TeX, LaTeX and Friends - 2018-09-13 (page 1 of 2)

« first day (2874 days earlier) ← previous day next day → last day (2356 days later) »

barbara beeton

01:41

@FaheemMitha -- a few people still use amstex, but if a manuscript is submitted to ams in amstex, it is rekeyed in latex according to ams style. latex won because so many people had adopted it for the automatic numbering and cross-referencing, and weren't willing to give that up. math could still be typeset without amstex, although painfully in many situations.

@Krishna -- there's a very simple explanation -- we are in many different time zones. i'm in a time zone six hours later than europe, and i know several participants who are three hours later than i am. never mind the handful of users in places like australia, which isn't close to anywhere else.

-- here's what it says on the tug website on the "board" list about spivak's participation: Spivak, Michael
Board member, 1980-1985
Acting Chairman, 1981-1983
Finance Committee, 1982-1983

@FaheemMitha ^^^^^ this was meant for you.

manooooh

02:42

@DavidCarlisle lol! Did you know that I actually write ${(x+y)}^2$ ? I noticed the 2 is a little above than without braces, but I thought the correct way to use superscripts and/or subscripts was adding braces, even if I had to write something like ${\left({(x+y)}^2\right)}^2}...$ . Even in 2018 should I leave this notation? Thanks!

@DavidCarlisle (but I don't use it in, for example, $x\in{\mathbb R\wedge...}$ , I don't use braces there, but I use it in $\lim{f(x)}$ because adds some extra space that I think is necessary for this type of math operator. Also in $\int{...}$ , $\sum{...}$ , etc. In these cases is that correct? I think yes, because there is no visual problem, but you know more than me :))

CarLaTeX

03:13

@Sebastiano Ciao! Happy TeXing!

CarLaTeX

03:25

@Krishna Someone else asked this before: tex.meta.stackexchange.com/questions/6203/do-you-have-a-life

manooooh

03:38

@CarLaTeX "Do you have a life?" sounds like a phrase from a movie. Oh wait, some people spends hours here. Is that question necessary? No? Hmmm... I don't know....

1 hour later…

CarLaTeX

04:45

@manooooh I don't remember a movie with that citation but it's very suitable here, lol!

2 hours later…

boycott.se - yo'

06:28

@manooooh well, there is basically always a visual problem with braces in math, as it kills space stretchability (I hope I remember this correctly)

Faheem Mitha

06:39

@barbarabeeton Thank you.

@barbarabeeton Could automatic numbering and cross-referencing not be added to amstex, then?

David Carlisle

06:56

@manooooh you shouldn't have {} or \left\right there, just (x+y)^2

@manooooh $\lim{f(x)}$ because adds some extra space the {} do nothing there, f is already a mathord. \sum{...}$, there should be no braces there, you make the following thing a mathord , and box it so that the white space can not be shrunk or stretched to help line breaking.

@FaheemMitha that's what lamstex did, but it was a decade too late, and no one used it. (@barbarabeeton)

Faheem Mitha

@DavidCarlisle I see.

3 hours later…

manooooh

10:24

@DavidCarlisle you're right... Take for example $\lim{(1)}$ vs. $\lim (1)$ . The first one adds some extra space (the correct I would say), when the last one it doesn't. I don't care line breaks, I'm asking for the correct way to use math operators

@boycott.se-yo' lol yes

@CarLaTeX :((

Krishna

10:46

@CarLaTeX About the sleep thingy, I thought adults needed atleast 7 hours and children about 9 hours of sleep. I cannot function the next day without sleep. I was genuinely surprised when I saw a few folks zealously helping newcomers here without much thoughts about their personal health

David Carlisle

@manooooh the extra space is so you can go \sin x and get a small separation it isn't needed and not added if you go \sin(x) so \sin{(x)} just means that you are fighting TeX's rules for math layout. that isn't necessarily wrong it is your document and you can make up the rules but certainly the extra braces are not intended, and not generally correct markup there.

Krishna

11:13

Frank Mittelbach just uploaded an interesting package to ctan.....widows and orphans

looks useful

David Carlisle

@Krishna we've been keeping ctan busy this week:-)

Krishna

Hmm... I noticed, latex-bin was upgraded in TL

On a cursory glance, Raphael Pinson's nowidow package seems to provide similar functionality.

David Carlisle

@Krishna yes i saw that in cygwin too, not sure what changed there.

Krishna

@DavidCarlisle Do you know how Frank's package compares to nowidow?

David Carlisle

@Krishna not really,

Krishna

11:20

I am of course trying to impress my viva examiners

@DavidCarlisle Ok. I thought they both affected clubpenalty and such values

Maybe I can use both packages in my thesis :)

David Carlisle

@Krishna not tried either but nowidow is basically a wrapper for setting 10000 penalties to prevent widows, the new package has some options for setting those but the main point is that it warns in the log when there are widows so you can fix them by re-writing or whatever. setting the club and widow penalties rarely does the right thing.

Krishna

I see.

So would you recommend not loading nowidow ?

David Carlisle

3

A: Compromise between too large whitespaces and `\raggedbottom`?

