@UlrikeFischer well-deserved!

@samcarter_prepared_for_xmas ooh

@JosephWright did you see tex.stackexchange.com/q/575822/2388? The reason is naturally that \text_lower:n returns the result in a \exp_not, but I wonder how to handle this (and who should handle it, hyperref or l3text?).

quack

@PauloCereda boo!

@yo' ooh

Jonathan P. Spratte submitted an update to the menukeys package.

Version:  1.6.1 2020-12-19
License:  lppl1.3c

@Skillmon woooooo

@UlrikeFischer @samcarter_prepared_for_xmas almost 200 views for the extravaganza video!

@UlrikeFischer remind me, why the unexpanded?

@DavidCarlisle have a good week <3

@PauloCereda BJ had more for his video cancelling Christmas, so you still lose on the popularity stakes...

@DavidCarlisle oh

@PauloCereda traditional week for eating d...

@DavidCarlisle oh no

@DavidCarlisle ^^

@PauloCereda Ohh, that's great!

@samcarter_prepared_for_xmas yay!

@samcarter_prepared_for_xmas love your hat! :)

@DavidCarlisle Joseph wants to retain the char, and not expand further to the licr.

@PauloCereda :D Did you get the Sun or the Moon one?

@UlrikeFischer but then why does it work for typesetting, I could look but...

@DavidCarlisle well typesetting continues. Or do you mean why puenc doesn't kick in?

@samcarter_prepared_for_xmas the sun. :)

@UlrikeFischer i'm confused and can't really look now but if you typeset \unexpanded{....} then non ascii stuff in ... doesn't work does it?

@DavidCarlisle this here works fine for me:

\documentclass{article}

\begin{document}
abc  \unexpanded{æ}
\end{document}

@UlrikeFischer ah yes sorry still asleep OK unexpanded doesn't do much here. hmm

@DavidCarlisle wake up quack

@PauloCereda I should have had a cooked breakfast

@DavidCarlisle oh no

@PauloCereda yep, am now officially the second maintainer (after the last update was entirely done by myself as well). The new update just alters some internals so that the separator isn't x-type expanded anymore, so now you can use active separators (but only after you used \newmenumacro or \renewmenumacro with the active token as the separator, there is no category-code normalisation going on)

@DavidCarlisle \unexpanded does (roughly) nothing during shipout.

@Skillmon shipout wasn't my concern I was thinking it was like putting \noexpand in front of every token, but it's not

@Skillmon ... because like all the etex things it's more or less an expandable version of \toks@{...}\the\toks@

@DavidCarlisle yup

@DavidCarlisle this is utilised throughout expl3 for the x-expansion safe functions.

@Skillmon yes just thrown off by being asleep and working backwards from why it is failing in hyperref pdfinfo

@DavidCarlisle I forgive you :)

@Skillmon still leaves @UlrikeFischer's question of bets way to handle it:-)

@DavidCarlisle correct way, imho, would be to handle it at the top-level, like in her answer. That's nothing the l3text module should change, returning in \exp_not:n is the correct thing to do.

Brexit countdown... :(

@Skillmon it is not completly clear. The question is if \text_lowercase:n should contains something that tries detect to the (pdf) context.

@UlrikeFischer I have no idea how that should be implemented... :)

@Skillmon well basically it has to do only a \texorpdfstring{\unexpanded{xxx}}{xxx}.

@UlrikeFischer basically, but is that the correct thing to do in the low-level code?

@UlrikeFischer or the other way and keep unexpanded and add \expanded in the pdfstring version as this works pdfauthor={\expanded{\MakeExpandableLowercase{AbcÆæ}}}, % Does not work!

@Skillmon that is the question. But it is naturally easier to do it there: The code knows exactly where it inserts an \unexpanded that it can leave out, that is easier to do and more logical then to nullify it with an external \expanded as I did it in the answer.

@UlrikeFischer well, I'd say the low-level code should be agnostic and the high level code should add that abstraction, but not my decision.

@Skillmon seems hard though as not everything you want to put in pdfstring is going to be safe in \expanded is it? shouldn't the \unexpanded only get added in an expansion-only context (ideally)

@DavidCarlisle \unexpanded\neq\noexpand :) Agreed.

@Skillmon well the problem is that it is more difficult to do it on a high level, things are much easier if from the start on the code considers both pdf and text context.

@Skillmon my edit was faster:-)

@DavidCarlisle I already tried to strike it, but I can't remember which mark-up it is in this chat :(

@Skillmon this

@Skillmon I think three hyphens hyphens

Hi, I want to write a bilingual (Bangla+English) document in Overleaf (Google Chrome, Windows 10). But I found that the cursor isn't appearing in right place in the Overleaf editor. I guess it's due to the Bangla font. The MWE of my document is here:

@UlrikeFischer thanks. It's tildes almost everywhere else, I think... :(

@raf better raise that with overleaf support rather than here

@DavidCarlisle okay.

