@DavidCarlisle because those two were in the discussion on the github thread, sorry... @DavidCarlisle I created the PR for the lesson-08 rewrite.
@DavidCarlisle it did work with pandoc this way, so I hoped it would with Github's markdown as well... Could you push the working version, I'm still not sure what I should do. Should I build the entire table with HTML instead?
@Rmano A definite yesn't
@DavidCarlisle so many comments about my bad English.... :)
@Skillmonlikestopanswers.xyz no markdown tables will make it easier to keep the same style but I couldn't see a pure markdown way to get the | into a cell so just putting span around that will fix it up. I'll do that in your branch
@Skillmonlikestopanswers.xyz pushed to your branch
@Skillmonlikestopanswers.xyz mostly the comments were not that the English was bad so much as too chatty/colloquial for the style of the rest of the course.
@Skillmonlikestopanswers.xyz You could consider \usepackage[whole]{bxcjkjatype}
% tex.stackexchange.com/a/450483
\documentclass[a4paper]{article}
\usepackage[whole]{bxcjkjatype}
\begin{document}
\newcommand{\JaSampleText}{%
Direct input 日本語
or UTF-input \UTF{65E5}\UTF{672C}\UTF{8A9E}
works as well.}
\section{Kana and Kanji}
\begin{itemize}
\item \rmfamily Mincho (明朝) family: \JaSampleText
\item \sffamily Gothic (ゴシック) family: \JaSampleText
\end{itemize}
Hint: You can you use the well-known commands \verb+\sffamily+ and \verb+\rmfamily+.
\end{document}
@cis presumably not your fault but \UTF seems a strange choice of command name. The 日本語 form which is input in UTF-8 has no UTF command but the form using Unicode code points written as ASCII hex numbers not any of the Unicode Transformation Formats, uses \UTF
@cis well in xetex or luatex you could use \symbol{"65E5} or the primitive \char"65E5 or ^^^^65e5 syntaxes. UTF referrers to formats defined by Unicode consortium such as UTF-8 which is the encoding used by this site (most sites) and allows you to write 日本語 directly. But it doesn't really matter, it's only a name.
@DavidCarlisle should be better now. And there were a few sentence I had issues to follow today, just one day after writing them. So all in all it was not-so-good Genglish.
@DavidCarlisle "xetex or luatex ..." Oh no, no, I tried a lot with this general syntax this morning. So as long as there is no package with `\writedeown{<font>}{<unicodenumber>}` or something....
I also believe that the command can essentially only translate the Japanese "unicode numbers".
I think it's especially good that pdflatex also works without any problems.
@cis yes you need to use a package-defined command if you are using pdflatex, and unless you are the package author you can't really change the command name I wasn't suggesting you change anything, just commenting on the package syntax. But there are lots of packages that could have better syntax:-)
@JosephWright I'm trying to remember if I knew this before:-)
One important thing that escapes from \mathchoice is a math style assignment, so you can do \mathchoice{\scriptstyle}{\scriptstyle}{\scriptstyle}{\scriptscriptstyle}. I had thought about amsmath or LaTeX overloading those settings to do bookkeeping and know the current style, but the bookkeeping records will not get out. — Donald Arseneau7 hours ago
@cis sure I was only commenting on the name not the usefulness.
@cis but still if you are typing a lot of Japanese pdflatex is perhaps the least recommended of the common engines, xelatex, lualatex or (especially) uplatex have a lot more features to support non Latin scripts and a much wider choice of available fonts for japanese
@JosephWright source2e (or interface3) keep coming in to my head, but I think they are perhaps too far off, but on the other hand this is the last lesson so....
@UlrikeFischer, @DavidCarlisle I've just spotted a (Japanese) tweet about learning resources; I've suggested a translation would be good, but we likely need to decide where 'language-specific details' go
@DavidCarlisle well I wouldn't add xii.tex and ducks are probably too much of an inside joke. But there are quite a number of other things possible which would add a bit color and give a glimpse of the variety of latex to the end of the lesson, e.g. tikzpeople, or the xkcd imitations, or coffee stains, or tcolorbox, or something showing music.
@UlrikeFischer tcolorbox and music OK but not sure where, currently the last lesson is "more on documentation" that's more of a final lesson-16 "things you can look at after this course"
@PauloCereda I know you oscillate between wanting to be English and German and you actually are a bit Italian, but have you noticed which hemisphere you live in?
@PauloCereda yes but then google de-indexes the page so it won't be found by site search, just seeing if we can force it back...
@JosephWright this not very helpful advice:
If a page on your site is available and should be displaying in search results, it means Googlebot sees it as “thin” content. In this case, you need to work on improving the content on your page so you can avoid triggering a soft 404 error.
@DavidCarlisle -- Yes! Make real sure there's something about "it's essential that you know how to find the log." So many questions here show that too many front ends hide it.
