« first day (3622 days earlier)      last day (1355 days later) » 

5:58 AM
@ComFreek poor <insert package by @DavidCarlisle>, will never be used even though it is by definition bug free!
@VinayPandey beginner level questions for LaTeX? I'd say yes but no. You can ask a few simple questions, but keep in mind that chat isn't really an asynchronous thing, so questions are generally better suited for the main site. And chat shouldn't be abused for loads of such questions. Maybe you should first read an introduction to LaTeX (if you didn't already). Some folks frequenting this chat wrote one up at www.learnlatex.org (it is still subject to some changes, but it is already usable).
6:51 AM
@VinayPandey only really oneliners but you are free to ask beginner questions about Tex at tex.stackexchange.com If you mean this, the first version wasn't really answerable but I see you have edited it since, I'll look later:
Q: Place Index/Table of Content without page break

Vinay PandeyI want to remove the page break forcefully inserted by \tableofcontents and so far I have not had much success browsing through forums/StackExchange. In the sample image I have posted, I want the contents of page 2 to appear in the centre right after State of Uttar Pradesh and Others (given on p...

@samcarter_is_at_topanswers.xyz fair point
2 hours later…
8:48 AM
I have question regarding \insert and language
With the standard latex footnote insert
latex keep in memory the language in the point where the insertion is made
and, consequently, we get the correct hyphen in the footnote, even i we swtich the language


the first footnote is correctly hyphenated with greek rule
even if at the end of the page, we are not anymore in greek system
however, with reledmac inserts, it does not
I suppose we miss something in the reledmac code
but I have no idea where
@DavidCarlisle can't thank you enough for your patience. I have edited the question. I had earlier cropped the image to take up less space. Essentially, there is plenty of space on page 1 to show the table of contents but it is forcing it to go on page 2.
so any help will be useful
9:12 AM
@VinayPandey Note I would have known that if you had posted a usable example, but if you post code that does not run then only you know the output. as I commented under the question it page breaks as you are forcing sections to do \newpage
I apologise sincerely David. I am totally new to this forum (despite being on Stackoverflow for a long time facepalm)
@Maïeul latex really does nothing there it's just falling through the primitive tex \language feature. Probably best if you make a real example and post as a question where there is more space to try code examples
@VinayPandey no worries, but tex (far more than other programming languages) is highly context sensitive. If you show a fragment of C or Java then some functions may not be known but at least a user can parse the fragment, but in tex packages can change the syntax on the fly, so if you just show a fragment it's more or less impossible to say anything, so we almost always ask for complete test files
@DavidCarlisle I am still not sure why there is a page break. I am not creating any section before calling table of contents
@VinayPandey \tableofcontents (in article class, without knowing the class it is not defined at all, see above comment:-) is \section*{\tableofcontentsname}\input{\jobname.toc}
@VinayPandey tableofcontents uses \section too, to set its title. Remove the \newpage to test if it helps.
9:18 AM
@UlrikeFischer I probably should know but is this affected by internal using hooks in preamble discussion
A: How to use \AtEndDocument with the class exam?

David Carlisle You just need to add the text before the class adds the page break to the same hook \AtEndDocument{bar} \documentclass{exam} \begin{document} foo \end{document}