Using \clubpenalty10000 and \widowpenalty10000 is usually not the best way to avoid widow and club lines, especially with \flushbottom as they almost always make the constraints on page makeup impossible to achieve. A typical example is a paragraph in which a last line has a couple of words whic...

@Krishna I guess that was you leaving an issue on Frank's page?

Krishna

Yes. That was me :)

"although in luatex one could consider using the line and page breaking callbacks to automate this." - I am interested in this

How would one go about doing this in luatex ?

David Carlisle

@Krishna with difficulty:-) but in principle you can replace the entire tex line breaking algorithm by specifyng a lua function that does the line breaking, but you have to come up with a better algorithm and code it in lua....

Krishna

11:33

@DavidCarlisle hehe.... alright. So not happening for my thesis :)

Phelype Oleinik

Good morning everyone :)

So I updated TeXLive and one of the new packages has the following (quite curious) code:

      \begin{doublespace}
        \textbf{\small The \@doctype{} committee for \@author{} certifies that this is the approved
        version of the following \MakeLowercase\@doctype:}

        \hbox

        \hbox

        \textbf{\headingsize\@title}

        \hbox

        \hbox
      \end{doublespace}

What is this supposed to do? And more importantly, why doesn't it produce an error?

David Carlisle

@PhelypeOleinik why should it error? Oh you mean the two \hbox ? That can't work!

Phelype Oleinik

@DavidCarlisle I tried to use an "empty" \hbox like that and got a ! Missing { inserted...

David Carlisle

@PhelypeOleinik yes sorry I was looking at \textbf{\headingsize\@title} which isn't an error, just poor style and didn't notice the hbox:-)

Phelype Oleinik

@DavidCarlisle :)

user 2646

11:43

to err is human

David Carlisle

@PhelypeOleinik presumably no one in Texas has passed, so the class hasn't been used yet.

Phelype Oleinik

@DavidCarlisle The thing is, the code apparently runs without problems:

\documentclass[masters]{utexasthesis}

\author{Me}
\title{hbox}
\supervisor{That guy}
\graduationdate{probably}{never}

\begin{document}

\maketitle

\end{document}

David Carlisle

@PhelypeOleinik give me a minute...

Phelype Oleinik

@DavidCarlisle Sure :)

David Carlisle

11:58

\documentclass{article}
\usepackage{setspace}
\begin{document}

\begin{center}
\begin{doublespace}

zzz

\hbox

\end{doublespace}
\end{center}



\end{document}

@PhelypeOleinik ^^

@PhelypeOleinik AAAAAAAAAAAAAAAAAAAAAAAAAARRRRRRRRRRGGGGGGGGGGGGGGGGGGGGG

@PhelypeOleinik \par is:

> \par=macro:
->\if@newlist \advance \par@deathcycles \@ne \ifnum \par@deathcycles >\@m \@noi
temerr {\@@par }\fi \else {\@@par }\fi .

@PhelypeOleinik which means \hbox followd by a blank line ends up as \hbox{\@@par}

Phelype Oleinik

@DavidCarlisle Ooooooooooo(many more)ooohhh

@DavidCarlisle Works by accident :)

David Carlisle

@PhelypeOleinik as long as you define "work" with care.

Phelype Oleinik

@DavidCarlisle LOL

@DavidCarlisle "Runs without having to type x in the console to exit" is more appropriate

David Carlisle

@PhelypeOleinik yes

@PhelypeOleinik which means doublespace wasn't involved and you can do

\documentclass{article}

\begin{document}

\begin{center}
x

\hbox


\end{center}

\end{document}

Phelype Oleinik

@DavidCarlisle As long as you define "can" with care.

David Carlisle

12:05

but you need that x

Phelype Oleinik

@DavidCarlisle If not in the console, in the source at least XD

daleif

12:30

Any advise on how to debug strange spaces in an area where there are fills?

user image

I have no idea where the space before the comma after Oldstyle 5 comes from, so is trying to debug.

David Carlisle

@daleif start with \showoutput and work back...

@daleif usual suspect is missing % but \showoutput will show that. as you'll see if that is a single space or multiple spaces of ~.3em....

daleif

@DavidCarlisle as far as I can see I've ended every line in the generating code by %. I'll try showoutput (had forgotten the name)

Hmm, it seems to be this one

.....\glue 5.0
.....\pdfstartlink(*+*)x* attr{/Border[0 0 0]/H/I/C[1 0 0]} action goto name{su
bsection.1.3.3}
.....\pdfcolorstack 0 push {0 0.2 0.6 rg 0 0.2 0.6 RG}
.....\T1/ppl/m/it/10 1
.....\T1/ppl/m/it/10 .
.....\T1/ppl/m/it/10 3
.....\T1/ppl/m/it/10 .
.....\T1/ppl/m/it/10 3
.....\penalty 10000
.....\glue 2.5 plus 1.49998 minus 0.59998
.....\T1/ppl/m/it/10 O
.....\T1/ppl/m/it/10 l
.....\T1/ppl/m/it/10 d
.....\discretionary
......\T1/ppl/m/it/10 -
.....\T1/ppl/m/it/10 s
.....\T1/ppl/m/it/10 t

David Carlisle

that is spaces:-) three 2.5pt glue will be three space tokens, almost certainly

daleif

12:45

Note that the other entries does not have this 3x \glue 2.5 plus 1.49998 minus 0.59998

Hmm, just found others. These are online entries in a toc. This is not at the end of a line, it is in the middle of the line. I'm guessing that the method for adding the comma is not working as intended

Skillmon

Is reading in an argument as a \vbox with an added \noindent good practise or should I do something like LaTeX does after its headings?

