« first day (1444 days earlier)      last day (2772 days later) » 

7:10 AM
@tohecz Foggy morning here. :(
@egreg It's sunny here, but I'm quite sleepy. I went to bed at 0:40, woke up at 1:10, and decided to do something since I was unable to sleep. Now it pays off :(
@egreg and moreover, I've got to use a bad LaTeX template for a while, which sucks :(
@tohecz Much worse than sleep deprivation.
@egreg yep. It sucks if a quite top-level bunch of journals uses so bad stuff :( In the manual, eqnarray* is mentioned, but not with the warning not to use it :(
status update from saxony: no fog, no sun, but lots of clouds
Good morning and meow to all :-)
7:19 AM
@Johannes_B hello :)
Good Morning from saxony too :-)
@egreg you really recommend using \colon in \delta \colon A\to B, x\mapsto x' ?
@tohecz I only use \colon in that context.
@egreg ok. And the 2nd thing: thin spaces in X=\{\, x\in A : x<1 \,\} ?
@tohecz Knuth recommends it; I usually comply.
7:24 AM
@egreg even in this case? (gimme a minute please)
@tohecz But \colon instead of : has a very precise objective (not making the colon into a relation symbol); the thin space is optional, provided the usage is consistent.
^^ I know you wouldn't make the outer () and \{\} big, but it doesn't change the fact that this looks ... odd to me
what is this quirky 2 for a sign? :o
@Serthy what? you mean \mathcal{Q}? :)
@tohecz It looks odd because of the alignment to = that hides the main part, that is the names of the sets. I guess a \left \right are being used, so you get two thin spaces between { and (.
@tohecz Train arriving in two minutes. See you later.
7:29 AM
@tohecz yes that thing, thanks
@egreg no, there is no \left\right
@PauloCereda Should've thought of that! Thanks :-)
@Serthy A fellow saxon who doesn't recognize a Q?
well, you got me
»In PSTricks we trust«
57350 Bounty spent
57599 Rep
0.9956770082813938 Fraction of rep spent
7:55 AM
@PauloCereda That works, thanks :-)
1 hour later…
9:14 AM
@egreg And do you put any spaces in \{0,1,\dots,b-1\} ?
9:34 AM
@tohecz No.
9:50 AM
@egreg ok, thanks
10:44 AM
@Johannes_B (s)he's currently spending bounty on a very worthy cause
I am using Texmaker editor. On the (Source Viewer) window when I want to jump to the editor side, it does not work with me, (by doing right click then “Click to jump to the line Ctrl+Click”).
How I can make this command work with me?
11:27 AM
@Jake Do you know how to upsample data in a convenient way? I created a colormap and used it on the addplot color based on sign question. But it keeps track of the last known point (obviously). Do you have any ideas on that? Here is some code
colormap={binary}{[1pt] color(0pt)=(red); color(60pt)=(red); color(61pt)=(blue); color(150pt)=(blue)}
-3.5 nan
-3.375 nan
-3.25 nan
-3.125 0.000121769833828
-3.0 0.000901599938148
-2.875 0.0014834598325
-2.625 0.00588102803508
-2.25 0.0720113632674
-2.125 0.0314393757219
-1.875 0.0306933267171
-1.75 0.07029144932
-1.375 -0.0342012822215
-1.125 -0.0171975784671
-0.75 -0.0405296273749
-0.625 -0.0330713662613
12:22 PM
Input: \* repeat (\@10@\,\xnax\)*\ batman!
Output: nananananananananana batman!
@egreg: my epic macro expander. :)
@PauloCereda It seems very useful! :P
@egreg :P
@JosephWright @ChristianHupfer Do you guys understand the request? latex-community.org/forum/…
@Johannes_B No
@JosephWright Use my macro expander, it's easy. :)
12:42 PM
\# define (\@\xgreet(\ynamey\)\x=\wHello \ynamey\!w\@\)#\ \%greet(\aegrega\)%\
Hello egreg!
@egreg: ^^ see how easy it is. :)
12:59 PM
@Johannes_B: Sorry, no
1:16 PM
When a number of an issue is 3-4, shouldn't it be typeset as 3--4? Because MathSciNet gives 3-4 which surprises me ...
@PauloCereda l3 syntax would be more readable
@tohecz if you're typesetting yes, but if you also want to put the range into some search tool, (or it came from there) a single ascii - might be easier to manage than an endash or two --
@DavidCarlisle well, the BibTeX export could deal with that correctly :)
1:35 PM
what could be possible fail when this does not work? pdflatex -aux-directory="C:/temp" -output-directory="C:/temp" (seems that these flags were ignored)
@DavidCarlisle You are advocating for L3? Wow. :)
@JosephWright: ^^ :)
@PauloCereda I'm a member of the team (although I've more badges for tikz)
@DavidCarlisle <3
@Serthy My advice, leave the auxiliary files where they are. As soon as they are moved to another location, something will complain not to find them.
