@PauloCereda Cwac!

How can I get separately both values?

Does anybody know of a PDF viewer for Linux which allows form fields to be filled in presentation mode? I usually use Adobe's reader on university hardware, but this is a Linux-only household, so that's not an option for Zoom-teaching.

Hi @cfr. I don't think I've seen you in chat before, but you've answered my questions on the site a number of times. Mostly about KOMAscript. Thank you for that.

@cfr Good household :) Other than that, sorry, no idea :( I'd try the usual suspects, Evince and Okular. If those don't, I'd use MuPDF, the GL-using version has basic form support (though you might not like the looks of it during form-input). It does support it even in fullscreen mode, I'll send screenshots in a second.

Looks during fullscreen mode:

Looks during form-field input:

After the input:

The yellow highlighting is part of the file, not the viewer. Noteworthy is, that it doesn't show the file during input but instead a small input dialogue. Free input fields, drop down lists, and tick-boxes are supported (maybe more, but that's what is part of this file and those all worked).

@Diaa it depends on the key (it can be just a value, that you can get with \pgfkeysvalueof, or a handler, like in this case).

\documentclass[border=10pt]{standalone}
\usepackage{tikz}
\usetikzlibrary{positioning}
\makeatletter
    \def\mynodedistance{\tikz@node@distance}
    \def\separe#1and#2{x is #1, y is #2}
\makeatother
\begin{document}
\begin{tikzpicture}[
    node distance=0.5cm and 3cm]
    \node[red](red) {\mynodedistance};
    \node[blue, below=of red]{\expanded{\noexpand\separe\mynodedistance}};
\end{tikzpicture}
\end{document}

I found that the key node distance set the macro \tikz@node@distance by looking at the source...

...and don't trust my command of expansion control, I just hope that is ok ;-)

...and well, x and y are the other way around. :-O

Hmmm, maybe \def\separe#1 and #2{x is #1, y is #2} (with spaces around and) is better.

@Rmano You have to read up to a marker, or else you don't know the complete second component but just the first token of it.

@Rmano redefine \separe to expand to {x is #1, y is #2:} to see this.

@Skillmon ach. You are right.

@Rmano also I'd put the \expanded construct in a macro as well, that looks cleaner, imho.

@Skillmon that would be a quark to stop the thing, yes?

@Rmano could be anything, \relax would work as well.

\documentclass[border=10pt]{standalone}
\usepackage{tikz}
\usetikzlibrary{positioning}
\makeatletter
    \def\mynodedistance{\tikz@node@distance}
    \def\separe#1 and #2\relax{x is #1, y is #2}
    \newcommand\getmynodedistance{\expandafter\separe\tikz@node@distance\relax}
\makeatother
\begin{document}
\begin{tikzpicture}[
    node distance=0.5cm and 3cm]
    \node[red](red) {\mynodedistance};
    \node[blue, below=of red]{\getmynodedistance};
\end{tikzpicture}
\end{document}

@Rmano ^^^

@Skillmon Thanks! I was already doing a quite convoluted \expanded{\noexpand\separe\mynodedistance\noexpand\relax}

@Rmano you don't have to protect \relax in \expanded, \relax never expands in TeX's mouth (as long as no evil genius did \def\relax{}).

@Skillmon @Diaa --- use that version, it's way better

@Skillmon Ah, yes. I added it because initially I did \def\stop{\stop} and used it as a marker, just to be naughty.

@Rmano don't do \def\stop{\stop} either, \stop is already defined :)

@Skillmon (blushing) yep, true.

@Rmano also, if you only use things as a delimiter for a macro, you don't have to define it at all. The reason why quarks are defined is so that \def\foo{\quark} evaluates to true in \ifx\foo\quark (which is a fast test to find some marker if it is assigned to some macro anyways).

@Rmano oh, and you don't have to put the end marker for \separe in the argument of \expanded here at all, you could as well use \expanded{\noexpand\separe\mynodedistance}\relax, the braces are removed by the expansion of \expanded, and since \separe only expands afterwards everything's fine :)

@Skillmon (Rmano, don't mess with adult's things and go back to your metamaterial sensors paper, which is easier). ;-P

Or on the bright side, one never stops learning... ;-)

@Rmano in expl3 the current development trend is to use scan marks for argument delimiters instead of quarks. And the scan marks are \let to \relax.

The advantage is that those a) don't create infinite loops when they're hit by expansion, and b) don't expand in TeX's mouth at all, so that you can use them in expansion contexts without losing your end-marker. Though I'd argue they should never be expanded anyways, as they are markers to some other functions which are expanded earlier... :)

@Rmano which reminds me that I should go back to my harvesters... :)

@Skillmon Thanks for the help. Does it also work even if I write just node distance = 15mm?

@Diaa no, test it --- argument of \separe has an extra }.

You probably should check if there is an "and" before trying to separe. No time now --- I have class.

