« first day (4369 days earlier)      last day (860 days later) » 

02:56
Hi, I need a little help. I created a dedicated Overleaf project so that you can toy with the code (https://www.overleaf.com/8354476178bdvvjjpvgpmg).

The problem is that biber complains about `warning: 2 characters of junk seen at toplevel` but I cannot find it. I tried all the normal "comment out" approaches and narrowed it down to two entries.

I cannot properly explain it, but it does not make sense (to me) when I dig deeper.

Going to bed now... (= giving up) :)
@Dr.ManuelKuehner There's an spurious comma at the end of line 8 and a } on line 9 of the bib file. Just delete those
@PhelypeOleinik Give me a sec.
@PhelypeOleinik OMG, thanks! Then the warning was misleading?
@Dr.ManuelKuehner No (to me). When Biber says “characters of junk seen at toplevel” it means that it found characters that don't belong to any bib entry, so you have to search outside the entries
@Dr.ManuelKuehner Brace match highlighting helps a lot in these cases :)
@PhelypeOleinik Ah, thanks. I did not read the message that way.
@PhelypeOleinik Again, thanks a lot! I was looking for Unicode stuff (unbreakable space, weird dashes, etc.)
@PhelypeOleinik Going offline now - you made my day!
@Dr.ManuelKuehner Oh. Biber should not mind those: it's supposed to work with Unicode
@Dr.ManuelKuehner 'night
03:08
@PhelypeOleinik Weird was that when I deleted the second entry then the warning disappeared!
@Dr.ManuelKuehner Hm... Maybe it only complains if the “junk” is between entries (not sure)
@PhelypeOleinik Anyway, that is a fight for another day :). But this is the reason why I did not understand where the issue lies.
 
8 hours later…
10:51
@MarcelKrüger it opens now. But just to keep you busy, this here errors with ! Argument of \__tl_rescan:NNw has an extra }. (it works fine with lualatex):
\documentclass{article}
\usepackage{acro}
\DeclareAcronym{cd}{
  short = CD ,
  long = compact disc
}


\begin{document}
abc %\ac{cd}
\end{document}
11:44
In LaTeX, I would like to set my essays title to the font Book Antiquea and size 16. How can I do that? I put the line \usepackage{newtxtext,newtxmath} to my document.
@JaakkoSeppälä If you use the fontspec package with lualatex or xelatex, you can use whatever fonts are installed on your operating system. How to change the font of the title, will depend on the class you use for your essays.
@samcarter I got an error: LaTeX Error: File `antiqua.sty' not found.
@JaakkoSeppälä If you use fontspec with lua- or xelatex, you don't need this package. You still haven't told us which class you use, but here a quick example how to change the main font of a dummy document:
% !TeX TS-program = lualatex
\documentclass{article}

\usepackage{fontspec}
\setmainfont{Book Antiqua}

\begin{document}

test

\end{document}
I ran `xelatex digikulttuuri.tex ` and the output had lines `Package hyperref Warning: Rerun to get /PageLabels entry.

kpathsea:make_tex: Invalid filename `Book Antiqua', contains ' '

! Package fontspec Error: The font "Book Antiqua" cannot be found.`
@JaakkoSeppälä Do you have Book Antiqua installed on your operating system?
I guess that I don't have it. I just have Linux Mint.
@JaakkoSeppälä sorry, no idea how to install fonts on mint
12:14
@samcarter I see. But thanks for your help!
 
2 hours later…
14:39
@mickep a comment for you to enjoy tex.stackexchange.com/questions/661994/…
@DavidCarlisle did you see the question with the more than 60 \expandafter's?
@UlrikeFischer no?
@DavidCarlisle arabic url. Very impressing, I wonder where they are from.
Oct 5 at 11:14, by David Carlisle
@UlrikeFischer looks fine, always good to have more \expandafter
@DavidCarlisle I found who to blame: tex.stackexchange.com/a/456682/2388
14:48
@DavidCarlisle Ooh, indeed. But is it right that luatex does not add ic? Doesn't that depend on some \mathitalicsmode?
@UlrikeFischer The good old saying "Viel hilft viel"?
15:13
@mickep yes it depends on that but latex and the user do not set it so ic isn't added in this example
@DavidCarlisle Yes, indeed.
@UlrikeFischer I blame @UlrikeFischer
16:06
@DavidCarlisle I felt the need to elaborate a bit: tex.stackexchange.com/a/662018/52406
@mickep why do you think I pinged you here? :-)
@DavidCarlisle Oh, I was tricked into it. (In fact, I already had started to write something when you first pinged me)
@mickep your answer is longer but my comment to Gaussler says the same thing
@Gaussler that depends on your desires — David Carlisle 27 mins ago
@DavidCarlisle Haha, yes, but without nice pictures. :P
 
