« first day (496 days earlier)      last day (3039 days later) » 

12:00 AM
@DavidCarlisle You have a very good trend. :)
broke the trend today though, either I wasn't very inspired or the questions were rubbish: I couldn't summon up any interesting answers
@DavidCarlisle There's the Batman question. Or the "local footnotes". :(
12:17 AM
as I said:-)
@egreg Reformatting footnotes is very annoying. Someone needs to write a nice interface to the footnote mechanism that goes beyond footmisc.
I really must get round to looking at tikz one of these days,
@DavidCarlisle That will make your reputation line go completely vertical.
It's the biggest rep bang for the buck.
@AlanMunn you mean if I lean enough tikz I can draw my own graph?
@DavidCarlisle Hehe. That too.
12:21 AM
So I can leave TikZ to you. Good night.
@egreg Buonanotte. :)
12:33 AM
@AlanMunn Indeed I have drawn two arcs and things went crazy.
12:57 AM
@egreg: should we write a checkurls script too? :) tex.stackexchange.com/questions/47476/…
9 hours later…
9:57 AM
Currently, we have three Batman-related questions. Can I add a tag? <3
1 hour later…
11:24 AM
@egreg: I have a funny solution for the IEEEeqnarray question:

    &p(t) = \sum_{k=1}^M a_k \cos(k w_f t + \phi_k) = \sum_{k=1}^M h_k(t) \label{eq:pli} \\
      &    h_k(t) = a_k \cos(k w_f t + \phi_k)  \label{eq:plih}

every second pdflatex-run it works ;-)
@MarcoDaniel Actually David's solution works, but a \noindent should be added: \noalign{\noindent where} or, perhaps better, \noalign{\hbox{where}}
It does because ultimately the environment reduces to an \halign to \displaywidth
@egreg but noalign produces and extra line. ?
I assumed he wanted an extra line (I wondered about \noindent but really you need to fix up the vertical spaceing to undo the fact that tables/alignments are usually set with no lineskip, and if it's just one word, probably not worth it...
@MarcoDaniel Yes, of course. It would be wrong putting "where" at the left margin of the second equation.
@egreg Yesterday I learned -- everything can be requested ;-)
11:39 AM
@MarcoDaniel "Customers are always right", but they end up with what the vendor sells them. :)
OTOH, I see no reason to use an IEEEeqnarray environment in that case. Two equation envs would be just right.
@egreg I prefer the command shortintertext provided by mathtools
@MarcoDaniel It tries doing the adjustments David was mentioning
@MarcoDaniel yes as I say that one (and the ams version) take a lot more care over vertical space and penalties and stuff, but if you just... oh @egreg just said same
@DavidCarlisle It allows multiline stuff too
11:54 AM
@egreg yes but the chance that tex's line breaker needs to hyphenate "where" and put it over several lines seems slim:-)
Do either of you see any white gaps in tex.stackexchange.com/questions/47492/…
@DavidCarlisle At the largest magnification I can see one. But it depends on device resolution, I guess
I remember worrying a lot about that sort of thing when I dd the original colour support but drivers (and pdf tex engines) seem to do a pretty good job it seems of snapping the panels to the right pixel alignment normally.
Friends, what's the best way to add a space before "script"?
\newcommand{\version}{Version 1.0g from March 7, 2012.}

\title{The \checkcites\footnote{\version} script}
Without \footnote, I was using \checkcites\ .
@PauloCereda \checkcites{}
@egreg I tried, but the * from \footnote "break" the space.
I added a ~ after it, but I was thinking if there's a better way.
12:02 PM
@PauloCereda I would you \xspace
A: \newcommand and spacing

Marco DanielYou can use the package xspace. To avoid the known bug you have to use \usepackage{xspace}[2006/05/08] \makeatletter % usually \check@icr is \@empty and therefore % not appropriate for \xspaceaddexceptions. \begingroup \text@command\relax \global\let\xspace@check@icr\check@icr \endgroup \xsp...

then if it doesn't work you can blame me :(
@PauloCereda The footnote marker in \title is set with \rlap; so \checkcites\footnote{...}\ should do
@egreg Uh-oh, no space added. :(
@egreg using rlap is clearly nonsense, it assumes I think that the foonote marks all come from \thanks at the right hand edge of a vertical list of author names.
hmm what's the easiest way to get 8 rep... if palindromes are what we strive for....
