« first day (466 days earlier)      last day (4460 days later) » 

12:08 AM
seems that it'll take a while...
I'll check it in the morning. hope it'll work. thanks anyway.
 
 
9 hours later…
adn
9:09 AM
Hi all, I have a strange error using \ifx.. could someone give me some advice? Is this a correct place to ask this type of thing?
 
@adn: I would suggest asking a question on the site, that's kind of the point of it.
 
adn
yes, but.. i don't want to bother with something so specific
because I think is some error on my part
 
@adn What's the error?
 
adn
I'm testing something in this way \node (a) \ifx\a\pgfutil@empty\else at(5,5)\fi {something};
so, if the \a macro is empty I want to not print the at part
but I get the error Package tikz Error: A node must have a (possibly empty) label text \node (a) \ifx
Actually, that macro holds the position for that at, so if it is empty I would like to not print the at part.. but I don't know why it is dying at that part.
 
@adn I suspect that TikZ doesn't do expansion: it looks whether it finds at; in your case it doesn't, but finds the \ifx token which is illegal. One should know some more context, in order to give thorough advice.
 
adn
9:17 AM
Ok, then I will post a question... I though it was an dumb error on my part...
Thanks :)
 
you probably need \pgfextra or something
you can't just put any TeX code in a TikZ path
 
adn
@wh1t3 Thanks, I made a proper question now.
 
9:50 AM
@egreg ah I pooped here to ask if I should delete my comment or let it stand, and I see you'd pre-answered the question already:-)
popped
 
10:04 AM
@AndrewStacey: Looks like you are right with regards to the label question. The manual says something about this on page 195 (point 2), but it isn't specific as to how to solve it...
 
10:36 AM
@wh1t3 That comment gave me hope, but digging in the code I couldn't see how to reset the anchor - the obvious things didn't work and there didn't seem to be a hook to override it. I have a hack that I'm about to post ...
 
10:47 AM
@AndrewStacey: I was pretty surprised that just passing anchor=north as an option to the pin didn't work. Nor did the every pin style. Apparently the "automatic selection" code is ran afterwards. But that makes me doubt the note in the manual, all a bit strange.
 
11:02 AM
@wh1t3 I expected that to work as well.
@wh1t3 Looking at the code, I think that one could modify the auto-anchor routine to only run if an anchor hasn't already been specified. Then anchor=north and so forth would work properly. It's only used for anchoring labels/pins and for anchoring nodes on lines.
 
@AndrewStacey: Hmmm, I just find it strange that that is not the default behaviour, in my opinion that comment in the manual sort of suggests it should be.
just saw your answer, excellent :)
 
@DavidCarlisle Keep the comment. It's still relevant. (I presume you're talking about the reference to the \makeatletter question on adn's question.)
 
11:30 AM
@AndrewStacey, thanks
 
@DavidCarlisle Yes, it should be kept.
 
 
1 hour later…
12:54 PM
Sorry for the editing spree; I guess I added all the repos I could find: meta.tex.stackexchange.com/a/1182/3094
 
1:38 PM
0
Q: How to collaborate on a TeX file using VCS with a collaborator who is not using VCS?

PhilipPretty self-explanatory; I'm thinking of @AndrewStacey's answer, in particular the part about working with collaborators who do not use VCS. I'm hoping for a more detailed explanation / step-by-step workflow for how, e.g., to use GIT to make it "seem" to me that a potential collaborator who is no...

I'm not sure. "Too localized", "Not a real question" related to TeX?
A possible answer would be, "Use a diff tool." :)
 
1:53 PM
Just crashed firefox a couple of times and in revenge, it's decided to use nynorsk as my speling language.
@PauloCereda You missed the TeX-SX launchpad stuff.
 
@AndrewStacey Terribly sorry! :( I'll add it.
@AndrewStacey: I updated braids and tqft. :)
 
@PauloCereda Thanks! Incidentally, I was exceedingly chuffed to see both of them mentioned in the TUG "Treasure Chest" of new packages on CTAN.
 