Mhh, the problem is I have to work on a server from now on and it takes 300% of the time now to generate a pdf. Does it help if I wouldn't delete all output files after a pdf is generated?
1:46 PM
@Serthy Little hint: adress people by adding the @-sign in front of their names (you can tab-complete).
@Serthy LaTeX needs those auxiliary files in the next run (to get a toc/lof/lot ...).
@Serthy There are tools like latexmk or arara that can compile the document with the needed steps (one call of arara can lead to multiple calls of latex) and later clean the directory.
2:06 PM
i need help drawing a Koaxial cabel. who can help me?
@MomoTh 'lo!
i need help drawing a Koaxial cabel. who can help me?
@MomoTh You posted the question. You can improve it by making the code to a complete and compilable working example.
i posted i tried i improve every time but i need help... i have to give my documentation tomorow but this cable isn't ready... please
@MomoTh oh I have seen the question. Well, it quite fits into the category "do it for me". The point is, we can "do it for you", but this way, you don't learn anything, and that's something we don't like.
2:09 PM
@MomoTh But the important thing is to be patient. An answer takes time.
it's not do it for me... i tried 1 Week and i read every thing tried tried tried but no solutions...
@MomoTh If the deadline is that near, two questions:
1. Why are you coming up with this now?
2. Can't you just include the picture in your document?
that's why i m here in chate
@MomoTh well, in all honesty, is there any difference between somebody drawing the cable for you, or you taking this wikicommons CC-BY-3.0-licensed (i.e. free to use) picture? commons.wikimedia.org/wiki/File:Coaxial_cable_cutaway.svg
i will include the picture... but i tried 1 Week with TIKZ and the bes what i could do i posted it... i wann do it on my self thats why
dude i wanna the picture like that to learn from it!
ofc i will do later my own cables!
2:12 PM
@MomoTh Well, a starting point for you then could be something here: texample.net/tikz/examples
i just wanna know how to do this Cylinders and how to fill them how to cut a piece... it is not what i will put in my documentation
i tried in 1 Week all i saw all i asked all... but my friends my teacher and a lot of people couldn't help me... that's why i m here
do you understand the point?
the picture from wiki is just an example... do you wanna see what i wanna draw??? you will not believe it :D
@MomoTh So the question is »How can i draw cylinders?«
not really
this is my Cylinder
\tikzset{Cylin/.style={cylinder, draw, cylinder uses custom fill,
cylinder end fill = black, cylinder body fill = black, minimum height =3cm,
minimum width = 4cm, opacity = 1, aspect = 2.5}}
\node[Cylin] (#3) at (#1,#2) {};
the problem if i have a lot of Cylinder how can i cut fill ... it mus be real
how can i see it really 3 D
@MomoTh I do understand your point. However, understand that you won't really learn anything if we give you the solution straight. You need to get to learn the techniques.
@MomoTh there are true 3d tools for TikZ, but I'm sorry, I don't really know them :(
Mannn... i need the solution to learn from it
2:16 PM
@Johannes_B: Elternabend ... wenig Zeit für TeX.SX :-(
hhhh na ja ned wirklich
kannst du mir mal helfen?
@MomoTh As i said before, you need to be patient.
spar dir diesen Kommentare bitte
@MomoTh There are a couple of guys here (maybe ten or twenty) that will take up this task. Not everyone uses TikZ or PSTricks.
@MomoTh you can edit your last message by pressing the up-arrow ↑
i m waiting... its important to me to learn this cauz i spent one week trying
2:20 PM
@MomoTh well, I think that now you need to turn your code into a full minimal example, then you'll quite likely get a helping hand. But as @Johannes_B says, it may take some time. And in general, we aren't really happy with people saying "I need this and that quick please", which is exactly how your question originally looked like.
@tohecz can you please stop complaining??? i dont really care if you are happy or not! i need help if you can help than do it if not than be quite please
Hi everybody! I got another one of those pgfplots "Dimension too large" questions... I read about 15 questions and answer about the topic, but I really cannot find the problem. Should I still ask this as a question or look for another place to find help?
@MomoTh well, you may not care if I am happy, but I suppose that you care if all the other people that are able to help are happy and willing to help. Period.
