to the home page?

Other two just bumped up: tex.stackexchange.com/questions/283031 and tex.stackexchange.com/questions/383476

@JouleV I did not know how to link a question with closure? should I do with comments for closure as off-topic (which, I really dont want to do so) :D

*There should not be quesiton mark in my first sentence (typo).

tex.stackexchange.com/questions/474476/… has there been any progress on the modiagram-package? I'm still getting errors, even with v0.2g

@Raaja I think the OP's question has anything to do with Drawing on an image with TikZ. However, you should do what you have been doing:

> I'm voting to close this question as off-topic because this is a just-do-it-for-me question. And I downvoted the question for that reason.

@MaxR Don't think there has been a new release

@egreg I fear it's me making some old anwers community wiki. For some reason they get bumped to the home page but vanish after refreshing it a few seconds later.

Ok noted @JouleV

@HenriMenke Concerning TikZ feature requests, I guess you can close github.com/pgf-tikz/pgf/issues/607 (LaTeX does issue the warning).

@campa Why should tex.stackexchange.com/a/263275 be CW?

@DavidCarlisle of course, still caught me off guard when I tried whether some code of mine works in plain.

@Skillmon as documented here tex.stackexchange.com/questions/450509/…

@HenriMenke Just noticed that the pgfmanual gobbles a control sequence I put in | entirely :(

@egreg Because it's basically copy/paste from the documentation; no real work from my side. (Unless I'm misunderstanding the sense of CW...)

@marmot sorry if you got the wrong impression. I think the pgfmanual is really great. And of course it is much more exhaustive and descriptive than the xparse manual. But xparse is only about a better \newcommand and though it has lots of code lines, its scope is very narrow. I think 13 pages of documentation for \newcommand with extras (sorry if that offends anyone in the L3-team, xparse is great) is not too small.

And the expl3 macros are explained on 300 pages of interface3.pdf. Some of those are used by xparse but aren't part of it.

@campa You did the work of reading the manual and extracting the relevant information. ;-)

@campa if that was the sense of CW there would be loads of CW. Approximately 50% of questions can be answered by reading the manual.

@egreg thanks :)

@Skillmon and the rest is undocumented features. :)

@GermanShepherd Here's an abstract version where you don't need to guess the widest entry

\documentclass{article}
\usepackage{enumitem}
\usepackage{xparse}
\usepackage{showframe}

\ExplSyntaxOn
\NewDocumentEnvironment{tabdescription}{O{}b}
 {
  \__germanshepherd_tabdesc_measure:n { #2 }
  \begin{description}[
    #1,
    labelwidth=\l__germanshepherd_tabdesc_dim,
    leftmargin=\dim_eval:n { \l__germanshepherd_tabdesc_dim + \labelsep }
  ]
  #2
  \end{description}
 }{}

\dim_new:N \l__germanshepherd_tabdesc_dim
\seq_new:N \l__germanshepherd_tabdesc_seq
\box_new:N \l__germanshepherd_tabdesc_box

If you don't use TL 2019 or MiKTeX, there is a variant to apply.

\documentclass{article}
\usepackage{enumitem}
\usepackage{xparse}
\usepackage{environ}
\usepackage{showframe}

\ExplSyntaxOn
\NewEnviron{tabdescription}[1][]
 {
  \__germanshepherd_tabdesc_measure:V \BODY
  \begin{description}[
    #1,
    labelwidth=\l__germanshepherd_tabdesc_dim,
    leftmargin=\dim_eval:n { \l__germanshepherd_tabdesc_dim + \labelsep }
  ]
  \BODY
  \end{description}
 }{}

\dim_new:N \l__germanshepherd_tabdesc_dim
\seq_new:N \l__germanshepherd_tabdesc_seq
\box_new:N \l__germanshepherd_tabdesc_box

@egreg Great! I do use MikTex.

@egreg "This causes the whole document to be in italics from that point on." at least it's not all in math mode, so you get some spaces:-)

@GermanShepherd The first code requires an up-to-date distribution

@DavidCarlisle :-D

@egreg I have a query on tikz grids. Care to take a look?

Im not sure what tag it should go under

@GermanShepherd I guess that marmots are better in this kind of thing.

Ohh, a tikz question that @marmot have not yet answered. And it is more than 1 minute old!! :S

@mickep marmot doesn't seem to be online. Any directions/ideas are welcome.

@Szabolcs normally I simply use where pdflatex.exe to find binaries.