@JosephWright -- It does say that there is a .log file, but does.t say where to find it. I guess that's too variable to be specific, but if it's possible to add a hint, even if only to ask the potential helpers where to find it. Maybe a community wiki question here on the main site, "Where can I find the log file?", could gather information on the various front ends.
@cis ? in xetex or luatex the characters just pass straight through (more or less) there isn't any code, in pdftex there will be some macros decoding the number and working out which font that corresponds to, the details are specific to the package you are using. (for cjk packages that are mapping large numbers of characters)
@DavidPurton I have played with your checkbox code and have now a pdf here github.com/latex3/pdfresources/blob/test-new-hooks/experiments/…. Unlike your code it doesn't create a new xform for every checkbox but only one per appearance. And it allows to group the checkbox (the one with the same number can be checked together.
@DavidPurton But I'm not sure if one can integrate this into the current hyperref drivers - that would probably need quite some rewriting as I had to split the code for the field dictionary and the annotation dictionary.
@JosephWright yes. Moving catalog, pageresources etc under "pdfdict" worked out very well. But it started quite some thinking about how to organize/name the other things.
@DavidCarlisle I was thinking about translations of learnlatex.org: we could have language-XX pages for more 'open-ended' language-specific additions, then link to them from the index for that language
@UlrikeFischer Aren;t steams meant to be compressed or not based on the object compression level? How does this work with a 'mixed' situation?
@JosephWright this is all a bit muddy. You can find references that the metadata stream should not be compressed. But I'm not sure if this is really required. With luatex you can do it with the uncompress keyword and then you get a partly uncompressed pdf. With pdftex one can locally change the compresslevel. with xetex and dvips it is imho only possible to not compress at all. You can try with hyperxmp:
@cis a true type font isn't a readable file so you can't really look at the code, it internally stores curves and lines for each character but it's more like a graphical editor you can't really see how the curves came to have those shapes, it may have been generated by something or the designer may just have dragged the control points around on screen until it looked nice.
@JosephWright language- would be easier as then just need to look for fixed strings where we are already looking for more- and lesson, but en- would work (as that string will need passing in anyway as the directory to use)
@JosephWright it will do (although you'll probably offend someone whatever you call it, but "languages locales scripts bcp 47 tags and other related concepts" is probably a bit long
@JosephWright hm, verapdf says only that the stream should not contain the filter keyword, but I would have to build some test files to check if that means that we don't need to care about stream compression.
@JosephWright while testing the overleaf I noticed that the project at learnlatex.org/en/lesson-13 doesn't actually have a bib file so you get a no bib file warning....
@yo' the test url I had given to your support address earlier worked straight away but I had to re-arrange things a bit as initially I had all the subsidiary files first (as I just edited the filecontents loop) but overleaf is a bit unhappy if the first item is a bib file so I had to re-arrange a bit, simple cut and paste once I'd worked out what needed moving where:-)
@DavidCarlisle ah! Well, is there something to note down? We can definitely either make a quick fix, or just make a note about the issue so that it's not forgotten :-)
@yo' no I think it's implicit in the docs (and same as the multi-file version that you had before working for http uploads, it's not unreasonable that if you submit 10 .tex files it needs a hint as to which is the top level entry point (or at least I think that's the case, putting it first doesn't hurt certainly:-)
@yo' no but I need to read them to add the engine= parameter to the post submission, as the only interface the end user has is editing the text block so the comment interface seemed most natural
But I was going to show you the %!TEX comment working for overleaf submission learnlatex.org/en/more-14 was working but it looks like i just broke it, I need to check, probably me not you:-)
@DavidCarlisle but I'm mostly off now. If you keep the project as is, I might have a look tomorrow, but basically running it through UTF decoder (such as VIm) should suffice to see the issue.
(other issue is when the files are too large, but that's not the case here... :D )
also, if you ask about it at support@overleaf.com, it'll probably get passed to a developer who might have a better idea...
I don't see anything unexpacted first hand, but I really gotta go, I get up earlyish tomorrow
I'm glad that the multifile thingy works! With these little items, feel free to prod me here in chat, I'll either know how to help or know someone who knows :-) Good night!
@UlrikeFischer LuaTeX does not insert an explicit discretionary if a word ends with a hyphen. Your Java-\textrm{API} is equivalent to Java-\/API\/ where the \/ inserts a kern node of subtype italic correction. In contrast to what the LuaTeX doc claims, a kern node always ends a word except if the subtype is fontkern (The manual claims that italic correction is the exception). So as far as LuaTeX is concerned, the kern ends the word.
@MarcelKrüger it is a bit of pain. The question came from here tex.stackexchange.com/q/545274/2388. But I don't have an idea what to do here, apart from saying to use the babel shorthands.