Should wrapfig work on PNG images?

I should have asked the problem that I am facing on this problem very earlier

Also does it work on bxjscls, lualatex, and graphicx?

I really have no ideas what I am failing

I was supposed to submit the document by today

However I try to make a document of just one page, the figure goes to second page: WTF

@nrgmsbki4spot1 There's no restriction on content: most likely you have a size issue

@nrgmsbki4spot1 yes, if correctly used, it works. Please take a moment reading your intervention though: do you think you gave sufficient information to help us solve the problem? Do you think that the tone of it is the best one to have people rushing to help? I suggest you prepare a MWE for the main site, somebody I'll surely look at it.

@samcarter Looks like the beamer release is going OK

@samcarter Did you enjoy the emails about colour?

@JosephWright more than 2 days without anybody screaming about broken presentations -- that's surprising! :) Thanks for pushing out the new release!

@JosephWright Very interesting to read! My head just oscillates between the different approaches, the more one thinks about the colours, the more complicate they get ... I've been thinking about the problem of the block body having to know about the title colour. Maybe one could delegate setting the colours to the block environment instead of the boxes? Maybe one could then get away with local colours (but making them available as soon as they are defined)?

Anyone know how to tell Evince (Ubuntu) what font to use when trying to render a PDF using helvetica without helvetica being embedded into the PDF? For now the text is just blank.

This helped

@samcarter @JosephWright I have been pondering about it too, and I think the main problem is that one has actually two types of colors: things like "red" "darkgreen" "corporate-color-1" are global colors that should at best be predefined in the preamble. But colors like "alert" or "item" or "block.bg" are semantic colors that one perhaps want to reset in an environment and keep local there. I think we are missing an interface for such local semantic colors and that's why things are awkward.

^^^ sounds like an advice from @DavidCarlisle

@UlrikeFischer Ah, that's a good point: I like the separation

@UlrikeFischer I'll think about it from that point of view when I have my lunchbreak here

@DavidCarlisle is the colour lead on the team: hopefully COVID won't stop him contributing here :)

@JosephWright you mean following my helpful contributions to @UlrikeFischer's search for an ideal colour model in hyperref?

@UlrikeFischer but I wouldn't use colour syntax for them they are more naturally just exposed as normal local commands with the names of colours (or more exactly anything to be used in the argument of a \color command) just as you don't use the font machinery to define the font used in headings or footnotes etc, you just define a normal macro that is called in that context that sets a font.

@DavidCarlisle yes. That is what I meant. One shouldn't try to get a local color by using the color machinery. Colors should be global and one should use something like \color{\alertcolor} or \color{\usecolor{alertcolor}} for a local color. The main question is if one can adapt e.g. beamer and the color code to such a model.

yes or (closer to what you do for fonts) \renewcommand\alertcolor{\color{red}} ... \alertcolor

@UlrikeFischer @JosephWright Could each colour maybe come in two versions, one fixed, one changeable? This way, the class/package author could choose at which places the original version must be used even if the user changed it? This would circumvent problems like the user redefining black and then complaining that transparency no longer works as expected.

@DavidCarlisle well the problem with commands is that pgf and beamer like to have color like "itemize subitemize duck alert.bg" ;-)

@UlrikeFischer a strange choice for a German author :)

@UlrikeFischer yes but if looking for a supply of "good tex constructs to use as a pattern for future design" I'm not convinced I'd start with beamer:-) Beamer may need forever some xcolor compatibility layer with color working as now but....

@DavidCarlisle Agreed: it has some great ideas but also some more problematic ones

@UlrikeFischer So one would expand the argument to colours before parsing - does raise the question of what \newcommand\mycolor{red!50}\color{\mycolor!40} is supposed to do (I get something similar in siunitx)

@DavidCarlisle Doesn't let you do a mix: \alertcolor!50 for a faded alert, for example

I'm going to sketch some things out

@JosephWright well it would be like getting an italic version of a current font: \alertfont\itshape so \alertcolor\color{.!50} (or whatever the . syntax for current color needs to be)

@JosephWright but in practice handling the color arguments in the commands rather than the color command itself may work out better I agree, but keeping close to font syntax has always been something of a design aim, and especially if you want to use coloured fonts in the end, it has some advantages perhaps

\documentclass{article}
\usepackage{xcolor}

\begin{document}
\newcommand\mycolor{red!50!white}
\newcommand\mycolorcmd{\color{red!50}}

{Text
\color{\mycolor!30}
text}

{Text
\mycolorcmd\color{.!30}
text
}


\end{document}

@DavidCarlisle, @UlrikeFischer, @samcarter The suggestions at a 2e level

@DavidCarlisle, @UlrikeFischer, @samcarter Doesn't work for l3color at present as it deliberately detokenizes the input

I think either form is viable, the link to beamer is trickier - for xcolor no change is really viable, but it would be for l3color

@JosephWright couldn't you insert a step for "local colors" \color{mycolor} -> if \l__color_mycolor_tl exists/is not empty, replace with value and repeat. If no local color exists look for a global color?

@UlrikeFischer You want to raise this on the team list or should I?

@UlrikeFischer That would be viable: I was wondering if I could pull off something similar to do a transition for beamer

@JosephWright then one would only need an interface to set such local colors. Probably \color_set_eq:nn could do that, and \color_gset_eq:nn could do a real global copy?

@UlrikeFischer Ah, that's slightly different. I was thinking that if we want to make a clear split, we only have 'managed' colours as global objects, then local/global 'stuff' is managed by individuals/packages as macros - they then don't have the namespace sharing

@JosephWright but if \color_select:n{mycolor} should also checks if this a local color and allow such an indirection then some minimal interface in l3color would be needed or not?