@HenriMenke I think I got the documentation done. Is there anything more that I should edit/add except for the code and a pgfmanual-en-module-parserx.tex documentation? Editing any index or the like?

@UlrikeFischer "where" shouldn't be available on Windows, unless you are using Cygwin.

@AlexG ^^^ . And I'm not using cygwin.

@UlrikeFischer Ah, thank you! Learned something new!

@AlexG You are thinking of which

@JosephWright Yes, you are right. I was a bit confused.

@JosephWright I mixed it up with whereis

@JosephWright one of the cases where MS had to divert from *nix simply to divert :-)

isn't tex.stackexchange.com/questions/483172/higher-contrast-needed on-topic and a duplicate of tex.stackexchange.com/questions/198721/…?

not sure if it would be worth it to reopen and immediately close as duplicate, but still

@Marijn Nothing in the question indicates that latex is involved at all. In contrast the older question was about how to manipulate images while including them into latex

@Marijn It is a question about viewers I think. Nothing to do with LaTeX and friends at all.

yes but here the answer could be "make a standalone document with a single \includegraphics and manipulate the decodearray"

and edit the question by putting "Using LaTeX, how do I" at the start, the OP would be happy to do so if that would solve his problem I guess

@Marijn as it is the question is not clear and has no use for other people. It doesn't make sense to reopen it. If you think it would help the user you can link to the older question in a comment.

@UlrikeFischer that's a good suggestion, I was just wondering :)

@DavidCarlisle sorry I do not understand. I can use xcolor, if that is better

@manooooh xcolor provides a general syntax for combining colours so you can use that to get the complement but with basic color package you just need to subtract from 255 so it's easy enough. for most purposes xcolor is used anyway (eg if you have tikz you have xcolor)

@Marijn no, the OP said it was a scan so no tex is involved at all.

@DavidCarlisle yes but tex could be involved as a solution to the problem

Guys, let's never forget:

@Marijn yes I saw your later comments, I suppose so. I doubt it's the answer the OP wants though.

@PauloCereda excellent package, why post it now?

@DavidCarlisle I was remembering the good times. :)

@PauloCereda when you had a thesis to write?

@DavidCarlisle oh no

Hm so apparently when you write a package you get a lot of upvotes

note to self: learn how to write packages :)

@Marijn you also need to insert a lot of undocumented features into the package so that people ask about it.

@DavidCarlisle I found today a nice one:

\documentclass{article}
\usepackage{wrapfig}
\usepackage{lipsum}
\begin{document}

\begin{wrapfigure}{o}{2cm}
\rule{2cm}{2cm}
\end{wrapfigure}
\lipsum[43]

\end{document}

lipsum...

@DavidCarlisle the interesting feature was that with \lipsum[43-44] it works fine ...

@UlrikeFischer \leavevmode\lipsum[43]

@DavidCarlisle that's one work around. But I made a bug report - it shouldn't use a group without a \par.

@UlrikeFischer yes I didn't look what it was doing I just guessed based on the output:-)

@DavidCarlisle good guess ;-)

@DavidCarlisle, @UlrikeFischer Who needs wrapfig ;)

\documentclass{article}
\usepackage{l3galley}

\usepackage{lipsum}

\begin{document}

\setbox0=\hbox{\rule{2cm}{2cm}}
\ExplSyntaxOn
\galley_cutout_right:nn { 0 } { 2.1cm , 2.1cm, 2.1cm ,  2.1cm , 2.1cm }
\leavevmode
\hbox_overlap_right:n{\vbox{\kern\dimexpr2cm-\baselineskip\hbox_to_wd:nn{\hsize}{\hfil\box0}
\kern\dimexpr\baselineskip-2cm}}
\ExplSyntaxOff
\lipsum[43]


\end{document}

(couldn't be bothered to do the wrapped figure properly ...)

Or without \leavevmode

\documentclass{article}
\usepackage{l3galley}

\usepackage{lipsum}

\begin{document}

\setbox0=\hbox{\rule{2cm}{2cm}}
\ExplSyntaxOn
\galley_cutout_right:nn { 0 } { 2.1cm , 2.1cm, 2.1cm ,  2.1cm , 2.1cm }
\hbox_overlap_right:n{\vbox{\kern\dimexpr2cm\hbox_to_wd:nn{\hsize}{\hfil\box0}
\kern\dimexpr-2cm}}
\ExplSyntaxOff
\lipsum[43]

\lipsum[1]

\end{document}

@JosephWright mixed in tex primitives ;-).