@AndrewStacey Really? That's awesome, congrats! :)
 
2:28 PM
@AlanMunn: can I add a feature request to the he-she package? :) In one of the Futurama movies (The Beast with a Billion Backs), there's a funny line where a new gender neutral is added: (Fry is presenting the monster who abducted everybody in the world) Yivo is the lover of all beings, male and female. But Yivo has no gender, thus Yivo has proclaimed that instead of "he" or "she," we are to use the word "shklee." And instead of "him" or "her," we are to use the word "shklim," or "shkler." :)
 
 
2 hours later…
3:58 PM
29
Q: Why are shortcuts like x += y considered good practice?

EpiGradI have no idea what these are actually called, but I see them all the time. The Python implementation is something like: x += 5 as a shorthand notation for x = x + 5. But why is this considered good practice? I've run across it in nearly every book or programming tutorial I've read for Python, ...

One of the comments: "Because everytime you write x = x + 5 a mathematician kills a kitten"
 
@PauloCereda No, we do voodoo with a fetish representing the scoundrel. :)
 
@PauloCereda Now tell me why x = x + 5 isn't a valid equation…
 
good point, I'm almost out of kittens.
@AndreyVihrov: In that context it's not an equation, it's an assignment
 
4:13 PM
@AndreyVihrov Programmers won't understand the point. :) However, my Algebra professor in first year said: "Equations don't exist"
 
4:24 PM
I guess x=(6.25)^(1/2) would be a solution...
 
@wh1t3 Uh.
Who knows about this curious number? \sqrt[3]{\sqrt{5}+2}-\sqrt[3]{\sqrt{5}-2}?
I should add that it's quite famous. :)
 
Well, it's aparently 1, but I have never seen it before
 
4:40 PM
@wh1t3 Just apply Cardano's formula to the equation x^3+3x-4=0. :)
 
4:52 PM
@egreg The author of tcolorbox answered. He will provide an English documentation. He couldn't terminate the publishing. He thinks not in the next week.
 
@MarcoDaniel Nice to know
 
@PauloCereda No. :-)
 
Hey everyone
Sorry for clogging up the exchange with basic questions today
@egreg it's that phi?
 
5:09 PM
\sqrt[3]{\sqrt{5}+2}=\phi^3, yes. :)
 
So looking at the Git vs. BitBucket choices in package maintainers, does that separate the amateurs from the pros?
 
Can somebody explain the optional argument type t of xparse. I understand it in the following way.
\documentclass{article}
\usepackage[T1]{fontenc}
\usepackage{xparse,expl3}
\NewDocumentCommand { \MyTest } { t{a} }
 { \IfBooleanTF{#1}
    { \texttt{#1} }
    { not given }
 }
\begin{document}
 \MyTest[a]
\end{document}
 
Hi guys. quick question: what is the common approach if you have a question (and the corresponding answer) that may be interesting to others? write up a question and immediately answer it? reason I'm asking, that I just answered a "non-bug" for LaTeX and thought that should have been asked here instead
2
And also if I may add, perhaps all the question from the ukTUG FAQ might be sensibly added as questions+answers here
 
@FrankMittelbach Everything is welcome. If you want I can answer the question.
 
@MarcoDaniel I have both already on my screen, so no need for others to answer
unless you have a better answer than mine of course ;-)
 
5:24 PM
Normally if you can answer your own question it is kind to provide them.
 
@FrankMittelbach I suspect that we already have questions for many of the UK FAQ questions on the site, so adding them wouldn't necessarily be a good use of anyone's time.
3
 
@FrankMittelbach We have also a badge for this ;-) tex.stackexchange.com/badges/14/self-learner
 
@FrankMittelbach Answering your own question within seconds is fine as far as I'm concerned. For example, look at the dates between the question/answer of @lockstep's entry:
18
Q: Continuous v. per-chapter/section numbering of figures, tables, and other document elements

lockstepSome document elements (e.g., figures in the book class) are numbered per chapter (figure 1.1, 1.2, 2.1, ...). How can I achieve continuous numbering (figure 1, 2, 3, ...)? And vice versa: Some document elements (e.g., figures in the article class) are numbered continuously. How can I achieve pe...