@UlrikeFischer Can I ask you: is there a key for changing the annotation font?

\setchessboard{
style=3x3,
showmover=false,
addpieces={rb2},   %largeboard
boardfontsize=44pt,% changes font of the pieces
%fontsize=15pt,% same?
}

@cis it is called label in chessboard, see 4.8 Labels

@UlrikeFischer Ok, thx

@UlrikeFischer I thought you were arguing that colours should be global, and that such local semantic things should be handled as simple macros? \color_select:n { \mycolor }?

@JosephWright well I think that "real" color should be global. And I think one need something to handle local semantic things. But I'm not sure yet what is the best way to handle such local colors. While commands works, imho from the user view it would be nice if something like \color{alertcolor.bg} would work too.

@UlrikeFischer I'll write something up for the team list: I'll try to summarise

Later ... need to finish exam (questions)

@UlrikeFischer Does she allow a general chessboard.sty tip or is she offended later?

@UlrikeFischer I hope my mail summarises properly

@JosephWright changing the syntax to \mycolor might be really bothersome for theme author who don't know which class version their users have

@JosephWright yes. Just thinking about it ;-)

@samcarter No chance of document-level change: I'd have to find a way of making it work behind the scenes. What I'm aiming at is 'new' material: what's the interface we are promoting

I love Minion Math brainfans.com/brainteasers/…

@mickep :-)

New display of the main page? Lots of information that I don't need...

@AlanMunn A change in the wrong direction.

@mickep It's actually not useless information, but it's displayed in an overly intrusive way.

@AlanMunn I guess one gets used to it. I realized (after the change) that when I browse I look for upvoted questions (probably since they are more interesting to read). With the new design, that information is less obvious when scrolling.

@AlanMunn ooh, there is a main page?

@AlanMunn cough ... userscript ... cough

@AlanMunn Especially when it's information without a sort/filter option. Answers, votes, views are displayed, but I can only sort by Active, Hot (...), Bountied, Week, Month. Lots of vertical space empty. The question pages have still the traditional layout.

@AlanMunn I don't understand stackoverflows sudden obsession with view counts. They already put so much emphasis on this metric in the new profile, but what's the use of this metric? It seems to be mostly a function of age and the frequency with which bots like google visit the sites. Nothing really useful to say about the post.

@samcarter Perhaps because page views are KPIs for stakeholders, to visualize that

And a gamification score (# display, badges) to motivate users to create questions with a high view count.

@gusbrs As a followup to our conversation yesterday, do you find tesseract the best of the OCR tools?

Is it possible to ask a question with the same title of another one, now? ^^^

@CarLaTeX -- I'm not sure it was ever not possible. But obviously it's not a very good idea.

@barbarabeeton It was not possible before. Once I was asking a question and I got an error because there was another question with the same title

@CarLaTeX I never had such an error

@DavidCarlisle you should try the experience of asking a question

@CarLaTeX egreg promised to show me

@CarLaTeX "Please help me draw this" :)

@AlanMunn Indeed, I expect a thousand of "How can I draw this with LaTeX?"

@FaheemMitha Well, if you put commercial/paid ones on the table certainly not... Among the free software ones, it "appears" to be. Meaning here, that's more or less what I've found reading about it (which is something, though not really thorough), but I have not tested any other tool actually. So, truth is I don't have a great answer for this one...

@samcarter Perfect! Thanks for that link.

\documentclass[margin=5mm, tikz]{standalone}
\def\Spoiler{0}

% Get black and white horse...........................
\usepackage{fontspec}
\setmainfont{DejaVu Sans}
\newcommand\Sw{\char"2658}% ♘
\newcommand\Ss{\char"265E}% ♞
\newcommand\Scolor[1]{%
\xdef\first{#1}%
\xdef\second{N}%
\ifx\first\second{\Huge\Sw}\else{\Huge\Ss}\fi}

% Chess
\usepackage{chessboard}
\usepackage{xskak}

\storechessboardstyle{3x3}{%
maxfield=c3,
clearboard,
%startfen=b9,
%restorefen=current,
%labelleftwidth=1.5ex
}
\setchessboard{

@cis For professional looking chess pieces: github.com/samcarter/tikzducks/blob/main/duckpond/Chess.png :)

@samcarter Would at least have the advantage that the figures would then be of the same font and not of two different fonts.

@samcarter Shouldn't the orientation of the two colours be different? Right now both white and black are pointing to the left. But since they're facing each other in a game situation the black pieces should point to the right (or vice versa). :)

@cis you can always do something like \newcommand\Sw{\raisebox{-0.15em}{\WhiteKnightOnWhite}}.

@AlanMunn like this?

@UlrikeFischer Sort of. Maybe my ability to orient things in my mind isn't working, but I had in mind the white pieces pointing "towards" the black ones. So the black orientation seems right but the white pieces look like they're going backwards.

@UlrikeFischer Of course this is obviously mainly relevant for the pawns (ducklings?).

@AlanMunn as you can see it is not so easy on a flat board when white plays up and black down.

@UlrikeFischer Yes. I guess you need to reflect the white pieces. But the simple case doesn't do what I had in mind.

@cis as I already told you no chess player would write the moves like that. Why should a chess package support something that no professional would ever use? I would also never give white and black moves own numbers. These are half moves and the correct numbering is 1. Na4 Nb3 2. Nc1 Nc2 etc.

@AlanMunn Good point!

@AlanMunn Maybe they should all look towards their king and queen?