@Rmano I have time :)

@Diaa the following works with node distance = 15mm as well as with and with the spaces around and being optional (which is also the case in TikZ). It borrows the \tl_trim_spaces:n function from expl3. The node distance must not contain the token \separe@mark (that seems like a reasonable limitation).

\documentclass[border=10pt]{standalone}
\ExplSyntaxOn
\cs_new_eq:NN \trimspaces \tl_trim_spaces:n
\ExplSyntaxOff
\usepackage{tikz}
\usetikzlibrary{positioning}
\makeatletter
    \newcommand*\mynodedistance{\tikz@node@distance}
    \newcommand*\separe[1]{\expandafter\separe@\expandafter{#1}}
    \newcommand*\separe@[1]
      {%
        \separe@checkand#1\separe@mark\separe@yep and\separe@mark\separe@nope
          {\separe@xy#1\separe@mark}%
          {\separe@x{#1}}%
      }%
    \def\separe@checkand#1and#2\separe@mark{}

> Python creator Guido van Rossum joins Microsoft

One more reason for my dislike of Python... :)

@PauloCereda If this is what happens if a retired developer gets bored you should never retire ;)

@TeXnician indeed :)

@PauloCereda well, one for reason for not caring about tabloid news and enjoying my favourite scripting language :)

@yo' ooh Lilypond :)

@PauloCereda you really do like Scheme?!

@yo' heavens no. :)

@PauloCereda good. I was a little worried.

@yo' <3

@DavidCarlisle okay, we need to take the bus to UEA

@PauloCereda Saw that on Twitter yesterday :)

@PauloCereda what is UEA

@UlrikeFischer University of East Anglia, @JosephWright's duck pond.

@UlrikeFischer uea.ac.uk

@UlrikeFischer mail explaining the metajoke in a minute.

@Skillmon uber-expansion tricks ;-) I am sure it would be easier in expl3 ;-)

@JosephWright I will submit a motion that at least three observers from UK-TUG have to watch virtually the AGM taking place, as a means to audit the meeting. :) I vote @DavidCarlisle.

@TeXnician another installation in the bag without issues.

@daleif It should be a documented recommendation :D

@daleif ooh are there chips in the bag too? :)

@PauloCereda no, that bag is at home

@daleif ooh

@TeXnician At least it has helped on the stability

@daleif Well, you're not the only one I know who has suffered from stability issues during TL installation…

@PauloCereda, @egreg Just been offered Big Sur ...

@JosephWright Get it

@PauloCereda Plan to, but probably not right now: have a lecture at 2pm

@JosephWright Bah, what could possibly go wrong?

@PauloCereda I'll be doing a video-based session: I might disappear!

@JosephWright oh no

@TeXnician It is just strange my tlmgr is so much more resilient(?) aren't the installer mostly using the same code?

@daleif I really don't know about that…

@PauloCereda If you need some strange entertainment the next week, then Desert bus for hope starts today, see desertbug.org

@daleif ooh cool, thanks!

@JosephWright Week end job

@samcarter_preparing_for_xmas @UlrikeFischer oh no twitter.com/fiene/status/1327231053422075906

@PauloCereda so did Lamport, iirc. Do you dislike LaTeX?

@Skillmon pwnd

@Skillmon bloody macros :)

@Rmano not really, in expl3 I'd just use \exp_args:No instead of the two \expandafters and \exp_args:Nee instead of the \expanded{\noexpand construct.

@PauloCereda Oh, poor cat!

@samcarter_preparing_for_xmas oh no

@Rmano all the other things aren't expandably contained in expl3 as far as I can see, so either one would have to do this unexpandably, then expl3 would be of great help here, or it can only be used to aid in expansion control on two spots, not that much...

@PauloCereda chat.stackexchange.com/rooms/116188/… might interest you :)

@JosephWright are you sure you wanted to ping Paulo?

@PhelypeOleinik another one using a path to call a package: github.com/ho-tex/kvoptions/issues/8

@JosephWright ooh, an expert! ... or is it 'an ooh expert'?

@UlrikeFischer Sigh...

@UlrikeFischer Yes: macros

@UlrikeFischer I have it almost fixed locally, but there are some things to tidy up yet

@UlrikeFischer Though I have to look at kvoptions. It might need to adapt (I hope not) like xkeyval does

@JosephWright ooh

@JosephWright one should write a thesis on those thingies

oopsie I did write one

oh no

@PhelypeOleinik if there is something to change say it, but I had the impression that it is the general problem.

@UlrikeFischer Yes, the fault is on the LaTeX side, but kvoptions might expect a path where I'm removing them (I'm not sure yet)

@PhelypeOleinik I don't see anything referring to a path. And if I add this at the begin of the sty it works again:

\tl_set_eq:cc{opt@\@currname.sty}{opt@kvoption_test.sty}

@UlrikeFischer Ah, I think I was looking at kvoptions-patch. Ignore me :)

@PhelypeOleinik never <3

@PauloCereda <3

@PauloCereda Bad choice though :)

@PhelypeOleinik ooh a choice

Is there a way how to get the package name from the macro name? I'm aware of texdef, but for that you need to know the package name...