Hi everyone! I've just written a port of TikZducks for ConTeXt using Metapost/Metafun, so no TikZ anymore (I won't read a thousand-page manual just to do tweaks). However, I don't understand LPPL license (I'm reading it and I find it confusing), so I want to know whether or not it is legal. Just in case, I've not included the Overleaf duck. Can I ask here or should I ask it as a proper question in this site? Thank you in advance and sorry for any inconveniences

drive.google.com/drive/u/1/folders/…

As preposterous as it may sound, I want to use them for commercial purposes. I remark it just in case it's necessary

@JairoA.delRio usual not-a-lawyer warning but LPPL is designed to let you make and distribute arbitrary modifications even for commercial use, so long as you don't claim it is the original. It's a shame you didn't get rid of the ducks as well as getting rid of tikz:-)

@JairoA.delRio maybe asking @samcarter_prepared_for_xmas is a good idea :)

@DavidCarlisle I love ducks and my girlfriend (a duck breeder) as much as I hate TikZ (or, to be specific, its feature creep). Thanks for the clarifications. The license mentions maintainers, but it's not clear about porting to other formats.

@Skillmon I'll ask her. Thank you very much.

@DavidCarlisle you are mean

@JairoA.delRio As far as I understand, you'll just need a new name for your package (which makes sense anyway if you don't use tikz :) )

@DavidCarlisle oh

@DavidCarlisle How is the duck supply in GB going? I hope you don't rely on imports.

@samcarter_prepared_for_xmas might be difficult to import food from Brazil, but I hear we have a good supply in East Anglia

@samcarter_prepared_for_xmas Oh, true. Thank you for your answer. I'll upload a beta to CTAN soon so ConTeXt users will enjoy ducks too. Thank you again.

@JairoA.delRio los patos agradecen :)

@PauloCereda Los patos lo son todo ❤️ Yo les agradezco a ellos

@JairoA.delRio ooh <3

metaducks soon in CTAN :)

@PauloCereda I thought a plain "ducks", but I love your proposal more than mine. It will be "metaducks". :D

@PauloCereda Thank you a lot. I'm terrible with names.

@JairoA.delRio ooh

@JairoA.delRio so am I :)

@JairoA.delRio metaduck sounds great!

@samcarter_prepared_for_xmas @JairoA.delRio crispyduck would sound better

Anyone know of a "nice looking" underarrow construction? \underleftarrow{v} looks way to big

@DavidCarlisle revolting :)

@DavidCarlisle or duckalorange?

@UlrikeFischer so many variations possible, almost enough to have a package to control the options....

@DavidCarlisle oh no

@UlrikeFischer oh no

@UlrikeFischer you mean \duck[body=orange]? :)

@samcarter_prepared_for_xmas phew

@samcarter_prepared_for_xmas \duck[body=lacquered]?

@UlrikeFischer ! Package pgfkeys Error: I do not know the key '/tikz/lacquered' and I am going to ignore it.

@samcarter_prepared_for_xmas you forgot to load the kitchencolors package.

@UlrikeFischer Is it by the same author as the xkcd colour package?

@samcarter_prepared_for_xmas that could an idea (lurking while at a class)

*could be

θ60° 1d [romano:~/software … urce/xkcdcolors] master 1 ± grep -i roast xkcdcolors.sty
1d [romano:~/software … urce/xkcdcolors] master 1 ±

^^^no luck

But we have BurntOrange,C04E01;%

@UlrikeFischer Random weird thought regarding bitset whatever the name: Is there any case where it is necessary to set a flag expandably? If so, then it could be an actual set of flags from l3flag. Though it would not be so easy to switch them off expandably (not impossible though).

@PhelypeOleinik YES NOW YOU DON'T LOSE BALANCE <3

@PauloCereda OH YES! :D

@PauloCereda It's hard to balance a sombrero on your head when you're a green square though :)

@PhelypeOleinik LOL

@PhelypeOleinik this seems like serving for a hard edge case with comparatively slow code. Also, flags aren't boolean, as far as I understand.

@PhelypeOleinik and now I'm curious how you plan to switch flags off expandably... :)

@PhelypeOleinik well I don't set them expandably currently, so the answer is obviously no ;-). My main requirements is that I want an interface to set by name and something to output the arabic number (and this should be expandable). But I don't think that it would make much sense to define a bunch of flags for one set, this uses up csnames for not much gain, I mean storing everything in a string or a sequence or even a prop works fine.