@UlrikeFischer That does indeed help. Woo!!! Now, for the new page in article class, what should I use?
@DavidCarlisle yes. Your example will work, but
@VinayPandey ?
will change order, the foo will "wander" before the examp hook.
@DavidCarlisle I meant that removing \newpage in the redefinition of \section removed the extra page break
9:23 AM
@UlrikeFischer yes that's what I thought, OK so if you want it to come after the \clearpage you'll need an explicit label or rule or something. That's probably OK, I think...
@VinayPandey yes but then you asked what you should use, that depends on what you want, do you want normal sections to start a new page?
@DavidCarlisle I do
Also, if you don't mind creating a answer for my question. This way I should be able to mark the question as answered :-)
@Maïeul does reledpar end up doing \sbox0{\setlanguage{zzzz} stuff in this language} \unhbox0 if so you are falling victum to a premature optimisation in tex the program, language nodes are not properly inserted in an hbox as tex "knows" there will be no linebreaking/hyphenation.
@VinayPandey OK I'll put something there
@DavidCarlisle yes, your example should then be something like
@DavidCarlisle is there a way to remove the page break for the table of content and keep the page break for the rest of the sections?
9:28 AM
@DavidCarlisle no, It does not. So you say it should
@VinayPandey there is certainly a way. But basically it means that you want to use two types of section commands: one with page break and one without. So you need two definitions.
@Maïeul you should really better make a small example. It is much easier to investigate if one see the code.
Ohh, got it. So basically let \section remain as is, and then create another one called \section1 or something based on \section class
@UlrikeFischer yes, I should, but the example show it :)
i will try to do it this week-end
@VinayPandey I was just writing an answer on site:-)
@Maïeul no you need to avoid that construction to preserve language nodes
@Maïeul which example, you mean the greek one above? what preamble does that need?
Thanks @DavidCarlisle you have been so wonderful. I can't thank you enough for your patience :-)
9:37 AM
@PauloCereda Reasonable form of address ^^^ Your usual form of address >>> chat.stackexchange.com/…
orry @DavidCarlisle I start I conversation with some colleague
but I will open a a new issue
9:56 AM
@DavidCarlisle you said "it is bad style to put formatting in \zzzname macros", where do you suggest a common formatting should go ?
10:15 AM
@VinayPandey \contentsname should just have text for example if you add \usepackage[italian]{babel} it will be redefined to be Indice so you will lose all your formatting. That is exactly why they were introduced so language packages could translate all these text-only macros and they will work with whatever formatting is defined by the classes or packages used. Your class should instead define \tableofcontents taking the version from article.cls and changing it appropriately
10:37 AM
@DavidCarlisle oh no
@DavidCarlisle ooh
so @DavidCarlisle @UlrikeFischer now I have time for a full MWE (sorry, I did a wrong pastebin in the previous one)



as you can see, with xelatex, the latex classical footnote are correctley hyphenated
while the reledmac footnote (\footnoteA) are not
so I suppose there is is something wrong in the reledmac code, which either lost the hyphenation, either dont get it
11:04 AM



@UlrikeFischer @Maïeul ^
@Maïeul compare your original version log with \showoutput the first footnote after the text has \penalty 10000 and a \rule which is the \strut added for spacing, but the footnoteA version just has the \rule with no penalty but a string of characters needs to be followed by white space or a penalty (or similar) to be considered as a "word" for hyphenation so adjoining the \strut immediately after the text prevents hyphenation of the last word.
2 hours later…
12:49 PM
@DavidCarlisle oh! nice
thank you
@Maïeul basically you will have to replace various \strut by \@finalstrut\strutbox.
@Szabolcs Are you a Mac user? If so, Bibdesk is a great option.
@UlrikeFischer no boarder effect?
the archan of TeX typography are still mysterious for m
@Maïeul well the standard footnote command uses this at the end. It should be safe.
yes, indeed
I triy
12:59 PM
@UlrikeFischer whole thing done quite early in the 2e cycle:
% \changes{v1.0p}{1994/11/09}
%         {Revert \cs{finalstrut} to 2.09 equivalent (from ltpatch)}.
% \changes{v1.0x}{1995/10/31}
%         {Add \cs{nobreak} in horiz mode to allow hyphenation.
%                   internal/1931}
@UlrikeFischer and @DavidCarlisle that work with my mwe
but we the user example, I have still the problem
but itis a more complex example
....\TU/IFAO-GrecUnicode(0)/m/n/8 προφητεύσουσι
....\penalty 10000
....\glue 0.0
....\kern 1.33337
....\TU/IFAO-GrecUnicode(0)/m/n/8 ]
....\kern 0.0
is the \penalty 10000 enough to breek the world ?
@Maïeul yes... if there is white space before the word and the patterns give a break point in that word.
so yes indedd, normally, the patterns give a break point in that word, if we are in greek
\setmainlanguage[variant=french, autospacing=false, autospaceguillemets=true, frenchfootnote=true]{french}
\newfontfamily\greekfont{IFAO-Grec Unicode}