 
@AlanMunn 2 comments: a) i doubt you have all and b) and one of the good things about that source is that the answers are being worked on for a long period of time and the quality is very high
@MarcoDaniel getting a badge for that would be a bit rediculous in my case (as i wouldn't have had the question in the first place :-) but ok.
which reminds me of something else (probably a question for meta if it doesnt already exist. is there a way to acutally mark questions/answers as something like FAQ? it would be kind of nice to have a structure like that which you can actually browse
 
@FrankMittelbach Oh, I'm sure we don't have all, and I agree that the quality is excellent. I just think that most of the truly frequent ones have probably already arisen, and for the others, the law of diminishing returns make apply.
 
5:33 PM
@FrankMittelbach FAQs are typically searched by tag. For example, here's the FAQ entries relating to : tex.stackexchange.com/questions/tagged/tikz-pgf?sort=faq
You click on the tag, then click on the FAQ sort order.
 
@FrankMittelbach The site kind of generates its own FAQs dynamically based on views/votes of questions. So if you click on the Questions link, and then sort them using the FAQ link, you get an approximation of a FAQ. And as Werner just added, you can do this by tag, too.
 
@AlanMunn Agreed. One should justify the need to list something as a FAQ. It made sense to add the "This led to the package" tag...
 
@AlanMunn ok so the votes on questions (and or on answers?) make up the FAQ sorting. But I'm not sure this really is the way people use the voting buttons on questions. in my opinion there is still a concept missing here
 
...while adding something like "This is a FAQ" might be subjective.
@FrankMittelbach Since things seem questionable, meta is the way to go...
 
@Werner maybe. not saying this is thought out. Only if you go to questions and scan the faq order then many of the questions with high votes do not necessarily cover what I would expect (some do). And that tells me that something is not as it perhaps should
 
5:41 PM
@FrankMittelbach Yes, and this may be due to the FAQ sorting algorithm. There was a mention of this recently on meta:
15
Q: Does the faq tab work correctly?

Stefan KottwitzI like the idea of automatically generated FAQ lists. The mouseover text of the faq tab is questions with the most links. That's sounds sensible, links may come for example from duplicates or closely related questions. So I had a look at the {floats} faq.: Number 5 of the list: Article append...

 
@FrankMittelbach I also agree. That's why I said an approximation. But it's not bad. So for example, the top questions overall include only one that is inappropriately a FAQ. All the others are genuine FAQs (and probably on the UK FAQ too.) One thing that we've done as a community is to create canonical community wiki answers for certain questions.
 
but then perhaps it just doesn't quite fit the model of the site as a faq essentially needs some level of moderation typically
 
@FrankMittelbach That sounds right. As I mentioned, it may be a tad subjective.
 
We have 72 answers that explicitly mention the UK FAQ. (Searching for +UK +FAQ).
 
6:07 PM
@FrankMittelbach: Also note that voting for some people is very specific. The "formal" description for up-voting is "This question shows research effort; it is useful and clear". So, up-voting when considering "useful" may be more FAQ-ish, while up-voting due to "research effort" or "clarity" has little impact on the FAQ-ness. I'm not sure to what extent the votes form part of the FAQ sorting algorithm.
 
tex.stackexchange.com/questions/13370/… comes up as the top question on FAQ for tex-core ...
so much for "useful" "research effort" or "clarity" ;-) a closed question
 
Views seem to play a big part in this...
@FrankMittelbach Seems like your voting average is just under 1/day. So I guess you don't consider much as being "researched", "clear" or "useful". :)
 
6:42 PM
@JosephWright: The documenation of xparse should be changed. The optional argument t requires a token with catcode 11. I think this isn't clear in the documentation.
 
@MarcoDaniel No it doesn't
 
@JosephWright Curious I tested my example again and it works like expected.
 \documentclass{article}
 \usepackage[T1]{fontenc}
 \usepackage{xparse,expl3}
 \NewDocumentCommand { \MyTest } { t{a} }
  { \IfBooleanTF #1
     { given }
     { not given }
  }
 \begin{document}
  \MyTest a
 \end{document}
 
@FrankMittelbach I think that the tex-core tag is probably one of the worst ones you could have chosen: the site caters much more to a lower level of expertise in general. Remember there are about 9500 users, but the tail of low rep users is 8900. (see stackexchange.com/leagues/29/week/tex). Although reputation doesn't correlate exactly with expertise, (since it's also dependent on time on the site, and general participation) it does correlate somewhat.
So we have relatively fewer experts than non-experts on the site. But the tex-core tag is somewhat more biased towards experts than non-experts, I would think.
 