12:49 PM
@lockstep that one (or at least its accepted answer) is about document-wide class option rather than allowing for scalable arbitrary font selection at specific points, I think?
@DavidCarlisle Indeed. I removed my erroneous comment.
@PauloCereda The \checkcites\footnote{\version}\ \ script or The \checkcites{} script\footnote{\version}
As David says, the \footnote is supposed to fall at the end of the title
@egreg Oh my, fantastic! Thanks! :)
1:57 PM
Hello everyone!
I am looking at this solution to highlight my name in a reference list. The problem with the string replace approach is that I share a last name with another author. The problem with the first approach is that it messes up sorting by author names. I do not want to learn and switch to biblatex at this point. What other options do I have?
I'd prefer the first solution. Is there any way I could influence the sorting?
@Szabolcs I guess that the best thing is to ask a "follow up" question with some data to play with
@egreg You have a lot of experience with TeX. Do you think that the minimal effort solution here is to just bolden the names manually in the bbl at the very end? If not, I'll post a question in a while.
@Szabolcs For a one shot job, doing it by hand is probably the cheapest route
What are the two names?
I was stupid, I did not realize that it was possible to do the replacement on the full name.
It is okay now, thank you. I just need to make sure I use \'{a} vs á consistently, so the replacement can work.
2:36 PM
Hi all. I've a lua-related question. Can I ask it here?
Sure. :)
Not sure if we know the answer. :)
Cool :) Then I'll go to stackexchange, but I feel more comfortable here.
hehe :)
This is a `loadstring` related question. I don't understand the following. `print(({[0]=1,[1]=2})[1])` works and prints `2`. Then I'd like to use `loadstring` to perform the same kind of thing. So `f = loadstring('({[0]=1,[1]=2})[1]')
` Then `print(f())` and error: `attempt to call global 'f' (a nil value)` ???
Hm let me check it (I'm learning Lua too). :)
2:49 PM
I though that f would be def'ed to something like function f () return 2 end ?
@PauloCereda Thanks :)
@cjorssen I don't know if my assumption is correct.
I guess there's no attribution, so the value is lost.
if you go with f = loadstring('return ({[0]=1,[1]=2})[1]'), it works.
and print(f()).
Yes, it works but I don't understand why the other won't.
This one works too. f = loadstring('t = ({[0]=1,[1]=2})[1]') print(t)
Maybe because loadstring is executed inside an other chunk?
I think so.
We could ask our awesome Lua expert @PatrickGundlach. :)
I'm still in the Lua basics. /blushes
So am I! So I guess I have to wait for Patrick or another lua expert to sort it out? But I'm quite happy with your return workaround :)
It was trial and error. :)
3:07 PM
Thanks anyway! Does your @username will notify Patrick automatically (sorry, I do not fully understand how the chat work for now)?
It will, don't worry. :)
Cool... I just have to wait for the master to come :)
woo 1k again !
@N3buchadnezzar: congrats!
I gave away 1/4th of my points for an exellent answer =)
3:17 PM
@JosephWright There's something similar to a bug: suppose start.tex is a file and do
\ior_new:N \l_my_s
\ior_open:Nn \l_my_s { start.tex }
No error if siunitx is not loaded
4:01 PM
@cjorssen Are you using the Lua REPL? Then every line is a block. (But I have to leave soon)
Yay, @Patrick is here! :)
4:16 PM
@egreg Out of curiosity, I tried getnonfreefonts-sys garamond and it returned the following:
End-of-central-directory signature not found. Either this file is not
a zipfile, or it constitutes one disk of a multi-part archive. In the
latter case the central directory and zipfile comment will be found on
the last disk(s) of this archive.
unzip: cannot find zipfile directory in one of garamond.zip or
garamond.zip.zip, and cannot find garamond.zip.ZIP, period.
File 'garamond.zip' seems to be broken. Please try again.
4:40 PM
@AlanMunn This should be the correct syntax :(
@egreg The problem is that the garamond folder contains a ugm.zip instead of garamond.zip. I guess this needs to be reported to Robin.