David Carlisle

@Skillmon the latter: if you add a \noindent it makes a blank line at the start of the users argument very weird

Skillmon

13:08

@DavidCarlisle But who starts a braced argument with an empty line?

boycott.se - yo'

@DavidCarlisle if you add a \noindent it ~~makes a blank line at the start of the users argument very weird~~ constitutes a deadly sin

David Carlisle

@Skillmon if it's a paragraph level thing, why not?

Skillmon

@DavidCarlisle it's \ducksay :)

David Carlisle

@Skillmon well users will get what they deserve then:-)

@Skillmon but I meant if the argument is a paragraph, it's not unnatural to set it off with a blank line

Skillmon

@DavidCarlisle is \@afterheading the correct macro for that?

David Carlisle

13:15

@Skillmon yes (or the similar but more or less nameless code in \item)

@Skillmon can;t you just set \parindent=0pt?

Skillmon

@DavidCarlisle no, that doesn't seem like a good idea. What if somebody wants to use multiple paragraphs in the argument and does like the second one to be indented?

@DavidCarlisle I used \@afterindentfalse\@afterheading now. I think that's likely to be the best solution

@DavidCarlisle but it does also disable the easy indentation of \indent :( The world is a cruel place.

David Carlisle

@Skillmon no user should use \indent in a document.

@PhelypeOleinik @egreg has complained upstream github.com/linguistics/utexas-latex/issues/2

Skillmon

13:32

@DavidCarlisle why not?

@DavidCarlisle and how would you force the indentation of the first paragraph then (I mean in the argument of say \ducksay)?

David Carlisle

@Skillmon because in almost every place where you might expect it to work latex uses \@afterheading or similar so it doesn't work so it is not documented as a latex command in the latex book, it's just a tex primitive that you are not supposed to know about.

Skillmon

@DavidCarlisle I'm very sorry, but I know of it :) And this argumentation seems very logical to me.

David Carlisle

@Skillmon it simply shouldn't be a per-paragraph decision, if you want to over-ride the class decision and indent initial paragraphs the excellent indentfirst package will do that

Skillmon

@DavidCarlisle but how would you indent the first paragraph in a macro which uses \@afterheading like my \ducksay if you want the first paragraph be indented for reason?

@DavidCarlisle \hskip\parindent seems clunky.

David Carlisle

@Skillmon indentfirst package or a package-level option that just did it for ducksay and not normal headings, certainly I would not want to add markup in the document on each paragraph to over-ride it at that point

samcarter

13:37

Dear l3 experts: Are there any guidelines how warnings in l3 should look like? Will they all start like the following with * xxx warning:?

*************************************************
* widows-and-orphans warning: "orphan-widow"
*
* Orphan on page 1 (second column) and widow on page 2 (first column)
*************************************************

David Carlisle

@Skillmon \indent\indent works as well but probably you are not supposed to know that either:-)

Phelype Oleinik

@DavidCarlisle The simplicity in the name of the issue :)

Skillmon

@DavidCarlisle I'd never use it, \hskip\parindent would be way clearer in that situation.

egreg

@samcarter That's how they currently show; not that I really like it.

Skillmon

@egreg you love it instead?

Phelype Oleinik

13:39

@DavidCarlisle I just fear the day that I upload the class I wrote to CTAN and the dubious code that will be found in it :P

Skillmon

@PhelypeOleinik that's why we don't upload dubious classes we write for some institutes to CTAN!

Phelype Oleinik

@Skillmon Oi, mine isn't supposed to be dubious! :P

Skillmon

@PhelypeOleinik Or does your TeX-Distro include MRTthesis?

samcarter

@egreg Thanks, that's all the info I needed :) Some of the currently available log parsers don't like this either :)

Phelype Oleinik

@Skillmon MRTthesis? I didn't find that one...

Skillmon

13:42

@PhelypeOleinik because I didn't upload it.

@PhelypeOleinik it's here: gitlass.de/jonathan/MRTbundle

Phelype Oleinik

@Skillmon Makes sense :) I'm curious about that now

@Skillmon Ooh

barbara beeton

@manooooh -- here's what wikipedia says about spivak regarding the awarding of the steele prize: "1985 Michael Spivak for his five-volume set, "A Comprehensive Introduction to Differential Geometry" (second edition, Publish or Perish, 1979)." so the first edition was out well before tex was even a concept in dek's mind. spivak's cogent writing style is why he was asked to write the documentation for the math interface to tex.

Skillmon

@PhelypeOleinik it's not completely finished as it is not yet feature complete in comparison to its predecessor but pretty usable.

Phelype Oleinik

@Skillmon It's even written with expl3. Advanced!

Skillmon

@PhelypeOleinik not all of it. It's a pretty wild mix :)

Phelype Oleinik

13:45

@Skillmon I had no idea how to use expl3 half an year ago (not that I do now), so mine is written in a wild mix of plain and LaTeX.

