@leo did you update from admin command prompt ? see texdoc biber there are ways to test `biber test' using testfiles. check for 'TL bin path' in user account PATH.
@MarcoDaniel You did \let\@classoptionslisttemp\relax
What then happens is that your first \edef leaves \@classoptionslisttemp inside \@classoptionslisttemp, so the second one tries to expand the macro itself
Our question LaTeX Editors/IDEs is very popular, with close to 35k views as of Feb ’13. However, the answers vary a lot in terms of quality and the kinds of information they provide, which makes comparison harder than it would have to be. So I thought it might be nice to standardize the informati...
I want to show the text from a Python Interpreter in a Windows console. This should be plain text (like you see in a command line terminal). I also want to show Python code snippets with highlighting. To do this, I set my default listing style for coloring (\lstset) and then created a custom list...
I just tried to compile a Tikz file with around a million points in two plots (500k in each). Could this cause TeX capacity to run out? Because that is the message I'm getting.
@egreg I tested the above code for myself. It works for me.
@egreg The problem was that the :NN test simply took the first two tokens from the key, compared them (usually with negative result) and skipped to the recursion tail.
Heh. And even if the result was positive, you would still encounter \prop_map_break:, so that doesn't really matter. ;-)
@PauloCereda Psmith, the TeX bot: Good morning! I'm Psmith, the friendly TeX bot - the p in my name is silent, as in pshrimp. I'm here as a companion to our fellow users in the typographic land. As you probably noticed, I always reply under Paulo's account, but do not despair, I say, my replies are always preceded by my own name. Enjoy your stay at TeX.sx! If you need any help, just ask our chat residents. Cheerio!
I use lualatex and utf8 source. I want to use the Latin Modern fonts (including math). If I include e.g. the character ε in my source file; it does not appear in the output. I've tried changing the font to TeX Gyre Pagella (with fontspec) and the ε appears as I want. I assume the Latin Modern fon...
@Brent.Longborough No: I only use SVN nowadays for LaTeX3
@mhelvens Fix to CTAN once lualatex-math gets updated: there is a change in expl3 that will break the current lualatex-math release, so I need that to be updated on CTAN first
@JosephWright The hg learning curve is quite a bit better than git 's, I think, and I've never managed comprehensively to destroy a repository (done that with git, tho)
I tried asking earlier, now asking again. I've got a tikz file, generated by R's tikz driver, which has two line plots of 500k points each. TeX is giving a memory error, probably caused by this. Can anyone confirm this could be a problem?
I wonder why I'm constantly warned about the linguistic issue about the s/he ligature but not getting an actual ligature proposition which seems much easier to propose ...
@Brent.Longborough Hahaha, maybe bandits (haydutlar) would be more dramatic
@FaheemMitha Was it you that we have discussed the resolution issue with?
@FaheemMitha That it's very unlikely to display every point on PDF so you can just reduce the number of points :) But as egreg says LuaLaTeX can handle more memory demanding issues
@Brent.Longborough Yes they are but disgust is temporal. They might become addictive for future generations. As ienissei commented to my ligature question; (the only characteristic ugly fonts all have in common is randomly angled lines)
@percusse Hmm. I think the more math type journals do that. i think when i submitted to the journal of statistics and computation or something like that, they compiled it
Sorry, it was The Journal of Statistical Computation and Simulation
Taylor and Francis.
When biology journals see TeX they scream and run. So it depends.
But I don't have much experience in these matters. Anyone else care to weigh in?
@FaheemMitha Do you have any screenshot of what you are trying to plot? I'm curious about the figure. I'll see if I can find the past discussion we had here.
OK we had the discussion with texenthusiast before...
@percusse I guess I can post a question. I don't see anything like this on tex.sx. If I post, I'll include the image. It is fine with the R pdf driver.
@percusse So, he didn't post a question, then?
Hmm, in people's experience is it common to outsource typesetting? In my limited experience, this is the second time I've been getting emails from Indian typesetters.
@Qrrbrbirlbel It would be interesting to know why that wrong command works anyway. You were essentially doing \textit\bgroup<text>\egroup; roughly speaking, \textit{#1} does {\itshape#1}, so your input becomes {\itshape\bgroup}<text>\egroup; then \bgroup} is an empty group and \egroup matches the first open brace. A similar situation appears when you do \mbox\bgroup A\egroup, which seems to work correctly.
@FaheemMitha I wrote the class file used by the Journal for Machine Learning Research (jmlr.cls). My workflow uses pdflatex so I prefer .pdf images but I can use epstopdf on any .eps images. My major problems occur when authors submit tex files that don't compile without error as it causes my book building tool to fail.
