@PauloCereda It is gone ;-) The universe is in balance again ...

@ChristianHupfer ooh astrophysicists. :)

@PauloCereda Will @HaraldHanche-Olsen count that 'ooh' too? ;-)

@ChristianHupfer ooh more counting

@PauloCereda Indeed, more oooooooooooooooooooooooooooooooooohs :D

@ChristianHupfer oh

@ChristianHupfer Fortunately, I don't need to. We have computers to take care of that for us.

@HaraldHanche-Olsen Norway seems to be a pretty modern country then ;-)

@HaraldHanche-Olsen ooh

@PauloCereda @HaraldHanche-Olsen I think there are 10000 ooohs from @PauloCereda in one single week alone ...

Friends, I posted on c.t.t, but just in case: if any of you wants to translate arara to your language, please poke me!

@ChristianHupfer ooh statistics

@PauloCereda @DavidCarlisle will take care of ladin, I guess.

@mickep I might submit all strings to the Google Translate API first. :)

@PauloCereda I will provide the translation into the local German dialect spoken here --- no one will use, except of me, of course ;-)

@ChristianHupfer ooh please do it

[paulo@cambridge target] $ java -jar arara.jar test.tex -L qn
  __ _ _ __ __ _ _ __ __ _
 / _` | '__/ _` | '__/ _` |
| (_| | | | (_| | | | (_| |
 \__,_|_|  \__,_|_|  \__,_|

Hold yew hard, ole partner, I'm gornta hev a look at 'test.tex'
(thass 27 bytes big, that is, and that was last chearnged on
04/03/2018 05:41:38 in case yew dunt remember).

(PDFTeX) PDFTeX engine .........   THASS A MASTERLY JOB, MY BEWTY

Wuh that took 0.63 seconds but if thass a slight longer than you
expected, dunt yew go mobbing me abowt it cors that ent my fault.

@mickep ^^ an Easter egg from Nicola :)

@PauloCereda ooh!

@mickep ooh surprises

@ChristianHupfer I doubt the Stackexchange computers are located in Norway. (Here's the count, by the way.) But we are connected to the Internet! In fact, I first had access to arpanet back in 1984 or thereabout. Oh wait, that count includes others who said ooh and mentioned @PauloCereda at the same time. Not sure how to avoid that.

@HaraldHanche-Olsen Ok, I was ten years old at that time ;-)

@HaraldHanche-Olsen ooh a conundrum

@PauloCereda English as spoken by Pirates? ;-)

@ChristianHupfer and a bottle of rum

@PauloCereda and wooden legs....

is it well documented that math type after \overbrace might not be what you expected

say, what was the april fool's joke this year on TX.sx and on CTAN.org?

kinda missed it, I guess

@daleif No, not well documented; \overbrace{...}^{...} makes an Op atom.

@daleif Always better to brace the whole thing.

@naphaneal I finished my thesis a duck on the lower corner. :)

@egreg I'm messing with my own version of \overbrace so I'll just grap the ^ as well and wrap it in a \mathord. BTW: do we have a method of getting the class (or what ever it is called) at a given point?

@daleif Not really

@PauloCereda really? I thought, that was a new permanent feature for the site. I kinda like it.

guess you fooled me there

@naphaneal I am not allowed to comnent, people strongly believe I am somehow involved. :)

@PauloCereda well, you're the resident duck expert, so...the assumption is close

@naphaneal exactly. <3

@egreg @daleif Because \overbrace uses \limits, which requires an Op. So it makes some sense that it has to be that way.

@HaraldHanche-Olsen ahh, nice to know

@LoopSpace Sorry for the long delay! I did a few tests with the new tikz ducks library and as far as I can see it works perfectly! I just merged it in the master branch and added a short passage in the documentation.

For the moment I stashed the addition of transform shape, as I'm currently thinking about different possibilities to automatically fit the text into the available space. I also move the super duck from the library into the example file.