\pstart \edtext{Ὁ}{\lemma{ὁ}\Afootnote{om. S J}}\footnoteA{blabla} Θεὸς γὰρ ἀψευδής ἐστιν ὁ Υἱὸς τοῦ Θεοῦ. Εἰ δὲ \edtext{ἀναισχυντεῖς}{\lemma{ἀναισχυντεῖς}\Afootnote{ἀπαισχυντῆς Pa}}, πάλιν προφηστεύσουσ
but that is the very strang point
it sems that the information about the fact it is in greek is "lost"
and what is the more strange, it is lost only with xetex, not luatex
@Maïeul hyphenation in lualatex is different in many ways
but if I try with "all the document" in greek
or if I add \newpage sss before \end{greek} I get the correct greek hyphenation (with xetex)
1:14 PM
@Maïeul looks like I don't have IFAO-Grec Unicode does the same problem show up with the font you used before?
I try
yes, same problem
original user question github.com/maieul/ledmac/issues/880
@Maïeul work calls I may look this evening
@DavidCarlisle many thanks. I am persuaded there is somewhere a bug in reledmac, but I have no idead where I should look on
@Maïeul Just blame @UlrikeFischer that usually works.
1:43 PM
@AlanMunn Yep, Mac user. Thanks for the tip! I should give BibDesk another try.
2:40 PM
plz help to fix this
@UlrikeFischer @samcarter_is_at_topanswers.xyz can you have a look at tex.stackexchange.com/questions/564794/…? This:
works ok on TeXLive 2017 but not on TeXLive 2020 (where the bold R is substituted...)
TeXLive 2017 moe or less ^^^
TeXLive 2020 updated yesterday ^^^
(using `pdflatex` on the same file). I have a warning in the second one, obviously.
\foreach \name/\uid in {a/P, gla/2, it/9P, ri/2, i/2, a/3P}{\textbf{\name} & \uid \\}
\end{tabular}` Can someone tell the error here?I have mangled the original data
@VisheshMangla Not easy, starts here: tex.stackexchange.com/questions/309765/…
Q: Parallel loop [python-like zip] in Tikz

plutonAn easy way to write axis labels in Tikz is: \foreach \x/\xtext in {-1,-0.5,0.5,1} \draw[very thin,gray] (\x cm,1pt) -- (\x cm,-1pt) node[anchor=north] {$\xtext$}; Is it possible to have something similar but where \x and \xtext do not get the values from the same list as follows ? \foreach \...

I used the above answer and modified the solution accordingly. Can you please tell what mistake I have done?
@Rmano I think this is this bx/b stuff
2:50 PM
@samcarter_is_at_topanswers.xyz yes, beamer should probably adapt. Or cmss.fd need an adaption.
@UlrikeFischer Hmm... my main system at the uni will upgrade this weekend (finally!), but this can bite me. Is there some simple workaround for the time being?
@UlrikeFischer If you tell me where to adapt ...
@Rmano try