Skillmon

@PhelypeOleinik and I'm going to steal and build upon your auto-frontmatter approach (but don't use it by default).

Phelype Oleinik

@Skillmon I still wonder if that was a good idea, but I won't change this now... But suit yourself :)

Skillmon

@PhelypeOleinik I intended to make it configurable and as I said don't use it by default, so that one still has the freedom to create the front matter very by one's self.

@PhelypeOleinik But that has low priority. The main issue which is still to be solved (and will then finally lead to being feature-complete) is the longtable option for MRTtab. Hacking longtable might turn out to be really hard though (I didn't start to seriously look into it) :( I hope @DavidCarlisle finishes the next iteration of it before I eventually tackle that problem, so I don't have to dig too deep :)

Phelype Oleinik

@Skillmon What do you use longtable for?

David Carlisle

14:05

@Skillmon well current rate of progress is a release every 20 years, when do you need the next release by?

manooooh

14:25

@DavidCarlisle but I think we have to be homogeneus in our notation. I noticed the following: some people often use ]...[ to write an open interval in math, right? So take this example: $=]1,2[$ ; the ] is sticking with =, and that's bad. How do I solve this problem? Adding braces! e.g. $={]1,2[}$ . This is the correct way I think, like the case where the limit is sticked with (

We have to add braces everywhere!!

David Carlisle

@manooooh no that's wrong

manooooh

@DavidCarlisle and how would you solve $=]1,2[$ ?

@barbarabeeton thanks for your time! I thought the book was released a few years later than LaTeX release, that's incredible. So now the question is, does Spivak wrote in PC his book? Did he use amstex or did he write by hand?

Skillmon

@DavidCarlisle next month or so :) Gonna hack into the existing code. Might get really ugly, I'll see.

@PhelypeOleinik I wrote a table-like environment which you can change to being longtable with a single option-key. That's great for big tables in the appendix. The tables get a uniform look with auto setting of captions, preamble, head-row and so on.

David Carlisle

@manooooh using {} makes things a mathord (like a letter) so you lose all spacing, so what you want is ] that is a mathopen so \mathopen{]}1,2\mathclose{[} but you don't want to type that all the time so you should declare something like \brakopen and \brackclose that have the right math class or (better) use mathtools \DeclarePairedDelimiter`

barbara beeton

@manooooh -- i really don't know what spivak's writing habits were when he was writing "calculus". since amstex didn't start to be created until 1979, it wouldn't have been available. (actually, spivak was originally only supposed to write the manual; as it turned out, he ended up writing the tex code too.) i suspect he might have used a typewriter in time-honored fashion, and used scissors and tape to rearrange parts of the text for the first edition.

David Carlisle

14:43

@manooooh no, almost nowhere (other than to delimit arguments of commands like \frac)!

boycott.se - yo'

@manooooh \mathopen{]}1,2\mathclose{[}

David Carlisle

@boycott.se-yo' what I just wrote:-)

boycott.se - yo'

@DavidCarlisle oh dammit :-) Actually, what you want is not sticking to French typography :-)

David Carlisle

@boycott.se-yo' well at leat we we both right:-)

boycott.se - yo'

and I would suggest \lopen and \ropen for \mathopen{]} and \mathclose{[}, eventually also \lclose and \rclose for [ and ]

@manooooh ^^

Phelype Oleinik

14:51

@Skillmon Ooh, I found it now. Looks interesting. I haven't defined appendices yet, but when I do I might as well take something from your class too :D

David Carlisle

@boycott.se-yo' Mais j'utilise toujours la typographie française

boycott.se - yo'

@DavidCarlisle I would guess "de la typographie", but I may be wrong

David Carlisle

@boycott.se-yo' you must be wrong, because google translate is never wrong.

Ulrike Fischer

@boycott.se-yo' No, imho David's sentence is fine.

boycott.se - yo'

@UlrikeFischer probably yes, I don't know. It was just a gut feeling without any linguistic reasoning.

Skillmon

15:06

@PhelypeOleinik don't. It tries to recreate bad typography as good as possible (it recreates existing Word templates)

Phelype Oleinik

@Skillmon "Interesting" the way it's done, not what it achieves. At least mine I could start from scratch, so it didn't turn out to be, as Alan Munn said, YAABNT :P

daleif

@DavidCarlisle hmm, seems to be caused by hyperref!!! At least if I remove hyperref, then it works as it should in my MWE. Hmm, but if I remove hyperref in the real doc, then the problem persist. Back to debugging

Skillmon

@PhelypeOleinik what means YAABNT?

Phelype Oleinik

Sep 8 at 19:41, by Alan Munn

@PhelypeOleinik Is this YAABNT template? (Yet Another ABNT template)?

@Skillmon ↑↑

Skillmon

@PhelypeOleinik don't know whether my code is good enough to be copied :)

Phelype Oleinik

15:17

@Skillmon Not copied. I prefer to write my own code when I understand what it does. I'm not good having ideas though, so I might search for inspiration ;)

Skillmon

@PhelypeOleinik notify me if you find something which should be changed :)

Phelype Oleinik

@Skillmon Sure thing! :D

Harald Hanche-Olsen

@manooooh There is also Lars Madsen's interval package. Personally, I think I prefer what @DavidCarlisle suggests, though.