@FaheemMitha I did some ridiculous stuff with extremely large number of points but I've printed on A0 so it was sensible to keep the data. I used LuaLaTeX for that.
@NicolaTalbot One problem with journals is that they are very restrictive about what packages are allowed. Sometimes they have no clear statement at all. It would be nice if they could just standardize on say TeX Live.
@percusse One option would be to get R to smooth the curve. It actually looks pretty smooth already.
@FaheemMitha 5000 points already comes pretty close to pixel resolution in a 5 cm linear axis plot. So it really hardly matters beyond that. After that you get weird artifacts due to downsampling.
Since down at the PGF level they are more or less \draw (0,0) -- (0.001mm,0.002mm); type instructions.
@FaheemMitha The jmlr class file is restrictive because the book has to import each article using combine. If different authors use conflicting packages the book won't compile and it can take me weeks to fix. Since I'm not paid for the work, this means that the book won't be published until I have the time to spare to work on it. (I do however list the known problem packages in the manual and the class file checks for them and issues an error if they're detected.)
@FaheemMitha I've used to different approaches depending on uniform or nonuniform intervals. One is the pgfplotstable solution by skipping every say 40 points. The other one is to give explicitly the important intervals in the domain where I use a coarse grid on smooth parts and finer grid on the tricky parts. .
@NicolaTalbot Oh, if the articles are combined into one book length manuscript, that does make things more difficult, but why is that necessary? Why not have a separate file for each article?
@percusse Ok, I see. Thanks.
@percusse I'm using the ggplot2 ecdf function, which doesn't do smoothing. I think the regular R one does. Not sure what I can do with that
@FaheemMitha I developed the jmlr class because I'm the production editor for the Challenges in Machine Learning book series that includes articles that have been published in the JMLR W&CP. The articles in the CiML book have a different page size to JMLR and have hyperlinks and cross-references between articles, and that's not easy to do if you just munged a load of pdfs together.
@NicolaTalbot Ah, I see. Yes, then conflicts between different packages can make things tricky. However, I don't think that is the case with most journals. I.e. they don't have cross-references etc between articles.
@FaheemMitha I don't know about the production process for other journals, but I know that many of them rely on volunteers who may or may not have a great deal of LaTeX expertise (or even an up-to-date TeX distribution).
@egreg Uh, it took stackexchange long enough to notify me about your message. But yes, I haven’t thought about what happens with \textit\bgroup. Though, we can break this easily with a definition like \def\textit#1{#1#1}.
@FaheemMitha No but they may have other constraints, a common one being that they convert the document to sgml/xml for archiving or post production or whatever, converting tex to anything that isn't tex gets hard very quickly if you allow authors to do anything other than write words.
@loverofstructure Re your flag about 'regaining control' of a question. There's nothing we can do: you'll need to ask the StackExchange staff. See tex.stackexchange.com/help
@NicolaTalbot Well, all the typesetters I've dealt with didn't seem to know much about anything. On one occasion I had to explain their own rules to them. On another, I had a glitch in a style file. I foolishly assumed the journal would know how to fix it, and fruitlessly passed messages back and forth for several months. Finally, I took it to tex.sx, and I got a dignosis and fix (if memory serves) in about half an hour.
@NicolaTalbot And these were commercial journals. Surely they pay their typsetters, even if they don't pay anyone else?
What exactly is a clean solution?
I guess it is easier to say what is not a clean solution.
The question came in my mind when I tried to made a Matrixnorm and used at first something
like (I don't remember what i made exactly)
| \! | \! |
But a friend told me that this isn't a clean s...
@DominicMichaelis well or if you actually want to ask how to define ||M|| then ask about that (oh I see you want |||) but I don't think the current question has any answer.
I would like to create a norm-like operator with 3 verticals bars instead of two. It has to look like $|||\cdot|||$. I would like to know how to decrease the spacing between the vertical bars such that it has the same spacing as the $\lVert .. \rVert$ command.
There is a command available in the...
@DominicMichaelis well that's the question that has no answer. It isn't a defined term. Saying a solution isn't clean is just a mild rebuke to the person who wrote the code, you can't ask for a definition of that.
@DominicMichaelis it could be anything from not working, to using 1000 commands when one would do, to having been written in vi instead of emacs. It's just like asking why are some programmers better than others. There is no reason it's just life.
@FaheemMitha Some do, but good (La)TeX typesetters aren't cheap and if they charge on a daily basis the costs can mount up if there are problems, so it's hardly surprising if journals decide skimp on costs. I know that editors and reviewers generally don't get paid. @Joseph do you know much about it?
