Good night everybody!

What is this room?

@egreg O wow, there really seems to be a question for anything.

Hi!

@PauloCereda hello there!

@yo' Hi Tom! :)

@PauloCereda how well do you know fontspec, please? ah seems I'm fine, there's an answer by @egreg on the problem :-)

@yo' It should be easy to find which one, there are so few… ;-)

@yo' I know nothing. :)

Going to the beach, see you later.

@egreg I was looking for FakeSlant ;)

@egreg Yay, lots of sandcastles!

Beware of the water expansion, the castle should be robust. :)

@JosephWright Quick question, do you have some kind of every SIrange/.append stlye= in siunitx? I just know the normal \sisetup but not how to reach only certain commands.

I just used the TikZ syntax as I think, it is an easy way of explaining what I am searching. Not that I would query for such command.

@LaRiFaRi Nope

@LaRiFaRi There's no separate settings for different sub-commands other than those explicitly listed

@JosephWright Ok, would have been surprised if I had overseen that in your manual. Thank you.

@PauloCereda Don't worry, the expansion is limited, only 7\cdot10^{-5}\,\mathrm{deg}^{-1} :)

@JosephWright I guess, you have seen tex.stackexchange.com/a/261924 Maybe it would be nice to have a new feature: SIrange should check wether it should result in two different unis like here or if not. Normally, I have range-units=single loaded. Would be cool to get 3 to 5 Hz but 3 Hz to 5 kHz if the case of that post applies

But I guess, this is to much effort for little improvement...

the incompleteness of beautiful fonts drives me crazy makes me sad :-(

@DavidCarlisle, would you mind having a look at this one. None of us looking at it understands what is going on.

@LaRiFaRi It's also inconsistent

@JosephWright indeed. I would say: one of the cases where only bad solutions exist :(

@daleif Did you forget to attach a link?

Heylo all

@Johannes_B Seen this? I'm really not sure what to do

Is it true that the aux file is read twice for each compilation?

@1010011010 yep.

@yo' So this \captiononpage won't work the way I implemented it. :(

@1010011010 why wouldn't it? The second loading is at the very end, which means that you can ignore whatever it's doing

@DavidCarlisle, oops, forgot the link: tex.stackexchange.com/q/261884/3929

@Johannes_B yes, got distracted

Strange, it must be something else then. Can't figure out what, though.

@1010011010 got a MWE?

@yo' I'm not sure whether I can make it minimum. :-P

@1010011010 uh oh :)

@yo' The problem is that one algorithm requires another algorithm to make sense... They're mostly complex and it will be time-consuming to understand the full scope of it.

@1010011010 well, I can't help you then I think :)

@JosephWright hm, not in my taste as I use it just inline in text. But well, doesn't matter. No big need for that. Thanks.

@yo' Hm, there's one thing which may fix everything. I need to reset a counter after a page is constructed, but \AtBeginShipout{\global\setcounter\caption@counter@i\z@} is simply ignored. Is that normal? I guess I could MWE this problem if the reason this is ignored is not obvious.

@1010011010 I'm not sure it's a good idea to hook into shipout with this, but it should IMHO work. If you have a MWE, I can have a look

Actually, I found the culprit.

@1010011010 good :)

I still have that issue where I basically declare something 4 times, the counter is incremented 4 times, and then reset to 0. It should somehow count according to the value of \thepage.

@1010011010 this works like charm:

\documentclass[]{article}

\makeatletter

\newcommand*\yo@captioncounter{%
	\@bsphack
	\protected@write\@auxout{}{\string\yo@captiononpage{\thepage}}%
	\@esphack
}

\newcommand*\yo@captiononpage[1]{
	\@ifundefined{yo@cap@#1}{%
		\expandafter\gdef\csname yo@cap@#1\endcsname{1}%
	}{%
		\expandafter\xdef\csname yo@cap@#1\endcsname{\the\numexpr\@nameuse{yo@cap@#1}+1}%
	}%
}

\newcommand*\captionsonpage[1]{\@ifundefined{yo@cap@#1}{0}{\@nameuse{yo@cap@#1}}}

\let\yo@old@caption\caption
\def\caption{\yo@captioncounter\yo@old@caption}

@cfr I'm curious about your comment that you find forest syntax easier to relate to the structure than qtree. Apart from the interface to TikZ itself (which is clearly cleaner in forest) isn't the only difference in the syntax the . before the node labels and the fact that terminal nodes need to be bracketed in forest?