Currently trying to debug a piece of TikZ code. Why does \foo become contain \relax here?

\def\X{2}
\edef\foo{%
    \ifnum0<\X
        \ifnum0=\X
            foo
        \else
            \ifodd\X
                bar
            \fi
        \fi
    \fi
}
\show\foo
\bye

@HenriMenke missing space after 2 I guess

@HenriMenke yes I guessed right:-)

@DavidCarlisle But why does it yield \relax? Which token is it that turns into \relax?

@HenriMenke ot is expanding the second ifnum before the first looking for the end of the number and ends up with a frozen relax let me see if I can adjust it to b eclearer..

@HenriMenke vv

\def\X{2}
% \def\X{2 }
\edef\foo{%
    \ifnum0<\X
    \fi
}
\show\foo
\bye

@HenriMenke if you hit the \fi while still scanning for the number TeX inserts a frozen \relax

@DavidCarlisle Aha! Thanks.

@HenriMenke if the \X is coming from outside and you can not trust it to be a terminated number, then with etex you can do

\def\X{2}
% \def\X{2 }
\edef\foo{%
    \ifnum0<\numexpr\X\relax
    \fi
}
\show\foo
\bye

@Skillmon Not a flag per se (as that's not possible) though if the application so requires you can use a positive and a negative flag (much slower than standard flags, yes :)

@DavidCarlisle What I find really interesting is that you can actually omit \relax in \numexpr\X\relax and it still works. Looks like the rules for scanning for a number are different inside \numexpr.

@HenriMenke hmm I think maybe the numexpr absorbed the frozen relax....

@UlrikeFischer Sure, it was just a thought that crossed my mind when flags were mentioned. Ah, yes, your implementation is very memory-efficient, but with flags it wouldn't even be a fair comparison ;-)

@barbarabeeton andrewhartford.medium.com/…

@PauloCereda ööh

@samcarter_prepared_for_xmas <3

@PauloCereda What a great story. A very clever fix too!

@AlanMunn Indeed! The YouTube videos showing the fix are awesome!

@PauloCereda I'm a bit afraid of the videos YT will suggest after watching stone masons at a cemetery

@samcarter_prepared_for_xmas I don't bother anymore, my YT recommendations are already as wacky as possible. :D

@PauloCereda :)

@samcarter_prepared_for_xmas so... youtube.com/channel/UCYJdpnjuSWVOLgGT9fIzL0g :)

I posted the following question a few days ago. Can anyone help?

@PauloCereda -- Ooh!. A good item for my TUGboat column! Thanks!

To everyone, happy Winter Solstice!

@PauloCereda babel shortcut failure again "O

@user21820 probably it didn't get many responses as it is not very clear what you want.

@UlrikeFischer Help!

@DavidCarlisle how unfair!

@UlrikeFischer seems like cheating to me.

@DavidCarlisle ;-). Weren't you impress how I handled the pax issue? I had to check the code as it no documentation you can't look at ;-)

@DavidCarlisle People doing strange things these days!

@egreg this for example

@DavidCarlisle Amazing!

Or astounding, which one do you prefer?

Is it "well known" that xstring source is in latin-1 ?? :(

(/usr/local/texlive/2020/texmf-dist/tex/generic/xstring/xstring.tex
\integerpart=\count181
Invalid UTF-8 byte or sequence at line 35 replaced by U+FFFD.
\decimalpart=\count182
Invalid UTF-8 byte or sequence at line 79 replaced by U+FFFD.
Invalid UTF-8 byte or sequence at line 79 replaced by U+FFFD.
Invalid UTF-8 byte or sequence at line 83 replaced by U+FFFD.
Invalid UTF-8 byte or sequence at line 86 replaced by U+FFFD.
Invalid UTF-8 byte or sequence at line 102 replaced by U+FFFD.
Invalid UTF-8 byte or sequence at line 114 replaced by U+FFFD.

@MarcelKrüger I have tried to import the context files, but it errors.

warning  (node filter): error: ...texmf/tex/latex-dev/luaotfload/fontloader-2020
-12-21.lua:22193: attempt to call a nil value (upvalue 'flush_components')

To late to look now, but I pushed the test branches of luaotfload and lualibs

@DavidCarlisle did you check the documentation?

@UlrikeFischer no

@DavidCarlisle ;-). Who asked about it? And why did nobody complain before?

@UlrikeFischer minted question on main site, only xetex complains and it only complains in the log. I think actually the non ascii is only in comments but it's still a pain

@DavidCarlisle yes, fills up the log a bit ;-)