\usepackage may only appear in the document preamble, i.e.,
between \documentclass and \begin{document}
What does this mean and how to solve it , thanks a lot
@JosephWright last night I was too tired to code anything but I looked at the sources ...
if (fabs(det) < 1.e-8) {
WARN("Inverting matrix with zero determinant...");
return -1; /* result is undefined. */
}
@JosephWright which explains the cut off point as the determinant is the square of the scale factor, although multiplying numbers that small in tex without going to the expense of loading l3fp or similar is rather unreliable. It's perhaps reasonable for the driver to check this but in the fail case I think it ought to scale by something smal not leave the text unscaled
@JosephWright plan to make a test file along the lines of
@Sebastiano as we have said many times already you should ask a tex question preferably showing the tex code you have so far. If you just post a picture and ask people to draw it for you in tikz it will almost always be closed.
@Sebastiano so start with some tikz to make a sphere, if that fails, post what you have and ask about the error, if it works, try to add some dotted lines of longitude and if that fails ask about that error. that way your questions always have a test example to post and are asking about tex.
@DavidCarlisle Your observation is correct and right. I have seen on the TEX.SE site, on TeXample different spheres but with all my heart is not for bad will, and not for boredom or because I want to do the work to others with all sincerity. At this moment I'm also examining my students and unfortunately I'm having to do the work my colleagues should do.
@egreg Bugs are OT if solved, etc., but if not then having an answer saying 'This is a bug' is normally useful. On the matter at hand, I'm not sure if it is, though: isn't the description of the meaning of C-c C-a simply misleading? (I read it as just meaning it runs LaTeX.)
@egreg I suspect it's not a dupe, though, on this basis ;)
@JosephWright It's actually really quite hard to test from tex the exact condition where dvipdfmx will error you need to calculate xscale*yscale and check that its absolute value> 1e-8 without overflowing or underflowing tex arithmetic (and preferably without the cost of loading l3fp)
@DavidCarlisle A bit of re-ordering is needed on the L3 side (I need to back-out some code sharing and that then makes the logic a bit odd at present), so might be a day or so before I update that. See what the TL list response is then send to CTAN?
@JosephWright if I told my colleagues here that we were testing if a matrix is invertible by calculating a determinant I'd be kicked out of the building
@JosephWright as so often in numerical code you need to re-order the operations x y scale is invertible just if x and y are non zero, there's no need to calculate x*y, as you can invert just by calculating 1/x and 1/y so all you need is x and y to be within machine accuracy. A general inversion would give y/xy x/xy which mathematically is the same thing, but not if the product xy has underflowed the machine arithmetic. (even if it doesn't error you lose all numerical accuracy doing it that way)
@ChristianHupfer well... and costs around 45€ too, but I wanted to support the Lua project. The textwidth is a bit big (conversely, the character is small) but it looks already better than the LuaTeX manual
@ChristianHupfer Juventus, Fiorentina, Milan, Inter, Brescia and Bologna I think
maybe Piacenza too, but I don't know
@ChristianHupfer it basically was a pipe organ (the largest in the world) which was only painted on a church wall. The myth goes that a person told everyone that he was so good playing the organ that he could play great on any organ in the world. So they bet him there was this organ he could not play, the greatest organ in the world (Baggio's). Now when someone acts like a boaster, they say "yes, go to Baggio to play the organ" (in Milanese dialect)
@JosephWright we'll see... the online edition seemed kind of nice. I have quickly skimmed through the book and I saw that there are exercises (but no solutions!!!) that I don't like, but it's ok, I guess
@ChristianHupfer we could create one the great nonexistent organ of the fictitious city of Bielefield
@DavidCarlisle: You appear to be maintaining hyperref. Would you be inclined to accept a patch to its README.xml that turns it into (Pandoc) Markdown to make it look a lot less messy in PDF?
@wilx Ah... I was just looking into CTAN statements. Heiko Oberdiek is still listed as maintainer/author there. Sebastian Rahtz died unfortunately about one year ago :-(
@DavidCarlisle There appear to be two documents generated from the XML, LaTeX for PDF and a TXT. It is easier to get nice PDF out of Markdown than nice TXT out of LaTeX. But if you want just LaTeX for the README, it can be done as one time conversion.
I would still convert to Markdown and then to LaTeX for the one time conversion.
@wilx personally I would just have the ctan/github markdown and not have the xml or pdf, I can't see the point in a pdf readme really, but I'm keeping the setup as close as possible to Heiko's original. The readme is more or less "one time" as you say but the ChangeLog.xml goes through more or less the same process and needs to be built each time. But anyway I wouldn't want to make installing pandoc a requirement to build the package (I don't have it for example)
@wilx I don't see that the format would improve the readme at all, the source is XML and the text and latex are generated from it via a couple of simple xslt transforms, if you have suggestions for changing the format of either they could be implemented in the xslt without changing the source format, surely?
@wilx I suspect that essentially no one has ever looked at the pdf readme so I'm not too concerned about its format:-) But suggestions for improvement welcome!
@wilx the current xml is pretty dire it has to be said:
<p>Included are:
a) `backref' a package by David Carlisle to provide links back from
bibliography to the main text; these are hypertext links after using
hyperref.
b) nameref' a package to allow reference to the *names* of sections rather
than their numbers.</p>
</section>
@wilx but if I do anything I'd drop the generation stages altogether and have the changelog as a plain text GNU changelog file and the readme as a github readme.md, most likely (but other issues first like fixing some of the code issues reported....:-)
@Johannes_B OP: I does not work -- Johannes_B: What does not work? Please post a MWE and the error log -- OP: It does not work .... Johannes_B: Please read what I wrote a few minutes ago ... OP: It does not work ;-)
@daleif No seriously, I remember that I read a comment by Heiko Oberdiek rather recently on this issue, but I don't remember which question this was and what the precise reason for this limiting of expansion on the first macro only is.