@yo' In my case, I have to additionally lead values to an algorithm, which leads to hardcore complications. I must say I'm considering giving up on this and just using a local logic tree for a specific scenario...

@1010011010 Well, the call \captionsonpage{\thepage} is expandable, so there shouldn't be much problems. But you're not really explaining what you do, so ... :)

\documentclass{classnofonts}

\usepackage[demo]{graphicx}
\usepackage{letltxmacro}
\usepackage{atbegshi}
\usepackage{ifthen}
\usepackage{everypage}
\usepackage{lipsum}

\makeatletter
\newcount\caption@counter@i \caption@counter@i\z@
\newcount\caption@counter@ii \caption@counter@ii\z@
\newcount\caption@counter@iii \caption@counter@iii\z@
\newcount\caption@counter@iv \caption@counter@iv\z@

\LetLtxMacro{\oldcaption}{\caption}

\def\captiononpage#1{%
\expandafter\ifx\csname npos-page-#1\endcsname\relax

along with gist.github.com/TobiasBouma/798444f698dd637cafa2

The values in the aux file are correct, I need to sort the position at the first instance of a float on whatever page. I can only do it at declaration, somehow. Or I still need to hook into the float placement algorithm (I honestly see no other way).

Expected output should be figure 123456 (obviously) with pageheight being the deciding factor.

@1010011010 I still don't get it -- you need to know whether the float is first/second/third/... on its page?

I need to know when the first float on page x is typeset, in which case I need to take special action.

This gives the desired output on a page by page basis:

\documentclass{classnofonts}

\usepackage[demo]{graphicx}
\usepackage{letltxmacro}
\usepackage{atbegshi}
\usepackage{ifthen}
\usepackage{everypage}
\usepackage{lipsum}

\makeatletter
\newcount\caption@counter@i \caption@counter@i\z@
\newcount\caption@counter@ii \caption@counter@ii\z@
\newcount\caption@counter@iii \caption@counter@iii\z@
\newcount\caption@counter@iv \caption@counter@iv\z@

\LetLtxMacro{\oldcaption}{\caption}

\def\captiononpage#1{%
\expandafter\ifx\csname npos-page-#1\endcsname\relax

Here I just spam \sortpos every chance I get.

Well, you seem to simply want to number figures on a per-page basis, right?

@PauloCereda \DeclareRobustCastle

@egreg ooh :)

@1010011010 Well, you seem to simply want to number figures on a per-page basis, right?

@1010011010 Would this do the trick?

\documentclass[]{article}

\makeatletter

\newcommand*\yo@captioncounter{%
	\@bsphack
	\protected@write\@auxout{}{\string\yo@captiononpage{\thepage}}%
	\@esphack
}

\newcount\yo@figcnt

\newcommand*\yo@captiononpage[1]{
	\advance\yo@figcnt 1\relax
	\@ifundefined{yo@cap@#1}{%
		\expandafter\gdef\csname yo@cap@#1\endcsname{0}%
	}{}%
	\expandafter\xdef\csname yo@cap@#1\endcsname{\the\numexpr\@nameuse{yo@cap@#1}+1}%
	\expandafter\xdef\csname yo@fig@\the\yo@figcnt\endcsname{\@nameuse{yo@cap@#1}}
}

\def\thefigure{%

@HarishKumar No worries :-)

@1010011010 (Well, the version I posted expects no other floats than figures in the document, but it can be easily adjusted)

@1010011010 that almost certainly isn't what you want to do!

@daleif perhaps but I also need to pack bags:-)

@DavidCarlisle Nope, it's not.

@daleif In my opinion that's something for Frank Mittelbach.

