Any Windows users?

@ArthurReutenauer Apologies, yes but only did it last week!

@ArthurReutenauer Yes, some time ago

@WillRobertson t @barbarabeeton There is some precedent, mathml (normatively) defaults to italic for 1 character <mi> and upright for multiple (and has some non-normative weasel words to say that mathvariant=italic should mean math or text italic depending on length) I thought about mentioning that but have always considered it as mathml digging itself out of a hole of its own making, rather than a practice to be emulated... But perhaps you are in a similar hole....

@ArthurReutenauer yes at the conference hotel

@WillRobertson look what I have today ^^^^ You probably dispute my qualification:-)

@DavidCarlisle I got it almost two years ago. :P

@egreg: let's not talk about penalties. :)

@PauloCereda \brokenpenalty

@DavidCarlisle ooh :)

@WillRobertson -- "not too often" ... dreamer! i've just run a grep over the "in production" area of the ams proceedings. there are 168 articles "active", and in them, 459 instances of \mathbf; i haven't yet counted how many articles they appear in, but most are in definitions like \newcommand{\mbZ}{\mathbf{Z}} and there are quite a few like \newcommand{\zr}{\operatorname{\mathbf{zr}}}. (i'll do a count later; on a tight schedule right now.) what does that do to your suggestion?

@egreg @WillRobertson @JosephWright @DavidCarlisle Thanks, that’s what I expected anyway. On another note, at what time did you receive my email to tug2015-attendees? I sent it last night (22:42 UTC), but it seems it just got through one hour ago.

@ArthurReutenauer -- the received time here was 08:56:26 -0400. reviewing the headers, it looks like it was held up somewhere in the circuits of aarhus university for some hours. i can send the headers (by separate mail) if you'd like to see them (i'm on a tight schedule right now; no time to dig myself).

@ArthurReutenauer Turned up here

@barbarabeeton Don’t spend any time on it, I was just curious - 08:56:26-0400 seems to fit with the time I see in my headers. Thanks

Quack! :)

@PauloCereda -- maybe you could look at this one. i know you use inkscape.

@barbarabeeton Sure thing!

@PauloCereda -- thanks. inkscape may not be the answer, but the question, by a new visitor, is in danger of being closed, and there's no option to say "leave open". (maybe i'll take that up on meta after i get back from my midday assignation.)

@WillRobertson I don't know anything about what you are talking about, really. But in any case, do you know how ConTeXt treats all of that? I mean, since ConTeXt is aheadd of LaTeX in many things (math is not really developed, though, unless I'm mistaken), at the very least you could look at how they do things (in particular all this unicode related behaviour).

You probably know. But just to remind, in case it helps.

@Manuel What do you mean by “maths is not really developed” in ConTeXt?

@ArthurReutenauer They have relatively weak math support

@JosephWright I don’t think that’s really accurate. Aditya Mahajan has done really a lot of things with maths in ConTeXt. Admittedly I never looked into it in details, but it doesn’t sound right to say that maths support is not really developed.

@JosephWright That’s not to say that one can do in ConTeXt everything that is possible in LaTeX, of course. The number of ConTeXt users who really use maths is very small :-)

Hi, if I have a question in regard to code-review tag do I open a new question on meta or can I just ask here quickly?

@ArthurReutenauer Yes, also the math mode changes in LuaTeX have resulted in some oddities

@ArthurReutenauer I mean they can't match amsmath with LaTeX

@JosephWright That’s what I wouldn’t be positive about.

@Takeru Wow, 0% unanswered tex.stackexchange.com/tags/code-review/topusers

@Takeru Depends what it is

@Takeru Ask and ye shall receive.

@ArthurReutenauer arrived this afternoon sometime. by the way I think @yo' is using a different hotel

@DavidCarlisle Thanks, I already got a reply. I think this will end on Airbnb, and there are much less expensive hotels in the vicinity anyway (€130 for a double room is just ridiculous). @yo' Just out of curiosity, where are you staying?

@Takeru code-review is hardly ever used, better really to extract specific questions about specific issues rather than just post the whole thing and ask that someone review it.

@ArthurReutenauer try getting a room in Oxford:-)

@DavidCarlisle I never did try, but I’ve heard of people who ended up living in West London and commuting from there :-)

@egreg: No penalties, please. :)

I found big problem, may be someone can give me a hint please?

@Nasser not unless you say what it is

I changed all my build to use texfot, which is very nice as it eliminates so much output from the screen. But I found a huge problem. Here it is: one some files, it hangs.

@DavidCarlisle @ArthurReutenauer Hotel Hornung has good options on Booking.com

@Nasser never used (or even heard of) texfot but usually catch if redirecting output is if tex stops for input eg for missing file, it will wait forever expecting a prompt.