4:56 PM
@AlanMunn I tried and it works with sudo getnonfreefonts-sys garamond
@egreg: I sent an email with the script manual. :) Don't kill me, please.
@egreg I just tried it again and it worked. Very weird. Exactly the same command.
@cjorssen You should call these functions with a,b = loadstring(...) and always check a to be non-nil. If nil, then b has gthe error 4messatge. g
sorry for thettgbb t bvgv typos, my son is sitting on my lap gand dgghhhghhibihnhzhg hitting the keybgggogafffrdd
@PatrickGundlach awww how cute! :)
@egreg Bug in l3file which I spotted yesterday: will be fixed when I update CTAN
5:06 PM
@YiannisLazarides Hi Yiannis, I'm not sure that ArabXeTeX should be the recommended package for Arabic if you're producing regular Arabic documents. It's really intended for transliterated Arabic.
Oh my, epic English fail! I wrote "The script threats ..." instead of "The script treats ...".
5:23 PM
@AlanMunn It can depend on the repository that the multiplexer chooses.
@PauloCereda Found. :)
@JosephWright Just when an answer needed it. :)
@egreg I can do a CTAN snapshot now if this is urgent
@AlanMunn Thanks. I saw your answer as well. I don't really use it, but now and then we need to have a scribble of arabic in a document.
@JosephWright I guess it affects any document that uses \ior_open:Nn or probably \file_name_sanitize:nn. Is the latter the culprit?
If it's only for \ior_open:Nn it's probably not urgent
@YiannisLazarides Yes, I agree for that, but for most questions about Arabic here on the site, I think that the posters are producing Arabic documents rather than bits of Arabic in an otherwise non-Arabic document.
@egreg There was a typo inside some of the functions: I'd transposed parts of a variable name :-(
5:32 PM
@JosephWright So it's surprising it worked without siunitx!
In the secret game I'm working on (a TeX version of Monopoly), siunitx is very expensive to buy. :P
5:55 PM
@egreg: sorry for using parskip. :)
@PauloCereda I remember you have very good PHP experience - I got a task, perhaps you are able to contribute something?
It's getting a package name (in HTML) and returning the documentation path (via texdoc -l packagename possibly, tex distribution is on the server where texdoc can be called)
@StefanKottwitz Sure!
Let me grab my dev notebook. I'll write something in a minute. :)
Could be useful for any TeX site, for an easy button or link code to documentations without searching for a link
Great :-)
@PauloCereda No hurry, my family keeps me a bit busy besides the computer ;-)
@JosephWright You posted your answer to the beamer question just as I was pasting my (equivalent) solution.
@AlanMunn Sorry :-)
6:07 PM
Is there an equivalent to texodc but for .sty files? It would be helpful to type something like, for example, texsty adjustbox and get adjustbox.sty displayed.
I meant texdoc.
@JosephWright I've posted mine anyway, since it's a one liner in the source and doesn't require a new version of beamerbaselocalstructure.
@GonzaloMedina We could write one. :)
@GonzaloMedina less `kpsewhich filename.sty`
@GonzaloMedina you could make a shell function or script with an argument for the name
# texedit - find one or several tex related files
# and open them in the editor
if [ $# -eq 0 ]; then
echo 1>&2 Usage: texedit file1 [file2] ...
exit 1
gedit `kpsewhich $@`
exit 0
@StefanKottwitz Thank you! Something like this should be part of the distribution (just like texdoc), don't you think?
@GonzaloMedina would be good. I use such scripts, such es texgrep, texls, texcd for such purposes
Perhaps I could read over it, document it and send it to the TeX Live team
6:12 PM
@StefanKottwitz Thanks again. Helpful scripts!
@StefanKottwitz Sounds great!
@JosephWright In fact, you went too far in your fix. You can't remove the \blx@unitpunct\blx@postpunct\newblock from the \apptocmd, since otherwise the punctuation of the title gets the wrong formatting. (Although I suspect that having this code remain will also create a problem if the title is empty. There's something not quite right with this code altogether, I think.