2 hours later…
17:45
@UlrikeFischer Fixed with the new engine I just uploaded (You can update by rerunning the installer)
@MarcelKrüger just curious, but what was wrong? the scantokens implementation?
@DavidCarlisle @mickep @egreg : All of you provided excellent and helpful answers to tex.stackexchange.com/questions/661994 . Being a noob, I'm completely clueless on whom to give a checkmark. So unless you all agree on one and the same answer, the question will receive no checkmark from me at all.
@AlbertNash Thanks. I do not care for checkmarks, but the other guys usually fight about them, so I suggest offering it to one of them. In fact, best först give it to one of them and then a day later change it to the other. ;⁠)
@AlbertNash better to give it somewhere or the system will have it as unaccepted. Obviously the rule if egreg and I both answer a question is not to give the tick to egreg.
@DavidCarlisle : Ok. Then I tick an answer of @egreg to another question of mine, so as to maintain the balance between you two.
@mickep Thx! Done.
18:52
@JosephWright \cs_set:Npn \__pdfmeta_xmp_generate_bom:{\codepoint_str_generate:n{"FEFF}} works fine ;-)
@JosephWright \codepoint_generate:nn{"20AC}{} works with pdflatex but errors with lualatex because of the missing catcode. Should that be unified somehow?
19:04
@AlbertNash and @mickep -- I've started an inquiry about this offline, although I'm starting with the text situation rather than the math. I'll report herre whatever I find, and intend to consider a report in TUGboat as well.
@barbarabeeton Interesting! But are there problems with the text fonts? To me it looks more like this is a math italics problem. I don't think the gyre gang was lazy, but rather that DEK was very careful.
@mickep -- The original Computer Modern has this problem. I first noticed it when setting a paragraph in italic, and a sentence in the middle starting a new line was noticeably indented. It can be noticed readily in a poem where every line starts with a cap. It's kind of a reverse italic correction situation. I really don't know the reason, and have never asked Knuth, but will try to find an occasion to do so. Maybe something can be found in Volume E, but that's in a box at the moment.
@UlrikeFischer Yes, scantokens acted like scantextokens, especially not adding \everyeof.
19:29
@barbarabeeton Yes, I have seen that as well.
19:41
hey folks, like there's \bye in plain.tex, \end{document} in LaTeX and \endinput in general, is there some command like \foobar such that whenever a macro \bar is being expanded, reaching \foobar stops the expansion? like \endinput for a macro
like "break" in a for loop in most languages
@DanielDiniz no not in general, tex works by macro expansion not function calls so the body of macro is inlined in the input buffer with no mark at the end you could jump to
hmm I see! I think it could be implemented but as an engine primitive, probably impossible using TeX definitions. so \def\bar{abc\foobar def} would expand \bar into "abc"
19:57
@DanielDiniz you could more simply do \def\bar{abc} ... similarly any macro where you add \foobar you could add \foobarend at the end and skip to there.
the use of that would be in conditionals, like in "more normal" languages where you write things like "if my_value == 0: break" to skip the rest of a loop. so you could put a line "\ifdim\mydim<0pt\relax\expandafter\foobar\fi" in the middle of a macro
@DanielDiniz as I say you can do that now, expl3 mapping constructs have such a break for example,
@DanielDiniz ^^
@UlrikeFischer The intention was that the second argument is required ...
@UlrikeFischer Arg #2 is ignored by pdfTeX but is expected for all engines
20:27
@DavidCarlisle ahh very interesting! i'll take a look
@JosephWright Did you do something special with latex2e's develop branch some hours ago?
20:53
@JosephWright yes I know. I only wondered if it should error with all engines if empty.
@barbarabeeton Thanks! Following tex.stackexchange.com/a/662032, just in case you have any influence, and for the purpose of testing, could we get some real-world products of 𝑌 with anything in testmath.tex in the long run? It'd be great to have (dotless, with central dot, or with cross) multiplications from the left, from the right, or from both sides. (I don't know any real examples myself but could imagine that folks might be using stuff as 𝑋𝑌𝑋̅ and 𝑋𝑌𝑋^⊤.)
21:11
@AlbertNash -- I no longer have easy access to a searchable archive, and I'm not actually a mathematician. (I've just hung around with them for a long time.) If you ask nicely at math.stackexchange, somebody might actually know something useful. Let us know here when you post something -- there are enough mathematicians here to support such a question so it won't get ignominiously closed. (@DavidCarlisle, @egreg, @mickep, @daleif)
@MarcelKrüger Not that I know of!
@UlrikeFischer I can arrange that I guess
@MarcelKrüger Made a local branch, nothing else deliberate
@JosephWright you don't have to ;-). I was only playing around and trying the stuff.
@JosephWright does only pdftex use the intarray and luatex not?
@JosephWright Just Github being weird then I guess. It showed me that something changed but refused to give any of the other information it normally includes.
21:47
@barbarabeeton Thank you, I see. In this case, please kindly forget my request. I posted my own test cases in tex.stackexchange.com/questions/662029 in the meantime (without any products, though).

« first day (4369 days earlier)      last day (860 days later) »