@NicolaTalbot Reviewing is as you say unpaid: my former boss does get paid (plus an assistant) to run an editorial office for a journal on a part-time basis. Certainly the publishers have staff (I have several friends in that area). Typesetting I know less about: everyone in my area uses 3B2 I think.
@DominicMichaelis sorry it is still not answerable (and seems like enough people agreed that it got closed). Don't let this put you off the site though, we are usually fairly helpful I hope:-)
@JosephWright Since (La)TeX is such a specialist subject I think there's a danger that journals might withdraw the option to submit articles in LaTeX if it causes them too much hassle.
I would like to know what the difference is between a clean solution and a "not clean solution".
So what properties does a clean solution have?
Is there something like a list ?
Or what makes a specific solution unclean?
For example I guess
\def\boxphi{\fbox{\ensuremath{\varphi}}}
is no...
@DominicMichaelis Quid ergo est tempus? Si nemo ex me quaerit, scio; si quaerenti explicare velim, nescio. I find this quotation from St. Augustine can be applied in many situations. If you have a solution before you, it's usually easy to see if it's unclean (provided it does not involve many lines of code). But defining "uncleanliness" is rather hard.
I once asked a friend of mine (he is making is PhD) how difficult a question hase to be for math overflow, he told me if you think there is an answer, it is to easy
@FaheemMitha Tend to let go some questions for others to get some rep even if you know the answer. It sounds too narcissistic(?!) but I mean in a very very casual way. Quite a few TikZ questions for example are easy enough to be answered by anybody. So no need to jump in immediately.
i want to use the \autoref command (from the hyperref package) in the document-class tufte-book.
mini-example:
\documentclass{tufte-book}
\usepackage{hyperref}
\begin{document}
\section{Testsection}
\label{sec:test}
\begin{figure}
\rule{20pt}{20pt}
\caption{Testfigure}
\label{fig...
@FaheemMitha Well it's kind of greeting every user. Might work on some. Might not on others :-) At least that's how I feel judging the answers on meta so far.
i found a hack (see my comment), but since I can't really explain it, I felt it better to leave it. I assume it's a combination of tufte and hyperref fighting for captioning/labelling rights?
@FaheemMitha It depends on the job. I don't much like working on projects where multiple authors have written dodgy LaTeX code, but I've had other typesetting tasks that have been quite fun, such as designing a catalogue for an art exhibition.
@percusse yes indeed, I thought that would work. but as in my comment, it is a good work around, but I'd be interested to know why the problem exists :)
@FaheemMitha My primary job is a publisher, so most of the typesetting I do is for my own books. I haven't put my name on the consultant list as I don't generally have time for consultancy. The art exhibitions were charity fund-raising events, so I did that on a voluntary basis.
I'm inserting a subplot created by MATLAB into my TeX document, and in the bottom left hand corner the four letters "ches" appear. There is nothing like this in the subplot, and I don't think I have any issues in the TeX document that is creating this random message. Has anyone come across this...
Im using `book` document class. I want the even pages headers: 12 Cool chapter the odd pages headers: Cool section 13 So I have done `\renewcommand\chaptermark[1]{\markboth{#1}{}}` and `\renewcommand\sectionmark[1]{\markright{#1}}` to use `\leftmark` and `\rightmark` as arguments of some `\fancyhead` commands. But after do so, I still get: 12 Chapter 1. Cool Chapter and 1. Cool section 13
\documentclass{article}
\usepackage{enumitem}
\begin{document}
\begin{description}[style=nextline]
\item [First:] Description of first item.
\item [Second:] Description of second item.
\end{description}
\end{document}
I want to know what the fonts are in this pdf. Which package to use?
What exactly is a clean solution?
I guess it is easier to say what is not a clean solution.
The question came in my mind when I tried to made a Matrixnorm and used at first something
like (I don't remember what i made exactly)
| \! | \! |
But a friend told me that this isn't a clean s...
i want to use the \autoref command (from the hyperref package) in the document-class tufte-book.
mini-example:
\documentclass{tufte-book}
\usepackage{hyperref}
\begin{document}
\section{Testsection}
\label{sec:test}
\begin{figure}
\rule{20pt}{20pt}
\caption{Testfigure}
\label{fig...
I'm inserting a subplot created by MATLAB into my TeX document, and in the bottom left hand corner the four letters "ches" appear. There is nothing like this in the subplot, and I don't think I have any issues in the TeX document that is creating this random message. Has anyone come across this...