@AlanMunn Sigh. I will take another look at this, then
@JosephWright Unfortunately my knowledge of biblatex internals is mainly theoretical; I know that you generally shouldn't be letting low level biblatex commands to print punctuation, but at the same time I'm not quite clear on the relationship between \newunit and the actual printing of the punctuation (which is asynchronous). This is why the formatting gets applied incorrectly, but I don't know how to fix it.
6:50 PM
@AlanMunn Life is rather more complex than I'd like as formatting in biblatex should really be sorted using separate style files. However, that would require one beamer-specific style file for each biblatex style: not realistic.
@JosephWright Right. And both packages are quite complex, and somewhat labrynthine in their code base.
@AlanMunn That's not really the issue here: it's more a conceptual issue, as the beamer changes to standard bibliographies are already a hack.
@JosephWright Yes, I understood, and agree with you.
@GonzaloMedina texdoc adjustbox.sty works as well ;-)
@Stefan: I've been playing a little with a script. It's possible, though I'm not sure if it's the best approach. IMHO it would be better if we create a DB with all package infos, then retrieving in PHP; a crontab script could search the TEXMF tree and update the database accordingly.
I'm not a fan of system calls. :(
7:02 PM
@PauloCereda same for me
but sounds complicated
@StefanKottwitz I'll take a look. :) I did something like that before. :)
@PauloCereda I've got both PostgreSQL (separate server) and MySQL (same server)
I guess PostgreSQL on the dedicated server would be better for using on different sites
@MarcoDaniel Unfortunately texdoc latex.ltx doesn't
@StefanKottwitz Indeed.
@egreg pity ;-) Maybe we can suggest this behaviour for the next TeX Live version.
7:13 PM
@StefanKottwitz: I'll find a way. :)
@egreg texdoc source2e ;-)
@StefanKottwitz: oops I have an idea already. :P
7:29 PM
@PauloCereda Great, what is it?
@StefanKottwitz 1. I'll try to write a terminal app to be put on crontab. This app will navigate in the TEXMF tree, find every bit of source files/scripts/documentation, then populate a MySQL/PostGreSQL db with the results. 2. We will have a PHP page that allows the user to search for everything his heart desires (e.g, Gonzalo's suggestion on finding the .sty file). :)
@PauloCereda isn't there a texdoc database which we could convert ...
@StefanKottwitz I'm looking for it too. :P
I just like being compatible to texdoc, even if it's extended
@MarcoDaniel that's true; I only use texdoc for documentation and never tried it on files with .sty extension. Thanks.
7:38 PM
(because experienced users know the texdoc keywords)
@StefanKottwitz I guess I found it! texmf-dist/ls-R
ls-R is clear regarding kpathsea, I did not know that texdoc would rely only on this
texdoc texdoc did not show me that yet
@StefanKottwitz Hm I'm not sure it's the one texdoc uses. Apparently, the Lua script iterates in the doc directory after all.
@StefanKottwitz When you converted this answer into a comment, my and Peter's comments then got added to the original question. This seems like a bug, don't you think?
@PauloCereda type: texdoc texdoc
7:44 PM
@StefanKottwitz Great! Thanks!
@AlanMunn I wasn't sure if the non-answer referred to any of the answer, so I converted it to an answer to the question, and I transferred your comments too so the poster would get your information, I even added @name to your comments to be sure
@PauloCereda texdoc uses aliases, which is important
/usr/local/texlive/2011/texmf-dist/texdoc/texdoc.cnf onmy system
@StefanKottwitz So you transferred the comments yourself, it wasn't automatic? If so, then I don't think transferring the comments makes too much sense, since they now appear as if they are directed to the original questioner.
@StefanKottwitz (I see though what you mean by the @name.) That makes it a bit clearer.
@AlanMunn Yes, the OP sees that they belong to the comment
@StefanKottwitz Ok. That makes sense.
@AlanMunn semi-automatic: I chose to convert the comments as well, then I added @name to clarify
7:52 PM
At ALL ;-) What do you think about the style of this question: tex.stackexchange.com/questions/47537/…?
8:02 PM
@StefanKottwitz: Plan B: I'll take a look on running texdoc in a restricted sandbox, even if I have to write the sandbox. :)
@PauloCereda B sounds good too
What do you usually do about underfull vboxes?