@samcarter ooh secret identity

@daleif have a look at bm (it spends most of its time trying to figure out the class so it can re-instate it later (it is relatively painful:-)

@DavidCarlisle ooh the author must be very clever

@PauloCereda I doubt that

@DavidCarlisle argh, I'll just make a mental note.... The minimal width of the overbrace, that is limited by the design of the font?

@daleif yes (you could in a macro check by hand and for small enough arguments use a scriptsize overbrace, perhaps...)

:43772568, hmm, that might be interesting. In my current project we have some small arguments, I'll see what we can come up with

@egreg @daleif I have reported this before, but the response from Karl is that it is not enough a bug.

@yo' @daleif In my opinion it's bad syntax; it ought to be \overbrace{formula}{top} to begin with; using ^ (or _ for \underbrace) doesn't really make much sense.

@egreg well, it's inherited from plain (where the bug obviously happens too). And DEK doesn't like the arguments thingy.

But probably Knuth took the idea from something already existing, like for \over.

One could always do something like this:

\newcommand\overbraced[2]{{\overbrace{#1}^{#2}}}

@HaraldHanche-Olsen Yes. And also something more in order to cope with different math styles.

@egreg I noticed that \overbrace explicitly sets its argument in \displaystyle, yes. Which is understandable, sort of. Who would want to use it elsewhere?

@HaraldHanche-Olsen well, that's more complicated I'd say. Certain thing is, you shall check the result and adjust as needed.

@yo' Not sure I understand what you mean by that. What result? What adjustments?

@HaraldHanche-Olsen well, \displaystyle anywhere automatically is doomed to be ugly, sooner or later.

@yo' It's easy to override the style. To automate it is a job for \mathpalette og \mathchoice. (Probably the latter, as you also need to deal with the second argument.)

@HaraldHanche-Olsen and I say that neither will do a good job. You will need manual adjustments anyway, like you sometimes need \dfrac, \tfrac and {dcases} and stuff.

So the point is, why bother so much?

@yo' I wouldn't. I think my \overbraced (and a similar \underbraced) is more than sufficient. As I said, it's easy to override: \overbraced{\textstyle…}{…} for example.

@yo' I think I misread you. I thought you called for more automation, but I take it you don't.

@yo' everywhere yes but by the time you have \overbrace it's probably not going to fit inline anyway so using \displaystyle is understandable (especially in original code when the time and memory overhead of \mathchoice was not to be neglected)

@DavidCarlisle this is what I advocate for: don't overthink it, let the user do what's necessary :)

@yo' such naive faith you have in users:-)

@DavidCarlisle I would not call it "such faith", I simply have less faith in automation :-)

Does someone know if accsupp-luatex.def has been excluded from the .tds.zip of the oberdiek bundle for some reason (missing in TL18 pretest too)?

@TeXnician put it down to incompetence

@yo' Well, given that it typesets its argument in a box, you can't get around picking a style. So the choices are between picking a fixed one and automating it. Given the former, using \displaystyle is clearly the better choice.

@HaraldHanche-Olsen you could use \mathchoice to re-use the current style but that has other costs

@TeXnician I'll fix, thanks for the report

@DavidCarlisle Yes, I think you made that point already.

@DavidCarlisle You're welcome and thanks for fixing.

@TeXnician well since I broke it, it seems only fair

@DavidCarlisle Wow, one of the rare cases when we can blame you ;)

@TeXnician fairer to blame Heiko:-) I had to reverse engineer a build script from the 100s of files in the collection and then didn't notice that adding a new option in one place didn't push the new file into the final tds zip:-)

@DavidCarlisle :D

@egreg @HaraldHanche-Olsen we could add Haralds version to mathtools with an explanation as to why it is better.

@HaraldHanche-Olsen @daleif See this

I currently use my own version where we control the height of the inner formula to get equal height where there are more overbraces on a line.