@UlrikeFischer don't you trust me? :-)

<3

time to go. see you later!

@UlrikeFischer Might be better. I'll try to summon him. (have been seeing too many MtG videos lately)

Can anyone summon Frank? I do not think I can summon him to the chat.

@daleif you can, if you know the dirty trick ;) But it may be better to simply ask @Joseph, or send Frank a mail :)

@daleif We can. :) @FrankMittelbach, we need you! <3

@DavidCarlisle Well for deep problems with colortbl and longtable I would suggest to ask you. Hopefully Frank wouldn't be offended ;-).

@UlrikeFischer, @daleif I'm on holiday and I haven't the faintest idea, ask Frank:-) But I did add a more minimal example to the question.

Off topic?

@egreg Well, I dunno, he mentions LaTeX in the text...

@yo' Thanks for your MWE buuut yeah, basically the issue is that I can't get something done related to the value of \thepage (which I use extensively). How can I access the value of the page at the moment some macro is typeset? Rather than when it is declared?

Your MWE kind of doesn

't need that jibber jabber.

@1010011010 you can, but only for writing to a file. That's why I do what I do :-)

@yo' So this can't be accessed at a given point in the document, only at the end or start of the document?

@1010011010 well, it's better to ignore the "at the end" part of AUX fiddling, you don't need to bother yourself with it. And yes, you can only access it at the beginning, that's why you have to store the information in a separate macro

@PauloCereda are you sure it reached him?

@yo' Maybe. :)

@PauloCereda do you know the secret way how to reach him "for sure"? :)

@yo' Only mods. :)

@PauloCereda no no

@FrankMittelbach Let's see whether this message reaches Frank :)

@egreg I don't think so. I've posted a comment with a possible duplicate. I think the person wants to know how to put verbatim type stuff in a nice box.

@AlanMunn Yes, I realized that

Proofs all done for TUGboat :-)

@JosephWright you're faster than me, but that's probably because I was late with the first version :)

@yo' All relatively straight-forward

@JosephWright well, professional copy editor/typesetter means easy proofs, usually :)

@yo' According to the newest information, it seems the issue is gone.

@Johannes_B that's ... unbelievable.

@yo' Preparing an MWE can save the day.

@Johannes_B woohoo, magic! ;)

@Johannes_B :)

The wikibook claims that markboth is only available with book. Can i get a quick ok to delete it?

@Johannes_B yep. Moreover, isn't it a completely useless information?

@Johannes_B Delete.

@Johannes_B \markboth takes two arguments, the first one sets the \leftmark, the other one sets the \rightmark. In a oneside document, all pages are considered to be right-hand ones.

@Johannes_B What's true is that in article.cls the \leftmark is never set unless one uses twoside, \pagestyle{myheadings} and \markboth explicitly.

@egreg vvv

\documentclass[twoside]{article}
\usepackage{blindtext}
\pagestyle{headings}
\begin{document}
\markboth{Hi, how are you?}{another test}
\blindtext[10]
\par
leftmark:~\leftmark
\end{document}

If i want to have different content for the left-handed pages, i need \markboth. At least this is the message the Wikibook is trying to say.

Thinking of deleting the manual approach, though needed for starred chapter stuff.

@JosephWright I asked you and @StefanKottwitz this a few days ago, perhaps you missed it. Is there any way for mods to add/change the duplicate that a closed question points to?

@AlanMunn I can reopen then reclose

@AlanMunn As I don't need other votes, that's a fast process

@JosephWright But does that then only allow you to replace the existing link with a new one? Sometimes it's helpful to have two links, and it does happen naturally when close voters suggest two different duplicates.

@AlanMunn That I know of, a mod swap will only leave one link

@JosephWright I see. Well this one should be pointing to this one instead of what it currently does.

@AlanMunn Sorted

@JosephWright And this one ideally should have two links, one to the current dup and one to this one also. I suggested that in the comments but it got closed before I came back to vote.

@JosephWright If you reopen it and I vote to close with one link can you mod close with the other and then both will appear?