I don't want the text, the content, to be a slave of the layout constraints. I do not want to change headings and structuring.
@AlanMunn: I guess biber has a potential issue with certain cached directories: tex.stackexchange.com/questions/47553/… :)
But I have a chapter title, then three paragraphs, then a section title. The three paragraphs are put on the first page, and the section on the next. The paragraphs won't fill the page, so the space between them is stretched.
8:21 PM
@PauloCereda Yes; it's mentioned also in the Biber documentation
@egreg Really? Alan and I found it was a bug. :(
I deleted my comment, since your answer is complete. :)
@PauloCereda It happens if you kill Biber while it's unpacking the necessarily files on a first run, and then needs to be cleaned up properly to get things working
@JosephWright Ah that makes sense.
@JosephWright Yes, but also at random times. I've never killed it, but it happens anyway.
@egreg I've never seen that
8:25 PM
Ah and people say computer science is an exact science. :)
@JosephWright The doc says it happens for new versions of Biber, but it's not quite so. I don't know if the cache has some "expiration time", but it happens fairly regularly.
Wow. temperantia just got a kernel panic. :P
@egreg This is really a problem then, because no one should have to navigate that deeply into /var/... to delete anything, especially those who aren't generally used to the command line.
@AlanMunn Tell it to the developers of Biber, who decided, in their wisdom, to code it in Perl using modules that they need to compile because they are not generally available.
8:41 PM
@egreg We should write a port in Lua. :)
@PauloCereda At least they should provide an escape that doesn't require the user's intervention.
@egreg Amen to the (implicit) 'don't use latexmk' advice. On the other hand, I've seen too many presentations/papers with (?) in them to think that for many people, not using it routinely is dangerous.
@AlanMunn Are you saying I'm too optimistic about human nature? :)
@egreg Indeed. Too much @PauloCereda, I would think. He has that effect on all of us.
@AlanMunn Aww thanks! <3
Pollyanna attitude! (Don't tell me it's YAGM.) :P
8:56 PM
@PauloCereda No, that's pop culture from the early 20th c. so I'm pretty much aware of it. :-)
@AlanMunn Phew! :)
@egreg: I found a nice use for parskip for my songs sheets. :)
@PauloCereda I'm contrary to spaced paragraphs in "normal text": they create horrible white bands on the page. OTOH, separating stanzas with space has been used for centuries. :)
To each job its tools.
@egreg I learned my lesson. :)
@egreg Indeed. :)
@PauloCereda For example, never use Perl for a big project such a Biber. :)
Nobody would do it. :)
@egreg fancychap excepted.
9:11 PM
@egreg haha so true. :)
@egreg This can be easily generalized to "Never use Perl."
The only contact I had with Perl was when Norbert wrote the region code support in the tl-install. I was scared. :)
@AlanMunn Is it a tool?
@egreg It's a tool for creating gaudy documents.
@AlanMunn I'd say "kitsch".
There's a palace at Sintra, near Lisbon, which is the monument to kitsch: Palácio da Pena
9:27 PM
@egreg Oh my! I'm still trying to understand it. :)
Pena in PT-BR can be pitty. :P
@PauloCereda It's really incredible. You turn a corner and find a completely different style.
@egreg I saw the article in Wikipedia. There are lots of halls with different stuff going on. :)
@PauloCereda And the exterior parts are a paste up of as many styles they could conceive.
@egreg I'd love to visit Portugal. :) (Specially Fátima)
@PauloCereda I've been there for too short a time, but it's a very nice place. Not in Fátima, though. I believe that Porto and Coimbra are places one has to visit; and Lisbon, of course, but I've been there.
9:41 PM
@Jake: Where are you? We need your help ;-)
@egreg Cool! I heard Porto has a very nice wine. And Coimbra has a nice gastronomy, specially pastéis de Santa Clara. :)
@PauloCereda Porto has also a nice bridge built by Eiffel
9:59 PM
@egreg Cool! :)
Mass time. :) I'll be back in an hour or so. :)
@MarcoDaniel: I'm currently travelling, so I only have very sporadic internet access...