\documentclass{article}
\usepackage{amsmath}

\makeatletter
\newcommand{\overbraced}[2][]{%
  {\mathpalette\over@braced{{#1}{#2}}}%
}
\newcommand\over@braced[2]{\over@@braced#1#2}
\newcommand\over@@braced[3]{%
  #1\mathop{
    \vbox{
      \m@th
      \ialign{%
        ##\crcr
        \noalign{\kern3\p@}%
        \downbracefill@{#1}\crcr
        \noalign{\kern3\p@\nointerlineskip}%
        $\hfil#1{#3}\hfil$\crcr
      }%
    }%
  }%
  \limits^{#2}%
}
\newcommand\downbracefill@[1]{$\m@th#1 \setbox\z@\hbox{$#1\braceld$}%

@egreg does that actually fix the type of the construction isn't this still a mathop?

@daleif No, this is always a mathord

@egreg Ah , of course we don't need to worry about the style of the exponent (as I stated above). It gets picked up by the standard mechanism.

@HaraldHanche-Olsen There's a #1 before \mathop just for that.

Couldn't we just wrap the whole thing in a mathord or would that break something? The @@ macro seems quite self contained

@HaraldHanche-Olsen @daleif Of course amsmath is required as it provides scalable cmex. With lmodern it would be horrible.

@percusse ^^

@egreg But isn't that superfluous? It merely select the current math style where that style is already in effect. Or did I miss something?

@HaraldHanche-Olsen We're inside \mathpalette

@HaraldHanche-Olsen Yes, it's not necessary.

@daleif It seems to me that someone who cares enough about performance to use the @@ macro can be trusted to do that for themselves. That said, I see no reason why moving the extra braces to there from \overbraced should break anything.

@PauloCereda I feel you need this :D

phdcomics.com/comics/archive.php?comicid=1351

@yo' unlike me, you are mean to @PauloCereda

@TeXnician an hour in and "oberdiek" bundle is still re-making (I hope it's right this time:-)

@daleif -- you've got mail.

@barbarabeeton so have you:-)

@DavidCarlisle -- yes, got 'em both, and thanks. (am now sorting out what's been done and updating the bugs list to say that these items have been taken care of. it's a rather long list ...)

@barbarabeeton by the way my \fractionseithdelims comment for amsxtra was a false alarm the code is skipped anyway if amsmath is loaded so it's not an issue, th eonly way to not load amsmath is to force amstex.sty to load and then you are really asking for old code so you get what you get

I think I need to try running four luatex in parallel, this is taking tooo long.

@DavidCarlisle -- indeed. nobody should be using amstex.sty unless the document was written before 1995. i hope you fixed your answer.

@barbarabeeton I hadn't added any text about that so your existing text (which doesn't mention anything about fractions) is fine

I'll read it tomorrow, I don't read my work email when I'm not at work 😎 hopefully I'll get time to read the other one as well

@daleif -- it isn't "work" mail. it's about the updated 'amsmath' documentation for tex live. (ask @DavidCarlisle about deadline.)

@barbarabeeton I already guessed.

How long should we wait to close questions in which OP doesn't provide an MWE after he was asked to do so? And for which reason do we close? Unclear what's been asked? Speaking e.g. of this:

@Skillmon longer than that:-)

Hey. I want to be able to ask a slightly more informed question on the main site. I got an error "Undefined control sequence" on /end{document}, where should I start digging?

@egreg Real Madrid played well, don't you think?

@kylergs the error message in the log tells you which command is undefined

@DavidCarlisle Quite decently

@DavidCarlisle \end apparently - it just seems unlikel

@kylergs no, what is the full multi-line error message from the log file (ignore any "summary" from your editor

@kylergs for example this generates an undefined command error in \end{document}

\documentclass{article}

\AtEndDocument{\zzzz}

\begin{document}

\end{document}

the full error message, from 1 to ? is

! Undefined control sequence.
\@enddocumenthook ->\zzzz

l.7 \end{document}

?

@kylergs the last command on teh second line

@kylergs OK so that tells you the undefined command is \nat so then it's just a matter of working out what used that command....

@DavidCarlisle Yeah I see that now, given a bit of trial and error I think it's \printbibiography

@kylergs in which case it's presumably an error in one of the entries in your bib file

@kylergs do you use texstudio?