Skillmon

Have a look at the output of the (unfinished) second iteration of the ducksay package:

user image

3

daleif

15:36

@HaraldHanche-Olsen it becomes cumbersome if you need to change the setup later on. interval is easier, especially also because of the color feature, then it is easier to see if you fixed the all.

Hmm, at least my spacing issue is not memoir related, now have a book based MWE with the same issue

David Carlisle

@daleif please don't report a hyperref bug:-)

@Skillmon nicely un-indented first paragraph:-)

Faheem Mitha

@barbarabeeton It doesn't look like Spivak was involved with TeX for very long. Not like some of you people.

CarLaTeX

43

Q: How to recover from TeX.SE (or, in general, TeX-LaTeX) addiction?

TeX.SE is addictive, we have recently had evidence. I don't think this is only due to the reputation mechanism. For example, the little games available on FB have a similar reward mechanism: you win points, and you can compete with your friends, but they don't give me any form of addiction (may...

samcarter

@boycott.se-yo' @barbarabeeton One of the very rare images of a haggis:

user image

2

The colour is inspired by ichef.bbci.co.uk/news/660/media/images/75904000/jpg/…

David Carlisle

@FaheemMitha started in 1979-80 or so and the ctan page for lamstex shows updates as late as 1991, most people would say a decade was a reasonably long time to be working on one software system

Faheem Mitha

15:49

@DavidCarlisle Was Spivak making the updates?

David Carlisle

@FaheemMitha I don't think anyone else ever understood the code so I assume so

Faheem Mitha

@DavidCarlisle Huh.

Given what I've seen of TeX code, that's a little terrifying.

David Carlisle

@FaheemMitha the internal coding structures are somewhat under documented. Nothing like the helpful source2e.tex.

Faheem Mitha

@DavidCarlisle Oh.

barbara beeton

@samcarter -- don't tell that to @DavidCarlisle!

daleif

15:51

@DavidCarlisle ohh, it is a hyperref bug. It is a sporadic space somewhere, in my setup I have subsection, subsubsection, subsubsection, subsection. The tocdepth include subsection. \l@subsection is defined to be an inline listing, adding comma using \ifhmode,\ \else\noindent\fi. Thus the space before the comma stems from the "empty" typesetting of the subsubsection entry.

David Carlisle

@samcarter but you didn't go for the "cut open and insides spilling out" look?

@barbarabeeton too late ^^

@daleif I just updated oberdiek's 90 something packages twice this week and started 19 new git repos for Robin yesterday, I was hoping to avoid hyperref for a bit:-)

barbara beeton

@DavidCarlisle -- there's no rest for the wicked.

David Carlisle

@barbarabeeton or apparently the virtuous

daleif

16:09

@DavidCarlisle Here is an MWE, I'm going to a wine and cheese tasting, so don't have more time now

\documentclass[oneside]{book}
\makeatletter
\renewcommand{\l@subsection}[2]{%
\ifnum\c@tocdepth > 1\relax%
A\ignorespaces%
\fi%
}
\makeatletter
\AtEndDocument{\addtocontents{toc}{\par}}%%% OK
\usepackage[colorlinks]{hyperref}
% test data
\newcommand\hmm{
\subsection{Test Test}
\kant[1]
\subsubsection{Test Test Test Test}
\kant[1]
}
\stepcounter{chapter}
\stepcounter{section}
\usepackage{kantlipsum}
\begin{document}

\tableofcontents

\bigskip

\noindent
The toc should show a list of A's with no spaces in between

The toc should show a list of As with no spaces, without hyperref it does not, with hyperref it does.

I don't have a github account.

Ulrike Fischer

@daleif if I did understand your description than you have this problem (without hyperref @DavidCarlisle !}

\documentclass{article}
\makeatletter
\setcounter{tocdepth}{2}
\renewcommand\l@subsection{\ifhmode,\ \else\noindent\fi}
\begin{document}
\tableofcontents
\section{xx}
\subsection{a}
\subsection{a}
\subsubsection{b}
\subsubsection{b}
\subsection{a}
\end{document}

@daleif in my example the source of the problem are the spaces in the toc after each contentsline entry.

David Carlisle

@UlrikeFischer this? \renewcommand\l@subsection[2]{\ifhmode,\ \else\noindent\fi#1{#2}\ignorespaces}

daleif

@UlrikeFischer try this with and without hyperref

\documentclass{article}
\usepackage{hyperref}
\makeatletter
\setcounter{tocdepth}{2}
\renewcommand\l@subsection[2]{A\ignorespaces}
\AtEndDocument{\addtocontents{toc}{\par}}%%% OK
\begin{document}

\tableofcontents

\newpage

\subsection{a}
\subsection{a}
\subsubsection{b}
\subsubsection{b}
\subsection{a}
\end{document}

Ulrike Fischer

@daleif This wouldn't do the same with hyperref. With hyperref \contentsline has one argument more.