@UlrikeFischer Yeah, I know ... like I said, couldn't be bothered to get it all right (should be a coffin)

@UlrikeFischer You want l3wrapfig I take it

@JosephWright well I guess the main problem is not the cutout, but the floating / left / right / automatic calculation etc.

@JosephWright but if you are looking for a challenge:

@UlrikeFischer Well yes, you have to make a judgement on the link between 'lines' and space (we have to work on lines)

\CatchFileDef {\l_tmpa_tl} {file.csv} {\shorthandoff{"}\ExplSyntaxOff \obeylines}

@UlrikeFischer Also the fact l3galley kills everything

@UlrikeFischer Urgh

@UlrikeFischer \file_get:nnN?

@JosephWright it works ;-)

@UlrikeFischer Not doubting it: catchfile is very good

@UlrikeFischer I should finish the catcode table stuff

@JosephWright the main problem is to get \ExplSyntaxOff when reading the file.

@UlrikeFischer Like I said, I think we really should look at catcode tables or similar: the question for this case is how they should 'look' (as there is a start but no end other than by grouping)

@JosephWright yes. This \CatchFileDef stuff is quite useful: I managed to import csv-files with newlines in the fields with it (and preserving the newlines ...).

TUG members enjoy the 'all members' mail?

@JosephWright ? I got no mail yet.

@UlrikeFischer Really?

@JosephWright yes ;-)

@UlrikeFischer members@tug.org?

@JosephWright normally I get mails addressed to all-tug-members@tug.org. Sometimes to my name. But I didn't get one today.

@UlrikeFischer Hmm

@JosephWright My husband didn't get a mail either. What happened?

@UlrikeFischer I'll see if there is an archive

@UlrikeFischer -- These are two different lists. "all-tug-members" is restricted to official announcements sent from the board. "members@tug.org" is a list that must be subscribed to. You can sign up (or read the archives) from the TUG website.

@UlrikeFischer tug.org/mailman/private/members/2019q2/000128.html

@barbarabeeton Ah, right

You two can start the pop band "The deleters".

@JosephWright I got no mail

@DavidCarlisle See the link to the archive I've posted above

@JosephWright authorization failed:-)

@DavidCarlisle I can copy and forward to you if you want.

@UlrikeFischer I'm going to try the password reset ...

@UlrikeFischer please :-) (no password reset received...)

@DavidCarlisle done ...

@UlrikeFischer might have guessed:-)

@DavidCarlisle I think it was obvious

Now I have to work out why I got no mail, @barbarabeeton am I listed as a member?

@DavidCarlisle -- I don't actually have access to those records. I'll ask Robin to send you the information.

@DavidCarlisle did you subscribe the list? I had to do it a few minutes ago to get into the archive.

@DavidCarlisle, @barbarabeeton, @UlrikeFischer I think it's the usual situation: some people do not want to accept that 'user' groups are today for developers

@UlrikeFischer oh there are instructions? No of course not:-) (I must have been subscribed at one point firefox thinks it has a password saved but it doesn't get in)

@DavidCarlisle, @UlrikeFischer, @barbarabeeton The reality is that a lot of meetings, etc., are going to end up as the LaTeX or ConTeXt devs + friends, and that is how it is

@samcarter all that work we did was for nothing. I forgot to check the design, and it had changed in 2017! So much effort. On the bright side: They could provide the design specs, and it is a LOT easier to implement. Got the base setup running, in 4 hours. Including using your titlepage page style trick, which is much more applicable to this newer design.

Now we also know that many fonts does not scale well. I have a base font size of say 20/24. I think I'm going to divide the entire design by 2, that seems better with beamer (I implemented it one to one).

@JosephWright at the Dante meetings that is not so clear. Naturally most of the around 50 people were more than "normal users" but real developers are not the majority.

@UlrikeFischer Sure, but DANTE is the exception. There's a critical mass issue. Do you get new users?

@UlrikeFischer The question remains though how one would encourage new users to join: there has to be something in it for them

@daleif Sorry to hear! You can just see the bright site: you are allowed to do one more beamer theme - yeah!

@JosephWright they get new users but the total number is slowly going down, as everywhere.

@JosephWright yes that's a discussion we have every year too.

@daleif 20/24 seems very big - does it come from a powerpoint prototype that uses a much larger paper size?

@UlrikeFischer Exacty

We submitted PGF 3.1.2 to CTAN on 04/04/19. I hope it will be picked up for TeX Live 2019.