@DavidCarlisle No, just a text editor and latexmk

@DavidCarlisle the .blg isn't showing any errors

@kylergs it wouldn't. the content of the entries is just opaque text to bibtex and biber but if there are tex errors in it you get errors when it's included

@kylergs the texstudio question was just because there is a long standing feature request for it not to mangle undefined command errors: sourceforge.net/p/texstudio/feature-requests/1068

@DavidCarlisle Found it! The reference I downloaded had journal = {\nat} for some reason.....

@kylergs :-)

@DavidCarlisle Thanks for the help!

Is there a way to clear LaTeX memory like one would do in the about:memory of a Firefox browser? I think that by loading loading a big-ish table several times so different domains get graphs fills the thing to the brim.

Alternatively, I could probably graph all of them individually into .eps files and input them in the final..

@vlg almost all relevant memory is cleared after each page

@vlg what error do you get (the error message shows you which of the stacks you have overfilled if you run out of space)

I made it so that three graphs fit one page, the TeX capacity is exceeded at the sixth?

Invites me to run a wizard to increase memory, too

@vlg hard to tell with so little information, but try putting \clearpage after the third

Yeah, doing that currently, you gave the necessary info

@vlg no not run a wizard, ask a wizard (he meant a human not some GUI help app:-)

Haven't seen one of those around recently

@vlg they are invisible

You're functionally invisible and a wizard in my eyes :p

@vlg I'm monochrome not invisible

@DavidCarlisle :D

@marmot You seem to be German. Are you on texwelt.de?

@HenriMenke No, I have never visited texwelt.de.

@marmot Maybe Till Tantau is not a “Zauberer”, but perhaps he owns a “Zauberflöte”.

@marmot Any contributions on TeXwelt are welcome!

@marmot I saw your message about the particle physics example on TeXample.net, about being offline, but it's online (the one I know)

@HenriMenke Btw. thank you for your contributions on TeXwelt.de! I read all. I subscribed to get an email for each post.

@HenriMenke @StefanKottwitz I just had a look and it seems to be a very similar crowd to here except for some Italian's and Englishmen who argue about pineapple and pizza ;-)

@StefanKottwitz The particle chart I saw on TeXample. Is this the same as texwelt?

@marmot It's just to enjoy the same TeX talk but in native German, easily digestible for students here. Local language means lower barriers and maybe more TeX friends.

@marmot Über Cricket könnten wir auch auf Deutsch streiten

@marmot It was always here: Standard model of physics

@StefanKottwitz Hmmh, but I could swear that a few days ago it wasn't. ;-)

@DavidCarlisle Cricket? What is that? Can one eat it? ;-)

@marmot It could be my fault. I hope I did not break something. I only added 10 examples.

@marmot Es ist alles hier erklärt cricket.de

OMG sorry, it's not my fault that the TeX.SE chat expands a tweet links with all included images

But it's about expansion, obviously.

@DavidCarlisle Cricket? What is that? Can one eat it? ;-)

@egreg Mille grazie for the edit!

@marmot Chiaramente non capisci il tedesco, proviamo un'altra lingua

@marmot You're welcome

@StefanKottwitz \begin{advertizing} doorcode \end{advertizing}

@StefanKottwitz When I click on the last link by Marcus Rentsch, I get to the Bernoulli picture. Just wondering if there is a copy&paste issue...

@marmot Wow! Asymptote is amazing. You know TeXample - I wanted to re-use the web code for an Asymptote gallery ( asymptote.info ) but figured that I would not have the time, alone on my own.

@StefanKottwitz I can imagine it is. Vielen Dank für all die Zeit, die Du da reingesteckt hast!

@marmot Oh! Yes, a c&p issue. I fixed it. Thanks!

seems to be a lonely place ;-)

@marmot More "tex" matches than "murmeltier" matches

@StefanKottwitz The number of "Murmeltier" matches increased by 1.

@marmot Great! There's a lot of TikZ ongoing too. Maybe you see an occasion to answer: texwelt.de/wissen/themen/tikz

Many times Elke already helps out, or Henri posts (but then with ConTeXt >;-) )