@DavidCarlisle No doesn't work. \renewcommand\l@subsubsection[2]{\ignorespaces} works without hyperref in my example, and with [3] with hyperref. But @daleif's example seems to be more complicated.

daleif

16:25

@UlrikeFischer but we are not messing with \contentline here, we are messing with \l@.. which is called by \contentsline, also in the hyperref case.

Unless I'm reading the wrong place (quite possible) hyperref just calls \l@... with two arguments

and hyperref wrapped around the first argument

Ulrike Fischer

@daleif Hm. Yes. But in any case: contentsline has a different definition with hyperref and you can't add the \ignorespaces to l@subsection, there is too much stuff after it. Perhaps \addcontentsline should write a % to the toc, or \relax.

daleif

This means that hyperref is wrapping itself around the if tocdepth test in the \l@... macro, so what happens if that internal part does not typeset anything.

@UlrikeFischer you are of course correct, because the wrap on the un-typeset subsubsections we cannot directly see the space and get rid of it. It can probably be flagged as a feature in hyperref, not much we can do.

David Carlisle

\endlinechar=-1 when reading the toc? (@daleif @UlrikeFischer)

Ulrike Fischer

@DavidCarlisle Yes just tried. Seems to work. @daleif

daleif

Come to think of it, no it is not wrapping it self around the if test, it is inserted into the if test....

Ulrike Fischer

16:45

@daleif @DavidCarlisle this here seems to work too:

\usepackage{hyperref}
\usepackage{xpatch}
\makeatletter
\xpatchcmd\addcontentsline{{\thepage}{\@currentHref}}{{\thepage}{\@currentHref}\relax}{}{\fail}

barbara beeton

17:04

for anyone interested: the webcast of the 28th first annual ig nobel prize ceremony can be viewed here starting at 18:00 eastern u.s. time. (there's a pre-ceremony act starting at 17:40.) should be fun. (we'll be there watching in person.)

user280247

17:26

Is it necessary to remove software-center-latex before installing vanilla latex?

user280247

Should I ask: how bad is to start vanilla latex installation without removing older version?

Phelype Oleinik

@santimirandarp I have both TeXLive 2015 installed with apt and TeXLive 2018 installed with install-tl from tug.org and they both work without ptoblems...

user280247

Oh, wonderful. I know very little of linux but different versions arent replaced, they are on different directories isnt it? @PhelypeOleinik

Skillmon

@santimirandarp installing to a different folder and everything is fine. Also you should then set your $PATH accordingly to not mistakenly run the wrong executable.

@santimirandarp yes, by default TeXLive installs to /usr/local/texlive/<version>/.

user280247

@Skillmon thanks a lot!

Skillmon

17:32

@santimirandarp As I said, just pay attention on what is in your $PATH (a variable which specifies where the OS/your shell is looking for executables).

Phelype Oleinik

@santimirandarp Yes, on different directories. apt installs under /usr/share/texlive, while install-tl installed to /usr/local/texlive/2018, as Skillmon said.

user280247

@Skillmon yes, that was my last question. That's what the computer executes when typing ´latex´, isnt it?

Skillmon

@santimirandarp yes. Just use your favourite search engine to find out how to change/set it. There should be plenty of good answers.

user280247

Fine, thanks both.

Skillmon

@santimirandarp and pay attention to its format. You can very easily introduce security breaches when fiddling with it in the wrong way (the results you'll find should mention those risks)

user280247

17:36

@Skillmon So setting the path is important. That's all I can understand XD

Skillmon

@santimirandarp setting the PATH in the right way is important. Your OS should have reasonable defaults. You'd have to make sure that the binaries you want to be executed are in a folder which is earlier in your PATH (so if you want the executables of TeX Live 2018 to be executed when you type latex, /usr/local/texlive/2018/bin should be searched earlier than /usr/share/texlive).

user280247

@Skillmon got it...

Skillmon

@santimirandarp opensource.com/article/17/6/set-path-linux doesn't look to bad for an introduction to PATH

user280247

@Skillmon oh and it's short. Great.

Skillmon

@santimirandarp unix.stackexchange.com/questions/26047/… looks good, too.

@santimirandarp especially this answer mentions some security aspects (the first answer is great for explanations)

Skillmon

18:18

How bad would it be if ducksay would require the array package? If super bad: Any idea how I create new column types without it (or inject code at the beginning of every cell of a column without it)?

marmot

@samcarter Why is (s)he smiling? I hope it is not because (s)he just ate a duck. (Regardless of how hard I try, I cannot see a duck on the pic.;-)

samcarter

18:44

@marmot All ducks are safe! He (I'll call him Jack) is just happy because he found other clockwise Haggis to run around with.

marmot

@samcarter Now I am relieved. ;-)

samcarter

@marmot :)

marmot

user image

samcarter

@DavidCarlisle I wanted to avoid nasty stains on the plaid!

@marmot Oh, carrots! Yummy!

Skillmon

19:09

What is the height of an ordinary tabular row? It seems \arraystretch\baselineskip is not quite right.

Ulrike Fischer

@Skillmon inside of a tabular you need \normalbaselineskip.

Skillmon

@UlrikeFischer no, I'm still outside the tabular, I just want to predict how many rows I'll need.

manooooh

19:27

@DavidCarlisle sorry for my late response! That also works fine, but I don't understand why using {} we lose all spacing. Could you give a counterexample, please?