(no idea if this is a good idea or not)
@samcarter_is_at_topanswers.xyz well, it works ;-)
In both TeXLives
@samcarter_is_at_topanswers.xyz better declare a substituation. And perhaps open an issue in latex2e, then we can discuss if it should be in ot1cmss.fd by default.
\DeclareFontShape{OT1}{cmss}{b}{n}{<->ssub * cmss/bx/n}{}
@Rmano ^^^
@UlrikeFischer Not a font expert, but in general I think that a b -> bx substitution should be preferred in general over a b -> m? Or not?
Where exactly should we open the issue? here: github.com/latex3/latex2e/issues ?
3:06 PM
@Rmano this is not how the code works. It doesn't try out a list (first bx, then sb, then ...), it falls back to the default. I mean assume that bx doesn't exist either, should it then try b again? or some heavier series first?
@UlrikeFischer Ahhhh, ok
@Rmano hm, wait. The substituation works without beamer. So something else is involved, I need to check this.
@UlrikeFischer Ok, will stay calm and go back to work on my class beamer ;-)
ooh Coca-Cola
@UlrikeFischer Shall I open a beamer issue to move this discussion there?
3:19 PM
@samcarter_is_at_topanswers.xyz wait. I'm just trying to figure out why the beamer setup changes the substitution.
overleaf.com/4163315319jkrfwcwrycxm link with edit permissions
@Rmano Thanks , I should have read the post carefully . I was looking at the output of the answer than the title of the error.
@UlrikeFischer if the document is using cm how come bold is using b not bx ? Then you wouldn't need a substitution?
3:36 PM
@DavidCarlisle \bfdefault is b, and the document uses \bfdefault to setup \mathbf.
@DavidCarlisle try this and compare with the output if you use beamer:

\sffamily abc
{\fontseries{\bfdefault}\selectfont abc}
{\bfseries abc}
@UlrikeFischer yes but why is it b I thought we arranged it was bx if you were using cm, but perhaps I mis-rememberd and that's just done with substitutions as you did here? (I know I could check but...)
Well, hello there :> I have a quick question here(I just started with LaTeX and this example isn't from me originally.)
I put some images with the tabular environment into my document and now have the problem that even with the positioning command h! its not compiled in the order I put it into the code:

\includegraphics[width=\linewidth, valign=T]{image.png}
& \begin{itemize}
\item some bulletpoints next to the image
\end{itemize} \\
@Syralia the only reason to use table is to allow the table to move, and by using [h] you are preventing [tbp] so not at top not at bottom not on a page on its own. so most likely the table is not allowed anywhere and will go to the end of the document
@DavidCarlisle Okay, something new to me. They actually used to appear where I put them in the code but after the third table, they skip the next section header and float below it
@Syralia h allows the table to be place at that position if it fits, but if say its height is half the text height and it comes anywhere below half way down the page it will not fit so latex will move it to a later page, to avoid ugly white space, that is the whole reason for having table if you just wanted a tabular in the document flow you could use tabularx directly.
3:48 PM
@DavidCarlisle well with the newest kernel, it is b\@empty, but somehow beamer does something different to article, and extends this to m instead of bx.
@UlrikeFischer so it's all @samcarter_is_at_topanswers.xyz's fault. OK
@DavidCarlisle ah thanks a lot! I would also create a page with newpage for LaTeX to put them into order; but that also doesnt work. Any advice? like:
all other tables
\new section
@Syralia very hard to say if you just give fragments, but [h] is always wrong at the very least use [htbp] \newpage does not stop tables you could use \clearpage but you most likely only need that at all as by using [h] you are explicitly preventing the table being on a page on its own (as you removed p)
@DavidCarlisle my code looks exactly like the example above 7 or 10 times in order, then theres:
3:54 PM
If you really must you could use [H] then it will not move but then you make me sad and curse the person who first implemented that option. One of teh highest voted qposts in the tag is:
Q: Keeping tables/figures close to where they are mentioned

artimessIs there any package or a method to force LaTeX to keep floating environments like table and figure closer to where they are declared?

@DavidCarlisle I wished but everytime I see b/bx I wonder ...
\subsubsection{a subsubsection}
%here comes something in
\subsection{a new section}
\item some item
@DavidCarlisle AH thanks a lot! :)
@Syralia .7\textwidth+.3\textwith+.3em is more than textwidth so you get a line break and one above the pther
@DavidCarlisle I thought 1 in sum is the whole page?
@Syralia you have put a normal inter-word space between the two minipages
3:58 PM
@DavidCarlisle Ah, thanks again! I should come here more often :D
typesets as hello world not helloworld :-)
1 hour later…
5:19 PM
@DavidCarlisle @samcarter_is_at_topanswers.xyz I think I found the place where it breaks. The problem is that beamer setups math fonts using \bfdefault, and this initialize the font and then substitution goes wrong. But I'm not sure if this a latex bug or if beamer should use \bfseries@sf instead:


{\fontseries{\bfdefault}\selectfont abc}
{\bfseries abc}
5:49 PM
Can anyone tell me why this produces different output for pdfTeX and XeTeX?
  \def\pdfliteral#1{\special{pdf:code #1}}




  \pdfliteral{\aa\space\ab\space\ba\space\bb\space\strippt\x\space\strippt\y\space cm}%

The list that TeX reports is exactly the same for both engines.
6:31 PM
@UlrikeFischer, @DavidCarlisle Fingers crossed ...
@JosephWright starting to build?
@HenriMenke You want pdf:literal not pdf:code (which is equivalent to \pdfliteral direct)
@UlrikeFischer Just done the merge, want to see that builds before tagging
@JosephWright I actually wanted to copy the new luaotfload but now realize I forgot it ;-(
@UlrikeFischer Oh: guess I'll do that in dev shortly ...
6:52 PM
@JosephWright I see, thanks. Unfortunately, the XeTeX driver in PGF uses pdf:code. How can I make these work together?
@JosephWright where did you find this? I was just skimming the docu for it.
@UlrikeFischer l3backend doc:-)
@UlrikeFischer What @DavidCarlisle said: I tried everything out then documented it :)
7:08 PM
@JosephWright but how did you got the idea to try out pdf:literal? You could not have started with pdf:a, pdf:b, ... pdf:z and go through all combinations ;-)
@HenriMenke The best I can offer is 'see how I did things in l3draw backend setup' - most of the time you can just swap the two as the CM doesn't matter
@HenriMenke See l3backend-code, first part with my comments about things that are wrong in the dvipdfmx docs :)
@UlrikeFischer It said it's like \pdfliteral, and it starts with pdf:, so I guessed ;)
@UlrikeFischer See also my comments about literal direct
7:40 PM
@JosephWright hm, you are saying that pdf:literal is equivalent to pdf:content, but this is not true: pdf:contents adds an additional q/Q pair (I had tried pdf:content, and the box is not rotated with it),
@UlrikeFischer Ah, quite possibly: should adjust that comment
@UlrikeFischer I've altered
8:38 PM
I see many people say their document is "compiled" by their chosen engine. What is the correct verb to use in this case? Is it "typeset"?
@LaTeXereXeTaL I used to argue that compile was the wrong word, as tex isn't a compiler, but it's a losing battle, compile is fine.
I have been guilty of that too but I understand it's not a compiler. I just wanted to make sure to use the correct terminology from now on.
@LaTeXereXeTaL the notion of the difference between a compiler and an interpreter and (as here a macro expansion system) are blurred these days anyway languages that are traditionally interpreted get just in time compilers and languages that were traditionally compiled get just in time compilers and it's hard to spot the difference. The main programming idiom is a macro expansion language, but for most people...
... (who are not in this chat) getting typeset output is more important than the macro processing programming language, but "compiler" is easier to say than "digital typesetting system customisable via an integrated macro processing language"
Excellent. I understand. Thank you very much!
@LaTeXereXeTaL overleaf claim to have 5 (6?) million users and presumably they all call it a compiler...
8:53 PM
That's one reason I asked. My students use Overleaf and I've been using the term to match the button, but in the back of my mind I kind of new it was wrong. I pride myself on pedantry so I needed to know.
@LaTeXereXeTaL ah I gave you a very pedantic reply so you should be happy:-)
It was perfect!
@DavidCarlisle oh yes we do :-) (both call it Compile and have over 6M users) -- technically, it is compilation, if you accept that PDF is the natural result of processing a bunch of TeX and related files. (You take a lot of files and you produce a single output file [and a lot of logs and other aux])
@yo' yes but that's compile in the natural language usage of organising a bundle of stuff, it isn't what would traditionally have been called a compiler in a computing context
@DavidCarlisle of I don't argue with this one :-)
but the thing is, the CS/IT people put special meaning to all/any natural English word that could be even closely related. This they have in common with the math people (and the physics ones, and chemistry ones, and ...)
9:05 PM
@yo' clearly English is a dead loss and you should use a proper language where you can make up specific words digitalesSchriftsatzsystemanpassbarübereineintegrierteMakroverarbeitungssprache
@DavidCarlisle I'm afraid that if I put this to our German i18n, it would break the UI :-D
@yo' g translate actually added some spaces but I assumed they were in error and removed them
@DavidCarlisle spaces? in German? superfluous.
@yo' my thought exactly
9:39 PM
@DavidCarlisle Is this the reason all these strange 16:9 or 16:10 displays became so popular? To accommodate buttons wide enough to fit your word creations?
@samcarter_is_at_topanswers.xyz no people don't use the graphical interfaces, just the character mode console, but the wider screen helps emulate a 132 character line printer
@UlrikeFischer Thanks for investigating! Replacing \bfdefault at a dozen places or so indeed gives bold font.
@DavidCarlisle Without the authentic sound, that's not very convincing emulation
Q: Is there a line printer sound simulator?