@MarcoDaniel it feels nice to be needed, though =)
10:15 PM
@Jake I modified your solution but I am can't improve tex.stackexchange.com/questions/47532/…
@AlanMunn: You can undelete your question tex.stackexchange.com/questions/47458/…
@MarcoDaniel Why? Yours (with the modification I suggested in the comment) does everything doesn't it?
@AlanMunn He wrote a new comment.
@MarcoDaniel I wonder if he has the version of your code before you added the listparindent parameter. Yours definitely works to do what I think he wants.
@AlanMunn I don't know I am working at an other code ;-)
@MarcoDaniel I've edited your answer with another minor fix (added parsep=0pt) in case that's the problem, and notified @jamaicaworm. Maybe that will solve the problem. If not, he's just not being clear enough for either of us.
10:43 PM
@MarcoDaniel He finally seems to be happy with my modification.
@AlanMunn Thanks.
I can cry ****
@MarcoDaniel ???
So that bashful package....What do people think of this?
@AlanMunn I want to improve my solution related to pgfplots -- Riemann Sum. With the simple example function it works. But I want to care of other functions.
%  \pgfmathparse{3*exp(-#1)*#1^3}%
%\pgfset{declare function={f(\x)=3*exp(-(\x))*(\x)^3+1;}}
\pgfset{declare function={f(\x)=sin(2*deg(\x))*exp(0.1*\x)+2;}}
    integral segments/.code={\pgfmathsetmacro\integralsegments{#1}},
    integral segments=3,
    integral samples/.code={\pgfmathsetmacro\integralsamples{#1}},
    integral samples = 10,
    integral min/.style args={#1:#2}{
10:58 PM
@Audrey Do you have a solution to the beamer+biblatex problem?
@AlanMunn There was an update of beamer few hours ago ;-)
I think so. Let me take a look at the update first.
@Audrey The update just did what Joseph has in his answer, I think.
The problem to me seems to be how to tell biblatex to flush the current punctuation out before the change in colour is done.
@AlanMunn Looks like Joseph used your answer by omitting \blx@unitpunct\blx@postpunct in the patches to the title bibmacro.
@AlanMunn Yeah - the patches to the macros have to set some punctuation before changing the color, but only if the last field is defined.
11:17 PM
@Audrey It seems that the problem is being solved at too low a level of command by beamer.
As Joseph mentioned earlier when we talked about this, there's really a need for a whole set of beamer versions of the biblatex styles.
@AlanMunn Whoa. That is a tall order. I think you just need to test whether or not the fields are defined - the unit punctuation features don't do all the work for you.
@Audrey Yes, it absolutely is a tall order, but otherwise everything will turn out to be a bit of a hack, I think.
I.e. you don't really want to patch the low level code for each possible field with some beamer code.
@AlanMunn Not sure what the right solution is in the end. For now I guess I'll just post another hack.
11:33 PM
@Audrey Good.
At every tikz user ;-) i can find the error. Every one who wants to test here the code:
        y filter/.code={%
                            \foreach \x in {0,1,...,\integralsamples}%
This is the relevant part and here the complete one

%\pgfset{declare function={f(\x)=3*exp(-(\x))*(\x)^3+1;}}
\pgfset{declare function={f(\x)=sin(2*deg(\x))*exp(0.1*\x)+2;}}
    integral segments/.code={\pgfmathsetmacro\integralsegments{#1}},
    integral segments=3,
    integral samples/.code={\edef\integralsamples{#1}},
    integral samples = 5,
    integral min/.style args={#1:#2}{
        ybar interval,
@Audrey I would recommend patching the commands as I did in my solution, rather than requiring a new version of beamerlocalstructure.sty (Too easy to forget it's there.)
11:55 PM
@AlanMunn It's relatively messy, but I'll make an edit.
Am I not able to read? See the comment to my answer here
Q: Combining \ifxetex and \ifluatex with the logical OR operation

RomildoI want to write in the preamble of a LaTeX document code that should be take into account by different compilers when the document is typeset. There are fragments that should be used: only by xelatex only by lualatex only by xelatex or lualatex only by pdflatex I am using the packages ifxetex...

I've later added the code example; but what did the commenter think?

« first day (496 days earlier)      last day (3039 days later) »