@barbarabeeton wow that sounds real! Over time their different editions have maintained the same format, especially the equations. It is surprising to see its great use to say things, use equations, create graphics, etc.

@DavidCarlisle no, almost everywhere! The most natural way to keep things spaced are {}; any other invention denatures the essence of LaTeX. Even TeX.SE uses braces for the logo!

@boycott.se-yo' that sounds good but I don't understand how to use those commands, thanks!

manooooh

19:54

user image

^^^^^^ strawberries with sugar!! 😋 😋

boycott.se - yo'

@manooooh well, for open intervals you would have \lopen 1, 2 \ropen, and for closed \lclosed 1, 2\rclosed, and for mixed one, you would mix these teo accordingly.

manooooh

@boycott.se-yo' ok, let me test it

@boycott.se-yo' it doesn't work ://. It says "Undefined control sequence." when I write $\lopen 1, 2 \ropen$ . I'm loading amsmath and amssymb packages. What am I missing?

boycott.se - yo'

20:11

You have to define the commands as I showed above (sorry I'm not at a conpurer and it's hard ti write down code)

manooooh

@boycott.se-yo' sorry but I can't find the commands definition here in chat :( (don't worry about that!!)

Skillmon

@manooooh

\newcommand*\lopen{\mathopen{]}} \newcommand*\ropen{\mathclose{[}} \newcommand*\lclosed{\mathopen{[}} \newcommand*\rclosed{\mathclose{]}}

I guess.

manooooh

@Skillmon ahh, as @DavidCarlisle said in his comment I think

Thank you :)

egreg

20:26

@Skillmon \mathopen] is better (similarly for the others)

Skillmon

@manooooh I think @boycott.se-yo' mentioned a package above, didn't he? Perhaps this package also contains definitions for those macros.

@egreg because {} turns it into an mathord (don't understand the reasoning)?

egreg

@Skillmon Because braces make a subformula

Skillmon

@egreg ok.

@manooooh keep @egreg's comment in mind (his answer to my code)!

manooooh

@Skillmon okkkkk

Anyway I don't understand why braces make a subformula, guys

egreg

@Skillmon Try this plain TeX file

$\mathclose[^2$ $\mathclose{[}^2$

\bye

2

user image

manooooh

20:29

@egreg can I hear your opinion of why I can't write $={]1,2[}$ ? What's the difference between {]} vs. \mathopen], etc?

egreg

@manooooh You might, but it's wrong nonetheless. And that Bourbakist notation is horrible and, in Knuth's words, perverse.

manooooh

@egreg well... I have not received a counter-example yet

egreg

@manooooh Easy

${]\sin\pi,3[}$

$\mathopen]\sin\pi,3\mathclose[$

\bye

user image

manooooh

@egreg oh my sin has mental problems here too!! That's a perfect solution, thanks!

@egreg why did you write \bye?

egreg

@manooooh That's to be compiled with plain TeX, it's faster for simple examples

marmot

20:37

@manooooh Because all his attempts to redefine it to \ciao got rejected. ;-)

manooooh

@marmot LOLOL don't say that!! :P

@egreg ok

egreg

@marmot \let\ciao\bye

@marmot Try \def\ciao{\bye} instead.

marmot

@egreg I think \xdef\ciao{\bye} will be even worse. ;-)

manooooh

@egreg anyway I don't know why MathJax ignores it; I usually test code in the rapid-display of math.SE site and there's no difference between ${]\sin\pi,3[}$ and $\mathopen]\sin\pi,3\mathclose[$ . I must always pay attention because if I use MathJax compiler these types of problems do not appear. Do you know why MathJax generates the same output in both cases?

egreg

@marmot No, it does exactly the same: an error and quit

marmot

20:42

@egreg Yes, but it is more effort to type.

egreg

@manooooh Because it's not TeX. ;-)

Ulrike Fischer

@manooooh mathjax is not tex. It can read tex syntax but it doesn't use tex to render the equation.

manooooh

@marmot I prefer \adiós (with accents ;))

@egreg @UlrikeFischer you have said it in unison! You should be a couple

marmot

@manooooh Even more errors. ;-) (I'd use ducks instead of the ugly [ and ] anyway. ;-)

egreg

@manooooh \let\adi\bye\adiós

manooooh

20:46

@egreg @UlrikeFischer ok, I will consider that. Thank you! You teach excellent

Phelype Oleinik

@manooooh With XeTeX or LuaTeX you can \let\adiós\bye :)

manooooh

@marmot hmmm... that's an interesting option, is there a duck-math-operator defined in amsmath? O.o

@PhelypeOleinik wow!

marmot

@manooooh \DeclareMathOperator will help you to do that.

manooooh

@egreg hahah nice! What's the meaning of \adi?

@marmot inside \DeclareMathOperator should we use \begin{tikzpicture}...\end{tikzpicture}?

marmot

\\DeclareMathOperator{\Lduck}{\tikz{\duck}} \\DeclareMathOperator{\Rduck}{\tikz[xscale=-1]{\duck}}

I don't know why there are additional backslashes, I didn't type them...

manooooh

20:50

@marmot maybe when you type \ inside ```

marmot

@manooooh But I did that above, too. Weird.

manooooh

@marmot ahh I forgot the tikzducks package!!