Output written on index.dvi (55 pages, 368284 bytes)...
(now the process hangs) I look at the ps -eaf and I see

me        3987   617  0 11:45 pts/1    00:00:00 perl /usr/local/texlive/2015/bin/i386-linux/texfot make4ht -u -c /home/me/nma.cfg -e /home/me/main.mk4 index.tex htm,pic-align,notoc*

me        3988  3987  0 11:45 pts/1    00:00:00 texlua /home/me/texmf/scripts/lua/make4ht/make4ht -u -c /home/me/nma.cfg -e /home/me/main.mk4 index.tex htm,pic-align,notoc*

@Nasser just run tex by hand and see if it drops to an interactive error propmt

@DavidCarlisle texfot I found about from this post:

@Nasser yes but it's not really a "big problem" is it, if it works fine, if it doesn't either don't use it or make a minimal example and report to the maintainer (who seems to be karl)

@DavidCarlisle everything works fine without using texfot. And it only hangs on this one file (this is little larger file). I've been using texfot all day on many files, no problem. Only this file hangs

@Nasser as i say run the steps by hand, see where it fails

@DavidCarlisle what steps? I am not sure how to find why it hangs on this one file. Here is screen shot:

!Mathematica graphics

If I remove texfot, then it runs ok.

@Nasser that looks like an answer to me:-)

I might have to ask @michal.h21 on this, may be texlua is not working well with texfot

@Nasser add --debug --interactive

@DavidCarlisle ok, will do

@Nasser no idea what they do:-)

ok, now I see lots of stuff on screen. it hanged on this line:

I think it is a problem. Since it only hangs on this file, which hangs lots of images. And may be make4ht/luatex and texfot are not piping things to each others right. I do not know how texfot works. But this only shows up with make4ht

Here is the ps -eaf

I think only Michale can figure this one. I do not even know where to look.

@DavidCarlisle I also thought about this but is not important to get the whole context of the preamble because of packages interacting and changing stuff "globally"? Or do you mean the specific warning messages?

I found something also. It hangs also when I use htlatex directly to build. i.e. not use make4ht. So the issue is not with make4ht. It is with texfot. It hangs also when tex4ht starts to write to the index.dvi always. So summary:
1. texfot pdflatex foo.tex  do not hang
2. texfot make4ht .... hangs
3. texfot htlatex .... hangs
And this happes only on this one large latex file. Not on smaller files.

@Takeru that depends, if you can make a minimal example that makes a warning and only loads two packages, then no.

@Takeru I wouldn't worry about it, just ask the question. Worst that can happen is that it gets closed as too broad or unclear what you are asking.

@Takeru You might be interested: mychemistry.eu/2013/01/latex-templates

@yo' Thanks, I really I’ll go with Airbnb, but thanks anyway :-)

@PauloCereda I believe that Knuth was thinking to Brazil when he gave \brokenpenalty its name.

@egreg :)

@Nasser I am afraid that I have no idea what's going on, I never head about texfot. I am quite busy until Wednesday, I may look at this later. you can mail me, otherwise I may forgot about it

@michal.h21 no problem, thanks. I think the issue with texfot, since it also hangs with htlatex itself and not just make4ht. I'll try to see if I can ask Karl Berry for help, but I need to first try to make a smaller example where this problem shows up.

@ArthurReutenauer I meant that Hans Hagen himself said that in the list many times: math in MkIV is not really developed (just temporary solution), when he needs it for a book, or if he has time, he would put some time into it. I didn't mean that it's plain wrong, but that's what the author said; what Joseph said is true, some “oddities” are found. It's kind of abandoned part as of this moment.

@ArthurReutenauer By the way, apart from what @GonzaloMedina told you about tilda (which I think it's correct) I actually have had a few teachers that said tilda (and others tilde). I'm from Spain.

I believe the word itself wasn't entirely correct, as Gonzalo pointed out, but I've seen it used.

@Manuel If you mean not really developed as part of the core, I guess that’s true, but Hans is not the only one writing ConTeXt - even if it sometimes seems so.

@Manuel Thank you about tilda, maybe that’s where my former maths teacher got it from.

@ArthurReutenauer Yes, I just took his answer as official, nothing more.

@Manuel I think that’s a mistake. Hans doesn’t make official statements, he just expresses his opinion. Obviously since he’s the main implementer his opinion counts a lot, but it doesn’t usually mean that it is the definitive answer about what ConTeXt should or shouldn’t do, especially for the “shouldn’t do” part - if someone comes along and implements a new feature, Hans is generally happy to integrate it in ConTeXt.

@Manuel OK, maybe not actually happy ;-) But generally speaking he’ll do it.

@ArthurReutenauer This is a misunderstandment :P My only comment to Will was that may be he coul look about how has ConTeXt sorted out this kind of problems although, hence my remark, it's not completed, is somewhat temporary solution that might need a rewrite in the future (but since they don't need math at this moment, well, there's no hurry).

@Manuel Yes, I understand what you meant now, I think we agree with each other. Also, misunderstatement is a funny word :-D

@Manuel @PauloCereda will giggle about this line ;-)

The following example give (slightly) different output when compiled with LuaLaTeX (height of the optional 3), XeLaTeX (line thickness) or pdflatex. Is that enough to report to someone?

% arara:lualatex
\documentclass[border=10pt]{standalone}
%\usepackage{lmodern}
\usepackage{fontspec}
\usepackage{unicode-math}
\setmathfont{Latin Modern Math}
\begin{document}
$\displaystyle
\sqrt[3]{\frac{%
		\frac{1}{1}
	}{%
		\frac{1}{1}
	}} $
\end{document}

Or a different question, can anybody confirm differences in the output?