@tohecz i dont care if you or the other people are happy or not. every one is some times happy some times sad! i m sad now i need help if you help me i would be happy if not i will be happy too cauz i tried my best! you are maybe sad now cauz i need help but you will be happy dont worry
@PaulPaulsen Did you check if you acciddentaly run towards infinity at some point?
@PaulPaulsen Try to play around with the domain a bit, the pgf internals round some numbers a bit strange.
2:31 PM
@Johannes_B I wouldn´t know where. I don´t do any calculations (pgfplots sure does, but I don´t tell it explicitly). All the axis limits and data points look right...
@Johannes_B I also tried "restrict x to domain"
bye guys see you and thanks
@MomoTh Sorry, I was just trying to give a well-meant honest advice. If you don't want my advice, it's your choice. But by not listening to it, I don't think you get any progress. (And this is my last message to that, this discussion starts to be pointless and it's on the edge of a troll.)
@PaulPaulsen Well, you can get this for some divisions. It's very difficult and almost impossible to debug, and it's one of the few things that make pgf a bad engine... :-/
I know my number range is touching, well, sensitive domains. But I successfully compiled a lot of similar figures...
I have [xmin=735784.474, xmax=735784.635] :/
@PaulPaulsen Please have a look at texwelt.de/wissen/fragen/2364/…
@PaulPaulsen have you considered simply re-scaling the numbers by 1000?
2:38 PM
@tohecz Making them bigger or smaller? Because if I make them smaller, the relative differences between the data points probably get to small, don´t they?
@PaulPaulsen ah damn, I didn't notice that the range is actually so short! Well, that's the problem. Then: have you tried shifting the data to be around zero? If that works, then the question is only: how to print the labels correctly ;)
@PaulPaulsen This is an interesting problem. You should prepare a thoughtful minimal example and ask a question on the main site. It might be helpful to other users.
@Johannes_B I will prepare one.
@tohecz I will try this later
@PaulPaulsen well, it's just an idea, coming from the fact that the problem here probably is that the ratio <step in the graph> : <the absolute position> is a technical 0 in pgf.
3:00 PM
@tohecz When I strip the first part of the numbers, so they result in 0.xxxx it works. So probably the numberyou suggested exceeds the upper limit of TeX.
I will post they question, if you want to answer it please feel free to do so. Otherwise I will add a wiki-answer later.
@PaulPaulsen I won't answer it. I think that the true question is: "How to re-add the prefix 735784 later?"
@tohecz Well, you could see that as a part of the answer ;)
It should be noted that all calculations must not exceed ±16383.99999 at any point, because the underlying computations rely on TeX dimensions.
@PaulPaulsen So jap, the number is too large. But i am with @tohecz that relabeling the axis is the interesting part.
@Johannes_B the @ doesn't work in multi-line messages :) @Paul
@tohecz Thanks, good to know :-)
3:06 PM
@tohecz It did for me ;)
@Johannes_B But I think an answer suggesting to change the datapoints should also adress this. In my case its not that hard, because I label them manually...
Probably x coord trafo or x coord inv trafo are the things to look for, since they also should take care of the ticks...
Q: "Dimension too large" while working with serial date numbers

Paul PaulsenI export plots using matlab2tikz and pgfplots from MATLAB to TeX. Most of my plots contain "serial date numbers", which are the days since January 0, 0000. As you can guess, the numbers get quite high, and the difference between the data points is relatively small. The following sample raises a ...

@tohecz I got the ping. :)
3:21 PM
@PauloCereda oh my :)
@PauloCereda @tohecz so did I! :-)
@PaulGessler ooh this is getting interesting!
I tried \pgfplotsset{
x coord trafo/.code={\pgfmathparse{#1-735784.4}\pgfmathresult}}

and I got some result that looks better, but it still raises the "Dimension too large" error.
@PaulPaulsen Use a different FPU?
@JosephWright FPU?
3:28 PM
@PaulPaulsen Floating point unit
@JosephWright @ChristianHupfer News: latex-community.org/forum/viewtopic.php?f=19&t=25237
Hm, the problem seems to be the x-limits and the xtick-positions. If I remove both, it gets plotted right (but with bad labeling).
However, if I leave them (both for the original values and manually transformed ones), I get the error.
I also tried some different numbers...
@DavidCarlisle Would you recommend to use package blkarray?
I'm having issues with the exam class and the package newicktree. For some reason, tree's aren't positioned correctly on the page when they're within the solution environment. See MWE here if you feel like poking at it: tex.stackexchange.com/questions/207131/…
3:47 PM
@Gregory Leave a blank line to start a new paragraph.