@samcarter Does that mean there will be a theme tikzlings?

@JosephWright but we had a nice view through large windows from the 8th floor over Darmstadt ;-)

@marmot If I remember the videos from last year's TUG correctly, there must already be some tikzducks theme out there, which Will used for his talk :)

@UlrikeFischer Oooh

@samcarter Yes, but only ducks is not very diverse.

@samcarter you could use the IFT-logo for such a theme.

@marmot Ducks are the personification of diverse - why would nearly all job offers search for m(ale)/f(emale)/d(ucks)?

@UlrikeFischer the crowd of tikzlings in front of the letters could carry the logo across the footline as some kind of progress bar indicator :)

@samcarter now we know what to mess, with so much easier

@samcarter Not for important jobs, though, like e.g. marmots.org/a-new-job-posting-marmot-keeper

@samcarter exactly. It is almost 34cm wide.

@marmot do you think the following is ok as an example in the pgfmanual?

Speaking of ducks, was at our natural history museum earlier. On display was one guys life time worth of collecting bird sculls (from already dead birds). Quite a number of duck sculls

@marmot mmm, maybe they excluded ducks because of the Ability to lift 22kg requirement?

@samcarter I think the task is way too important to assign it to a duck. ;-)

@daleif I can really recommend scaling down and using a natural font size like 11pt. In one of my first beamer themes I made the mistake to use the paper size from powerpoint and even though I used a font that scaled well, it caused all kinds of problems at places nobody thought of.

@marmot I'm sure a duck would make a great job in keeping the marmots happy!

@Skillmon There are (unsurprisingly) many things that I do not understand in this code, starting from the first line. (Side remark: some folks prefer \pgfkeys{blabla/.initial=7} and the to use \pgfkeysvalueof{blabla} rather than storing the stuff in an additional macro, whose name may potentially coincide with something that the user already has.) But with your skills you may write a really good answer to tex.stackexchange.com/q/482518/121799.

@samcarter that is what I'm going to do with it. Just wanted to have the basics running.

The mistaken one I did was designed by applying a scale factor to all the measurements I'd done with the PP default, then scaled to default beamer 16:9. It was just a lot faster here to go with the "real" sizes from the specification.

@samcarter Marmot prefer bees, who can produce honey liquor. ;-)

Is it palindrome this?

@Sebastiano The comma is not in the middle.

@marmot only mean people would complain about that

@DavidCarlisle Yes, you are in good company.

@marmot Is it only a binary number? :-)

@DavidCarlisle One cannot trust these ducks, can one?

@marmot trust no, eat yes.

@Sebastiano hexadecimal would be better.

@DavidCarlisle Neither nor.

@DavidCarlisle I am not able to read their syntax. I tried to write it on LaTeX and failed. Also, I do not want to put the numbers manually but using a formula i.e. if I have a \definecolor{original}{RGB}{45,2,7} then how can I substract 255 to each component not manually?

@manooooh as I wrote above, replace 45 by \the\numexpr255-45\relax and similarly for 2 and 7

@manooooh the complement in xcolor is simply -color:

\documentclass{report}

\usepackage{xcolor}

\definecolor{original}{RGB}{45,2,7}
\colorlet{complement}{-original}

\begin{document}
\textcolor{original}{XXX}

\textcolor{complement}{YYY}
\end{document}

@UlrikeFischer I was just coming to that bit (although I think it was in chat earlier:-)

@UlrikeFischer thanks!! I tried with the minus sign but I did not think about colorlet. Wonderful, and thanks to @DavidCarlisle too

@marmot I'm not sure whether I should put that example into the manual (and I don't think I'll have the last word on that), it might be way over the top and require way too many explanations just to show off stuff that are possible with my module. The question you linked to already has two answer, do you think they lack something? I have my problems understanding the question, to be honest.

@DavidCarlisle sorry, did I destroy the didactic structure of the lesson? ;-)

@UlrikeFischer no I was just seeing if xcolor had a manual, to check the exact syntax and you saved me posting the result:-)

@Skillmon I would avoid examples involving \def, \edef, \expandafter. (\tl_new:N, \tl_set:NV would naturally be ok).

@Skillmon I was also going to comment that the example uses a lot of non-latex constructs

@DavidCarlisle TiKZ is format-neutral

@DavidCarlisle yes it does. My bad. I don't think I'll include it. Besides the problems of plain vs. LaTeX (though the module can be used in both) it is simply too large and incomprehensible for the purpose of showing the possibilities of it.