Leo B.To improve "authenticity" of a mainframe simulator and for an additional nostalgic effect, I'd like to add sounds of a line printer to it. They should be similar enough to the original sounds for people to recognize familiar cadences of standard headers and/or footers of printouts. Before I attem...

10:07 PM
/wearing my compiler hat: it is definitely not a compiler
@PauloCereda democracy wins. It is now. unless you agree to always call it a digitalesSchriftsatzsystemanpassbarüberein
@DavidCarlisle I am fine with that term. We already have Entscheidungsproblem.
Overleaf is to blame for misusing the term. :)
@PauloCereda I think you meant Enten scheidungsproblem ?
@DavidCarlisle ooh
We could blame the Turing bloke. :)
@PauloCereda Hilbert's famous thesis on how to make crispy duck pancakes
10:13 PM
@DavidCarlisle oh no
@DavidCarlisle do you remember the dragon book?
@PauloCereda not well you know I did pure math not computer science:-) But I think I read the first version at one point
@UlrikeFischer I see your name mentioned on the MikTeX site, miktex.org/help. Do you know whom I contact if the site is not propagating my tokcycle package correctly. In particular, tokcycle.tex should be loaded into tex/generic, but instead seems not to be downloaded at all.
@DavidCarlisle it was probably around. :)
@PauloCereda Hopcroft and Ullman
@AlanMunn and Aho? Let me check...
@AlanMunn hm the edition I have does not feature Hopcroft... They have a book on automata theory though...
Good times. :)
10:22 PM
@PauloCereda only just: The first ed was in 86 and I'd have read it when starting the post doc in cs at Manchester which would be 87
@PauloCereda Ah, yes. Aho and Ullman then?
@AlanMunn bingo!
@DavidCarlisle ooh
I should have walked across the room. :)
@AlanMunn lovely!
We should have a similar cover for a L3 book: @Ulrike slaying dragons. :)
@AlanMunn ^^ you probably remember this one. :)
10:45 PM
@PauloCereda Yes, I have that too.
@AlanMunn ooh <3
@StevenB.Segletes open an issue here github.com/MiKTeX/miktex-packaging
@UlrikeFischer Thanks. Having trouble creating a valid account, but will try again.
@StevenB.Segletes if needed I can add an issue, but not tonight, it's too late now.
@UlrikeFischer I believe I created an account and issue successfully. Maybe tomorrow, just check if you can see it. Thanks so much.
10:58 PM
@StevenB.Segletes now you have a github account you can report bugs in all of @UlrikeFischer's packages.
@DavidCarlisle I'll take that under advisement, LOL.
@StevenB.Segletes I can see your miktex issue so I guess it worked
@DavidCarlisle Thanks so much for the confirmation.

« first day (3622 days earlier)      last day (1355 days later) »