@Johannes_B That did it; thanks!
I posted a first answer. Still couldn´t find a solution for limits and labels, though
@Johannes_B If you'll post an answer I'll accept it, or I can post the answer myself. Up to you.
@Gregory You can post an answer.
@DavidCarlisle Looking at lines 10/11 of blkarray.sty answers my question ;-)
4:02 PM
@DavidCarlisle Hey, blkarray.sty --> very cute package, easy to use (in most cases) - are there some updates planed? EDIT: Oh..... double-post: chat.stackexchange.com/transcript/message/18148387#18148387 ;)
Ok, I will leave. Thanks for the help, @Johannes_B and @tohecz ! Maybe I can update the question soon.
@Johannes_B I wrote it over a couple of weekends last millennium, I think it does something, but I'm not sure what:-)
@egreg Just saw your response to my comment. Apparently I don't the first thing about LaTeX. What should I be doing instead of \setlength{\parskip}{0pt}?
\usepackage{amsmath, amssymb, amsthm, dsfont}



a =
 & * & * & * & * & * \\
  * & a_{11} & * & * & * & * & * \\
  * & a_{21} & * & * & * & * & * \\
  * & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots \\
  * & a_{i1} & * & * & * & * & \leftarrow i\textbf{-te Zeile}\\
  * & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots \\
  * & a_{n1} & * & * & * & * & * \\
4:13 PM
@AndrewCashner The usual value is 0pt plus 0.1pt
@DavidCarlisle This works very finde - super package
@cis I'm always surprised if people get output from one of my packages.
@DavidCarlisle So, I am sure there will come an updatet version in the future ;)
@cis well in the end we just get @JosephWright to implement it in LaTeX3
4:16 PM
@egreg If I'd read your other comment more carefully, I would have seen that. Sorry, and thanks. Will try to go actually learn something.
@egreg I'll delete my comment
4:48 PM
@DavidCarlisle Probably
@DavidCarlisle Or Bruno
@JosephWright or even me:-)
@DavidCarlisle xor
@JosephWright oh yes, that
@DavidCarlisle Will you make UK-TUG 2014?
@JosephWright not sure yet, will let you know
5:02 PM
@DavidCarlisle OK
@DavidCarlisle You could talk about Fortran. :)
@JosephWright: Prod Nicola to talk about her new book and arara. :)
@PauloCereda I'll ask her if she's coming when I next see her
@JosephWright :)
@JosephWright: Where will the meeting take place this time?
5:38 PM
@PauloCereda or python
@DavidCarlisle Heavens no. :)
@PauloCereda Oxford: same place you presented
@JosephWright ooh how nice! :)
@PauloCereda UK gvt is getting all secondary schools to teach a "text based programming language such as python" which means that most of them seem to be using python (new cs curriculum starting this year at all age levels)
@JosephWright: How many keynotes so far?
@DavidCarlisle Dear Lord. I've seem some CS curriculum stuff being applied to Brazilian universities and IMHO it's quite atrocious.
Courses like compiler construction is totally removed from the schedule.
5:43 PM
@PauloCereda Erm, one
They are also ripping theory apart.
@JosephWright Oopsie.
Poke me if you need help.
@PauloCereda I'm sure we'd welcome an update via Skype
@PauloCereda well previously "ICT" had become more or less "Word and powerpoint" so it's got renamed back to CS and a requirement to teach some coding, python's not so bad, probably
@DavidCarlisle Indeed. I quite love mocking Python. :) What worries me is that a lot of important concepts get lost in the process.
@David: speaking of which, I will try to take a picture for you in my trip to SP this Friday. I'm sure you will love it.
1 hour later…
7:04 PM
Hello everybody :)
@HenriMenke hi there
Another expl3 question from my side. I'm trying to implement a macro for easy typesetting of tensors in relativity. My thought was to use \tl_replace_all:Nnn to replace all the ^ and _.
\cs_new_protected:Npn \tens_format:nn #1#2
  \tl_set:Nn \l_tmpa_tl { #2 }
  \tl_replace_all:Nnn \l_tmpa_tl { ^ } { \vphantom{#1}^ }
  \tl_replace_all:Nnn \l_tmpa_tl { _ } { \vphantom{#1}_ }
  \tl_put_left:Nn \l_tmpa_tl { #1 }
  \tl_use:N \l_tmpa_tl

\NewDocumentCommand \tens { m m }
  \tens_format:nn { #1 } { #2 }
But apparently the second \tl_replace_all:Nnn doesn't do anything.
I get double subscript error.