@AlanMunn I imagine so

@JosephWright Because although the OP seemed happy with the original dup, the one you just changed it to is much closer to the kinds of trees he wants, so having both linked would be helpful to others, I think.

@JosephWright Or we just leave it as you've done.

Can somebody check the documentation of standalone for containing the math option? I see the option in the class definition, but i am missing the reference in the doc.

Hey, which animal this photo resembles? @Paulo :)

@yo' :)

(And platypus is the wrong answer)

@yo' ooh

@yo' @PauloCereda Without thinking, just guessing: How many results does the main site report for platypus? :-)

@Johannes_B Potato platypus. :)

:-D

@PauloCereda what's up?

@FrankMittelbach Hi Frank! :)

@PauloCereda oh you did it! :D

@yo' <3

@PauloCereda not really here ... we just got a new instrument in the house

@PauloCereda now, I think I should wash up and go home

@FrankMittelbach ooh cool! :) Which one? Tell us!

@yo' Have a safe trip back home.

Kawai CA97

@PauloCereda it's 5 minutes walk :)

@FrankMittelbach oooooooooooh :)

@Frank: @Ulrike and @daleif were talking about this thread: tex.stackexchange.com/questions/261884/…

@FrankMittelbach ooooooooooooooooooooh

That sounds awesome!

Literally sounds too. :)

@yo' its more for my wife :-) but I found out that it works nicely as as bass amp :-)

@FrankMittelbach uh oh

I thought of getting one for me, but I ended up with a cheaper Yamaha model. :)

@PauloCereda that one is really expensive. IMHO more than necessary, but that's everybody's choice (and quality level)

@yo' Well, everything in Brazil is already insanely expensive. :)

@yo' allows me to play with her without getting an amp into that room

@yo' A rabbit? (Blame Wittgenstein.)

@PauloCereda you should start smuggling pianos :)

@PauloCereda will have a look but not before tomorrow morning

@FrankMittelbach you play bass + piano? Sounds interesting ...

@FrankMittelbach No worries. :) Thank you very much. :)

@yo' Including a double bass!

@PauloCereda well, I hope no true musician reads anything else in "bass" :-P

@yo' I play bass (did a bit of piano too way back but never really got into that). and a few other instruments (none good :-( ) but there you are

@FrankMittelbach I neither play any instrument really good, but I also play a number of them :)

but most often I like to play on people's nerves :)

@yo' there you are ... even did a bit of composition in my wild days

@FrankMittelbach cough tohe.wz.cz/c24.html cough

@yo' :-) difficult to read but possibe to listen to

@FrankMittelbach well, there's not much more than a disclaimer that I'm not really strong in making recordings (and the recordings are really not good)

I've got much more arrangements done, but I want to publish at least reasonable quality recordings with them. Also, it's commonly only the song arrangement, without any prelude, and I'd really like to include preludes to make the work more complete.

anybody TeXshop knowlegable? First time I use it on a mac and I don't seem to find the transcript window

@FrankMittelbach If you hit Command-T, the transcript window opens automatically.

@FrankMittelbach You'll have fun trying to use TeXshop on anything other than a Mac ;-)

(Not that I could answer that question)

@egreg doesn't ... I just get the output pdf window

@FrankMittelbach In the Console tab of the Preferences, you can set the console to appear always or only in cases of errors.

@egreg thought it must be somethign like that - thanks

@egreg hmmm, it says "immer" there already but I can't find the bloody thing ... maybe I should just close the lid and look at it tomorrow

@JosephWright isn't there onee on windows? I thought so

anyway time to unravel that tomorrow - have fun folks

@egreg: Maybe you will like this script: github.com/apaszke/termtile

@FrankMittelbach No, TeXshop purely a Mac program: I use TeXworks

@JosephWright Soon, Artemis, the best TeX editor ever conceived. :) L3-compliant!

now I go. bye!

Yay, cardiff is on its way to my humble home! :)

@1010011010 just use \thepage to get what you asked for, but that is fairly arbitrary number, and rarely what is wanted. If you have a long paragraph for instance all macros in the paragraph are expanded while the page counter has the same value, even if the paragraph spans several pages.