The file name parser that is used by package graphics requires that the extension is not hidden inside a macro. Only a macro at the start of the image file name is expanded once. — Heiko OberdiekMar 31 at 10:33
@UlrikeFischer in this case it did contain an extension. It is the result of a generating function. But we could just specify it without it. But why not with the extension?
@daleif @ChristianHupfer showed the comment from Heiko: \filename@parse (in the latex kernel) does only a expandafter. So with your two commands the extension is hidden. I don't think that one can avoid it. An \edef on a file name which could contain e.g. active chars would probably lead into disaster.
@UlrikeFischer in this case the user is under strict instruction that only ascii is allowed and no spaces in filenames. ;-) (this is auto generated reports)
@UlrikeFischer exactly and then it is not my problem. They user is just glad they can do this in a simple manner and not using Word (they are not LaTeX users)
I was looking at Bruno Le Floch's "unravel" package (see tex.stackexchange.com/questions/61010/… or texdoc unravel) -- it appears it implements large parts of TeX from within TeX (!). Question about LuaTeX: would this be significantly simpler to implement in LuaTeX?
@wilx oohh.... having done that you've probably read it a lot more carefully than me, my impression is that it's not really a "readme" at all more some kind of appendix to the hyperref manual. I was wondering if it should be cut back a lot with the more detailed stuff being added to the pdf manual, either a new document or added inside hyperref.dtx?
@DavidCarlisle I see, thanks. I thought LuaTeX was written to make things like this easier (see what's going on inside TeX). Maybe they just haven't seen a need to implement callbacks for this part of TeX yet I guess.
@ShreevatsaR all the callbacks really come from "the stomach" in texbook terminology, you can not from lua really see a macro (or create macros otehr than simple ones without arguments)
@DavidCarlisle I was trying to understand whether tabulary offered something that I would use more frequently than tabularx and finished in the code of the former, and I knew that some \Z or \z variant would be there :)
@DavidCarlisle Speaking of coloured background, the paracol documentation is a tour de force. See section 10 for some competition to the colortbl docs.
@DavidCarlisle yup, the problem is that I learned to talk a different way, due to American TV shows. I go back to English in Snooker season, thanks to BBC
@DavidCarlisle indeed
@DavidCarlisle anyway at the end of the final I started understanding Higgins… should I be worried?
@DavidCarlisle I'm curious whether I'll be able to understand people in London this time, and whether they will be able to cope with my terrible italian-american-jibberish accent
@Moriambar Like when the “official” guardian of the parking lot at the Piscine in Catania told me something about my scooter (or so I believe) and I couldn't understand a word. And neither did my local friend.
@egreg yup My parents went to Sicily once and weren't able to understand a word in a town near Modica: the locals tried Italian first and dialect the second time around, but no way.
@egreg Around 10 years ago on a local (i.e. from around Varese) newspaper, an article was published about some guys from Napoli trying to rob a bank, but they couldn't get what they wanted since nobody understand their dialect (and the people at the bank were able to call for the police while the robbers tried to explain what they wanted exactly)
@egreg maybe the article was forged. I know a guy who worked there for the sunday edition and told me it was kind of normal, in order to fill the pages, to forge some articles here and there…
@Sebastiano voting is arbitrary enough as it is without you making it worse. the idea is that you vote for answers that you think are useful not that you pick some people and randomly giv ethem a few votes.
@DavidCarlisle I do not have time at least for now. I have to correct some of my students' tasks and try to give a hand to the bricklayers before tomorrow they go to school.
@JasonBourne the default one, of course. But @PauloCereda pointed me to a font that I was trying to install on my work pc and whose name I don't remember…
LaTeX is so not designed for this kind of document. But if you must, must.
\documentclass{article}
\usepackage{graphicx}
\begin{document}
\setlength\unitlength{1cm}
\vspace*{\fill}
\noindent
\begin{picture}(0,0)\bfseries\large
\put(4,3){\includegraphics{house}}
\put(0.5,5){\parbox{5cm}{\ragge...
@Moriambar A lot of people think that recognize is the American spelling and recognise is the British spelling, but recognize is also used in Britain as the Oxford spelling.
@JasonBourne well only because the OED has (always?) not distinguished ize endings as American as a matter of stated policy, I think most people still think of ize as American spelling
@DavidCarlisle Oh, I think you just need your library card or something.
Talking about spelling, it should be 'just deserts' and not 'just desserts', a common mistake. But if a mistake becomes too common, it might become acceptable in the long run.
@Moriambar Haha. Well ok, the Sound of Music and The Simpsons then. :) That's probably an exaggeration, but I don't think it's really used much at all these days.
@AlanMunn nope. I learned most of it through FRIENDS, the TV-show. howdy is featured there (once), and since I own 5 cowboy-style hats, I find it well suited for me :)
@AlanMunn and remember: I watch spaghetti western in Italian. Apparently the only films I do watch in Italian now, but still!
Well, I think I'm gonna call it a day. It's been nice, but I gotta go. So long!
@DavidCarlisle ps: my battery is fine, I just wanted you to know that :)