@UlrikeFischer I think if I include expl3 code in the pgfmanual there will be exploding heads.

@JosephWright yes I know, but the examples ought to say what format they are using

@Skillmon I think the question is if we can teach TikZ vector operations. In 2D, you can do with tikzmath nice things like adding and subtracting vectors/coordinates, see e.g. the example on p. 695 of the manual. These operations work superficially in 3D, but you cannot extract the 3D coordinates of the result. Nor can you compute the scalar product in any dimension, nor the vector product in 3D, nor the result of a matrix multiplication (even though you can find the beautiful identity...

.... matrix on p. 999 of the manual.) Allowing one to parse vector operations would be a huge improvement over what currently can be done.

@marmot I don't think I want to implement that without the aid of L3. Which operations would you like to be there? Scalar and vector products, and matrix products. Anything more fancy than that? Keep in mind that TeX isn't build for such operations and the result could be really slow code.

@marmot and while we're at it, a simple 3x3 linear equation system solver?

@JosephWright But the example uses \begin{tikzpicture}. Such a mix is confusing.

@UlrikeFischer that's because the pgfmanual is build in LaTeX.

any beamer people around? what changed here?

@DavidCarlisle blame @JosephWright

Or the L3 people

Oh wait

@PauloCereda That is standard LaTeX team policy.

@DavidCarlisle ooh

I've created two smaller examples, I hope those are better suited:

@Skillmon hello! Sorry for the silly question, but how did you get the actual format style of the TikZ-PGF manual?

@manooooh by actually working on it :) It is on github and I currently write some documentation for a small module nobody will use I wrote. Once I'm satisfied with the results, I'll make a pull request.

hello guys

@manooooh link: github.com/pgf-tikz/pgf

@JosephWright TikZ is format-neutral at first glance but in reality it really isn't. github.com/pgf-tikz/pgf/issues/354#issuecomment-478804848

I'm using newtxtext and newtxmath, but when I see the pdf fonts it shows also CMR7 and CMR10. I'm afraid I am mixing some fonts by using some package wrongly.

@Skillmon yes yes I read the Henri Menk comment, that is awesome! I love people who share open code (I would not have imagined it from the TikZ-PGF package)

@manooooh it was just moved to github, the license didn't change, it was open the whole time, as far as I know.

@Skillmon oh, but to see the source code you needed to install the package and open the .sty, I guess. With GithHub it is online

@manooooh it was online before how else would you have been able to install it.

@manooooh Before GitHub it was on SourceForge: sourceforge.net/p/pgf/git/ci/master/tree

@manooooh it has always been online

@Sigur example code?

@DavidCarlisle, sorry, I'm not able to find one. I'm commenting out some inputs to find the file which produces that. I asked because maybe some \text or \mathrm could produce that.

@DavidCarlisle, I hope to find some more info soon since I've just discovered the file.

I did not even know it, I hope TikZ-PGF goes even further!

@DavidCarlisle yes, the default frametitle was changed "recently" - maybe related: github.com/josephwright/beamer/issues/535

@samcarter thanks

@DavidCarlisle, I found. Nice! It comes from a pdf figure inserted.

@Sigur ah;-)

@DavidCarlisle, now I have to check the figure code to discover, since if I remember, I used newtxmath to draw it with tikz.

@Sigur you should have stuck to picture mode

@DavidCarlisle nah, ASCII-art!

@DavidCarlisle, I'm not sure I could draw that

@Sigur should be easy in ASCII-art.

@Skillmon, it would be an adventure to try it.

                                           e
                                             3
                                           ·
   c                                       v c
     1,2                                      1,1

Oops, what a mess!

@DavidCarlisle, I found and fixed. Since my figure contains only math text (I mean, all texts are in $ $) I only loaded newtxmath and I forgot to load newtxtext.

Maybe the numbers come from text font.

@manooooh, did you copy and paste? lol

@Sigur hahaha where is the graphic available? I just made a sketch by hand

@manooooh, you are right, I sent only the png image. Sorry, it is not available.

@Sigur :) btw nice graph! I do not understand the differences between the dashed and dotted lines, why they are different?

@marmot How many bees will it take to fulfil the "ability to lift 22 kg" requirement?

(the marmot website might benefit from having a look into siunitx :)

@manooooh, the dotted are the great circles. the dashed is to highlight the edges of the solid region, but it must not be there, mathematically.