@StefanKottwitz golatex.de/viewtopic,p,74391.html#74391 Reading it a few days later. it doen't have that tone.

@AlanMunn There's also the need to leave spaces before closing square brackets. But the terminal nodes thing I find really difficult to relate to the tree structure. It just makes a whole lot more sense to me when these also get square brackets, as they do in forest. Also, there's a lot more room for error because, for example, node content almost always needs to be enclosed in curly brackets and I need to keep entering and existing maths mode in a really confusing way.

@PauloCereda ;) Caerdydd.... There's a Cardiff in Santa Cruz, CA. But no Caerdydd that I know of ;).

@cfr I see. I guess for me those don't make much difference. I actually like the . since it makes what the label is clear. Also you don't need spaced before closing brackets, you just need a space between a terminal and a closing bracket. So something like [.A [.B [.C D ]]] so long as there's a space after 'D'.

@cfr As Jason Zentz noted, the problem with forest for most of us is that it needs some serious tweaking to make trees look the way they typically are for linguistics, and so as a first recommendation to students, tikz-qtree is simpler.

@cfr Oh no! :)

@cfr: I looked at the pronounciation and someone has written this:

The Welsh name for Cardiff is Caerdydd, pronounced to rhyme with "fire breathe".

@AlanMunn Actually, it needs serious tweaking for proof trees, too. But qtree needed major surgery and I only ever managed to approximate what I needed - so that it was at least obvious what was intended, even though it wasn't quite right. With forest, I can get the right result and doing so is much less hair-raising than approximating that result with qtree. I've never used tikz-qtree so I can't say.

@AlanMunn I am using for compiler construction. :)

@cfr No Caerdydd, but there is a Bryn Mawr, Upper and Lower Gwynedd. Not to mention a few Bangors.

@PauloCereda The second bit is right - dydd rhymes with breathe. But caer doesn't rhyme with fire!

@cfr ooh

When I visit everybody in the UK, I will learn Welsh too!

@AlanMunn Yes. It is disconcerting. Maybe the difference is that those do not have Anglicised versions, whereas Caerdydd does?

@PauloCereda They'll be quite surprised in Edinburgh if you do!

@cfr Yes for sure it's a fantastic package. And qtree was good while it lasted, but tikz-qtree is miles better. So for most of my simple trees I'm likely to stick with that. I haven't quite decided. For fancy trees, forest is great for sure. And sometimes the compaction is useful too.

@cfr ooh bagpipes! :)

Paulo dwi!

@AlanMunn I'm not actually using any of them for logic at the moment because I'm not teaching it. I would probably either stick with qtree if using my existing examples (not likely as they were for seminars for someone else's course) or use forest, just because it would be easier than learning a new package. But I probably wouldn't teach proof trees anyway ;). I usually use Bergmann, Moor and Nelson (but forget how to spell their names).

@PauloCereda Ie?

@cfr Uh-oh, I'm lost. :)

@AlanMunn This is also why I'd like to see packages providing some standard styles as wrappers for forest. I think that could work quite well, but perhaps I'm simply deluded.

@PauloCereda Trueni ;).

@cfr I don't know that book. I've only taught a semantics course once, and for that I used a really nice little Cambridge book Logic in Linguistics by Jens Allwood, plus a semantics text.

@cfr No I don't think you're deluded at all. But I think Sašo want's to implement them as styles you can load like biblatex styles.

@AlanMunn It uses a fairly standard natural deduction system, but it sets things up in a way which gives students quite a lot of guidance in how to go about finding proofs.

@AlanMunn For logical trees, at least, a wrapper works better. Otherwise people have to remember to provide a phantom root, which is a pain. But I can see the attraction of styles where that is an option.

@cfr Right that makes a lot of sense.

Question.

Why does the TARDIS spin clockwise sometimes, but sometimes counterclockwise?

:(

@PauloCereda It's an optical illusion. Like this one

@AlanMunn He is a witch!

:)

@PauloCereda Last time I saw it, it wasn't spinning at all. It was just sitting on top of Cardiff Castle.