Skillmon

@manooooh he let \adi to by the same as \bye, it therefore is \bye. If you then type \adiós it's the same as \adi (because TeX ends its reading after it encounters \bye, everything afterwards is ignored, and ó is not a valid character in a control sequence)

marmot

@manooooh Just hope you didn't create your nice avatar with another package. ;-)

egreg

@Skillmon Exactly: the characters above 127 are assigned catcode 12.

manooooh

20:55

@marmot what really happens is that @samcarter created it for me; I just took a screenshot of the duckling :S

marmot

@manooooh Was that before the world cup?

manooooh

@Skillmon thank you! So when I use the command, should I invoke it using \adi?

@marmot I think so

@Skillmon I'm sorry, that was a nonsense

@egreg @Skillmon @DavidCarlisle: should I always always always use \mathopen and \mathclose? For example, $x=(\sin x)$ vs. $x=\mathopen(\sin x\mathclose)$

@egreg @Skillmon @DavidCarlisle ahá! I found that $x=\left]\sin x\right[$ works perfect! I would like to always remember what to use with this type of mathematical notation. Do you recommend, therefore, the use of \left] and \right[ even for this cases?

@DavidCarlisle I found a perfect solution: why instead of using $\lim\mathopen(1\mathclose)$ or something like that we don't use $\lim\left(1\right)$ (or for example $$\lim\left(\sin x\right)$$)? This is elegant, simple, understandable by all (fools and not fools), fast, works for any math operator, is consistent, etc.

I think that all types of notations to correct these problems are solved with \left and \right

Skillmon

21:18

@manooooh I think there was a question with an answer going into detail why one shouldn't always use \left and \right, but I'm a bit busy at the moment. @egreg and @DavidCarlisle are way better than I am anyway.

manooooh

@Skillmon oh ok, I will study that. Do not pay attention to this kind of weird questions that I ask! Make your stuff peacefully

Skillmon

Anybody interested in playing around with the new version of ducksay to help me debug it? (documentation is not yet done, maybe I'll get that done tomorrow, so testing could start tomorrow in the late evening)

manooooh

@Skillmon maybe you were referring to this question?

Skillmon

@manooooh I wasn't thinking of that one, but I think it gives enough reason by itself :)

manooooh

@Skillmon well, I red that the most common problem when using \left...\right is when we need to use multiline equations, but I never had to separate equations, so for the moment I will keep thinking that \left...\right are the perfect solution (until someone of you gives me a counterexample ;))

egreg

21:35

@manooooh No, ( is by itself an Open atom; since ] is a Close atom, you change its nature, if desired, with \mathopen]

manooooh

@egreg I don't understand. I thought we don't use \left, \mathopen, etc. because sometimes it isn't necessary; sometimes there's simple equations that don't need these (not multiline equations, no displaystyle, etc.). So \mathopen and \mathclose does not cover all possible cases to make the math better written? I'm going crazy with all the possibilities that exist to write a math formula...

Skillmon

@manooooh what he's saying is, that ( as an opening bracket already gets correct spacing and you don't need to use \mathopen( as it already behaves that way by default. The same is true for [ as an opening bracket. But if you want to use [ as a closing bracket instead, you'd have to tell TeX that in this particular case [ is not an opening thingy, therefore you use \mathclose[.

user280247

Do anyone know why vanilla texlive installation could crash? I've reinitiated installation three times :( and it starts from scratch

Skillmon

@santimirandarp it should tell you why it did crash. What did it say while crashing?

user280247

Any solution?

manooooh

21:46

@Skillmon ohhh! Yeah, you're right, in most cases ], ), } are for closing stuffs, but when combining these with another binary operator (like =) TeX doesn't understand it and treats =] like another thing. So in this particular cases we should write =\mathopen], right?

egreg

@manooooh Avoid adding \left and \right indiscriminately

user280247

@Skillmon i've found someone with similar problems tex.stackexchange.com/questions/321650/…

user280247

but there's no solution or maybe i didn't get it

egreg

@manooooh Anyway, \left<delimiter> makes the delimiter into an Open atom without the need of \mathopen, so \left] is good by itself.

manooooh

@egreg ... but \left] isn't the solution for all problems, right?

I should read all the answers you made in TeX.SE in addition to the manuals, and create my own document with all possible variants. This solutions has many alternatives... and I'm getting a little crazy :)

egreg

21:51

@manooooh No.

@manooooh Similarly, \bigl] will make an Open atom.

manooooh

@egreg thanks. Where can I find documentation to "open/close atom"?

Skillmon

Anybody else unable to upload images atm?

manooooh

@Skillmon me! I couldn't upload an image to the chat

Skillmon

@manooooh so imgur is at fault, not my end. I'm relieved.

« first day (2874 days earlier) ← previous day next day → last day (2356 days later) »

Transcript for

Sep '1813

$TeX - LaTeX$ TeX, LaTeX and Friends

\begin{chat}

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2025 Stack Exchange Inc; legal

mobile