Okay, it's because of the search string being _. Everything else get's replaced as expected. But I still don't know how to resolve this.
@HenriMenke ah, yeah, well, the problem is catcodes, I know a LaTeX2e solution to this, but that's not something you want
@tohecz Sorry! you are right... i had just a bad day! i m sorry
7:19 PM
@MomoTh No worries really, apologies accepted :)
@tohecz are you actually a moderator or a simple friend here :D
@MomoTh I'm just one of the folks here
@tohecz i like you!!!! you are so nice
@tohecz and by the way: a simple quastion: do you know what is that: \documentclass[border=5pt]{standalone}
hint: moderators (from all sites) have their name in blue here, and they have a diamond next to their name in the site itself
@HenriMenke We've not got a good interface for this sort of thing yet, so
\char_set_lccode:nn { `\@ } { `\_ }
\char_set_catcode_math_subscript:n { `\@ }
    \cs_new_protected:Npn \tens_format:nn #1#2
      \tl_set:Nn \l_tmpa_tl { #2 }
      \tl_replace_all:Nnn \l_tmpa_tl { ^ } { \vphantom{#1}^ }
      \tl_replace_all:Nnn \l_tmpa_tl { @ } { \vphantom{#1}@ }
      \tl_put_left:Nn \l_tmpa_tl { #1 }
      \tl_use:N \l_tmpa_tl
7:21 PM
@MomoTh yep. Standalone creates a "standalone thing", like a drawing, formula or whatever. And border=5pt means that the resulting document (PDF usually) will be just 5pt larger than the contents
@MomoTh I, on the other hand, am a mod :-)
@JosephWright ah yeah, the l3 interface to g'dol' tricks :D
@JosephWright and he's blue :) @MomoTh
@tohecz coool... hello friends
@MomoTh Hello
@tohecz Indeed: we want a better set up for this, it's just that we can't quite agree on the plan!
@tohecz did you see the answer of my quasion? he helped me to draw my cylinder. but i can just put it in apdf alone(\documentclass[]{standalone}). can i put the picture on my document
7:24 PM
@JosephWright Thank you very much! But isn't that just the \lccode trick in expl3 syntax?
@HenriMenke Yes
@HenriMenke We have plans to wrap it up in a better interface
@HenriMenke There is no good way to avoid this in some places in TeX
@JosephWright And maybe change the documentation, because for \tl_replace_all:Nnn it says that only catcodes 1 ({), 2 (}) and 6 (#) won't work, but apparently that's not true.
@HenriMenke No, that is quite true
@JosephWright But why didn't it work for catcode 8 (_) in my example then?
@HenriMenke The issue is that as set up your _ has catcode 11, but you are presumably searching for catcode 8 tokens
@HenriMenke Remember expl3 uses _ as a 'letter'
@HenriMenke If you arrange to pass _ as a catcode 8 token to \tl_replace_all:Nnn it works fine (which is what the lowercase trick does)
7:29 PM
@JosephWright Oh god, how could I forget that! I'm using it all the time. Aarg!
@JosephWright do you maybe know it?
Man do I love me some longtable these days. That package is awesome. — henry 9 mins ago
@DavidCarlisle I wonder who wrote that package. Maybe @egreg was the author. :)
@MomoTh ?
7:45 PM
@MomoTh Either use the code directly in your document, or include the PDF generated by that standalone document with \includegraphics.
@MomoTh that's exactly what you can do, yes. Using \usepackage{graphicx} and then \includegraphics{myfile.pdf}
@PauloCereda can't be him, it has no bugs
@DavidCarlisle ooh.
8:03 PM
@DavidCarlisle Ever tried to have a longtable starting in the same page as a float? Oh, sorry, that's a feature!
@egreg it works sometimes
visited 1000 days,
@DavidCarlisle "sometimes" seems to be your motto there, doesn't it? :D
@tohecz better than never
@DavidCarlisle Yay congrats!
8:30 PM
@DavidCarlisle 1301 days, 1301 consecutive. :P
@egreg 73 consec
@egreg, a nice round number I'm sure you will agree:
8:48 PM
@TorbjørnT. i tried to use my code directly but i shows an error
@tohecz but i wanna use directly the code in my code ^^
@DavidCarlisle Unmerited bounties, surely.
9:27 PM
@egreg s/Unm/M/
9:50 PM
@DavidCarlisle I'm no expert in sed.
10:19 PM
@egreg same as TeX then
10:51 PM
@DavidCarlisle Should I flag this one? ;-)
@egreg you could star it, if that's what you mean.

« first day (1444 days earlier)      last day (2772 days later) »