@yo' grep and good luck :)

@Skillmon yep, that's not really useful :-)

@yo' awk.

@Skillmon perl.

Use plain TeX and forget packages.

@PauloCereda input expkv.tex <- not a package as it isn't called .sty!

@yo' start from a preamble which loads every package and put the macro you want to know about after it. Then bisect through your preamble until you find the package which defines it. Is that better?

@Skillmon oh

@Skillmon and generate the incompatibility dictionary on the way.

@yo' good idea! :)

@Skillmon I'm afraid this would be EXPTIME.

@yo' ooh no!

(well, maybe only in theory, in reality you could probably say that you only care about incompatibilities of up to 5 packages, and then it's probably P)

@yo' we could try to solve this via FFT, then we'd be nlogn.

@Skillmon don't try to assess that P=NP using FFT!

(though I have no idea how this could possibly be solved via FFT)

@yo' use human memory and ask "do someone know where XYZ is from"?

@UlrikeFischer a nice suggestion to a user :-)

Gotta go, bye!

@yo' why not? not every problem is solved at best alone with some machines.

@UlrikeFischer well I suggest Googling "LaTeX <macro name>"

@yo' depending on the macro that often works fine too.

@UlrikeFischer yeah. Though usually you get some issues with the macro as first hits, and not its basic usage example, or even the pkg name :-)

@Skillmon bye mr. rabbit

<3

@yo' hmmm... googling "latex something" could be dangerous ;-)

@yo' texstudio has .cwl files for the most common packages that contain a list of macros for these packages

txs uses this for example to allow the user to open the package documentation from the macro

Mac friends, a quick Q

Where are local fonts installed?

@PauloCereda have a look into ~/Library/Fonts

@samcarter_preparing_for_xmas Cool!

Just out of curiosity: if a sentence ends in "?" or "!" and the last word is italicized for emphasis, is it customary to also italicize the "!/?" ? If I don't, it looks a bit funny, depending on the font.

@PauloCereda As a more general approach, you can open Font Book and if you right click on any of the fonts you'll get "show in finder"

@samcarter_preparing_for_xmas it's... a secret :)

I guess the fact that it doesn't look good could be considered a bug of the font. With Computer Modern it looks fine.

@PauloCereda Ohh, a secret!

That's from \usepackage{fourier}

A \, fixes it ...

I was just surprised because I am used to invariably excellent typesetting from TeX

@PauloCereda More likely /Library/Fonts (for some definition of local). I have no fonts in ~/Library/Fonts

@AlanMunn ooh semantics

@AlanMunn thanks :)

@AlanMunn I have fonts in my ~/Library/Fonts folder...

@PauloCereda I want my fonts to be available to any account, so I never use ~/Library/Fonts

@AlanMunn Hmmmm

@PauloCereda For the same reason I never install "local" TeX fonts into ~/Library/texmf

@AlanMunn I will not mention my recent TL installs are made in my user space. :)

@PauloCereda Well if you go that route, then it makes sense.

@AlanMunn it's actually... a secret. :)

@AlanMunn oh no

@MoneyOrientedProgrammer heh heh heh... i daresay some might be more tempted to skip the mask. :p

@Plergux De gustibus non est disputandum ;-)

@Rmano lol

@Plergux :-)

ooh another secret project

Ok, that similar question preview just saved me asking a question. I was assuming the issue was chemmacros when its effing lmodern of all things. WHAT. How does that still have a bug in it?!

@Canageek which bug?

That also comes up with chemformula ie \ch{[^{n}Bu4N][Au(CN)2] . 1/2 H2O}

@Canageek quite a few of the lmodern metrics are .. unexpected, changing them isn't trivial though, apart from doing the actual work you have to worry about the existing corpus of documents, changing the metrics changes everything

@DavidCarlisle Oh for sure, you you give it a [compatibility] flag though, or if that is too risky a [modern] tag

@JosephWright Did I make a mistake using chemformula instead of mhchem? I seem to spend a lot of time fixing it.....

@Canageek Yes, and the workaround is the same.

@egreg Already implemented it, which is why I didn't post a question

@Canageek It's not a problem of chemformula, but of Latin Modern.

Kinda wonder if I should have posted it, then answered it right away

@egreg Kinda is? Knowing the most popular font package has that bug I would have redefined that symbol within the scope of \ch to avoid it

@Canageek The package uses \textperiodcentered; if the glyph has wrong metrics…

@egreg I guess it would be somewhat hacky to detect if Latin Modern is being used, and pull from Computer Modern in that case and only that case.

@egreg I mean, does \cdot have that same problem? It makes just as much logical sense for the use

@Canageek The reason for that strange bounding box is really unknown; by the way, the period is higher in LM than in CM.

@egreg Also the [] are heavier in Latin Modern compared to Computer Modern, which surprises me, it wasn't mentioned in any of the articles I read about the conversion.

@egreg Wait, unknown as in we don't know why they changed it, or unknown as in the conversion processes messed it up and no one noticed until it was too late to change it?