@AlanMunn Great comment!
 
@MarcoDaniel Thanks for the ping. We will wait for the English documentation then.
 
7:35 PM
@egreg D'oh, I should have seen the similarity of this formula to Cardano's one (for x³+px+q=0) outright! :-/
 
8:15 PM
@egreg: I am trying to understand your solution. But I can't In the TeXBook the explanation of lccode doesn't allow any conclusion for me. I think (thought) I understand the meaning of lccode but that's all
\documentclass{article}
\begin{document}
\def\achar#1{\begingroup\lccode`!=#1\lowercase{\endgroup!}}

\achar{97}
\achar{`a}
\achar{"61}

\begingroup
\lccode`W=`H
\lccode`O=`e
\lccode`R=`l
\lccode`L=`l
\lccode`D=`o
\lowercase{WORLD}~WORLD
\endgroup
\end{document}
 
I'm back, yay! :)
 
8:32 PM
Can we please close this?
0
Q: Abbreviate "figure"

JanI'm using scrbook for the document class. Is there an easy way to abbreviate "Figure 1" in order to have "Fig. 1" ?

 
@lockstep Done. :)
 
@MarcoDaniel As I said, with this version of \achar, \achar{97} doesn't do more than \char97. The interesting bit is the other version.
 
@AlanMunn understood. however that doesn't explain that a closed one makes it to the top of the list
 
My dad wrote a whole text in bold font. /facepalm
 
@FrankMittelbach Because that particular question has 17k views, this is almost 10 times the next highest question in terms of views, and far more than is normal for the site. What that means is that the question was likely linked to via reddit or similar aggregator sites which generate huge views, and simply skew the automatic FAQ system. The reason it's closed is because it's not really an appropriate question for the site in the first place. But it's truly an outlier. (And mis-tagged IMO.)
 
8:52 PM
@MarcoDaniel I've added what you wanted
 
@AndreyVihrov Did you tire of your previous avatar?
 
@Werner true my voting average is low. I'm only start getting a feeling for this site and didn't felt the urge to throw votes around without knowing what it means and I'm not really looking that much around on it (or at least I try to)
 
@FrankMittelbach Yes, I agree. Time spent on TeX.SX definitely builds character type. My voting for answers is usually based (as a minimum) on the literal sense: Does the "answer" actually "answer" the question? In terms of questions it my be more subjective, since it requires expertise in the field of interest.
That's why it also helps if users submit images of the output. That way I don't have to wonder whether it actually does when I don't know much about the subject.
Jake's recent interview mentions that pictures help...
...and I would tend to agree.
 
@AlanMunn Kind of. But it also would show up everywhere I registered with my e-mail, through Gravatar, so at some point this became a little inconvenient.
 
9:10 PM
@AndreyVihrov I can see that. That's why I've stuck with the little abstract pattern too.
 
9:38 PM
@JosephWright Did you see the question about \cs_if_exist:NTF and \cs_if_free:NTF? My feeling is that the former would be the same as \cs_if_free:NFT(if it existed).
 
@egreg I think that is the case nowadays: I have a feeling it used to be different. I'll take a look back through the SVN (quite a way, perhaps to when I joined the team)
 
@JosephWright Maybe something related to being equivalent to \relax?
\cs_if_free:NT \scan_stop: { \cs_set:Npn \scan_stop: { AH } } :-)
 
@egreg Yes, there was something
 
@JosephWright Better yet: \cs_if_free:NT \scan_stop: { \cs_gset:Npn \scan_stop: { AH } }
Would you like to try and bet about the first crash?
 
10:06 PM
@egreg Sorry had to go for few minutes. Now I am back
My question related to the lowercase-part which includes the endgroup. I can't understand the algortih. Normally I would say \expandafter\endgroup
 
@MarcoDaniel There's nothing to be expanded after \endgroup.
The key point is that nothing inside \lowercase is expanded or executed; just the conversion is performed and the result is read again.
 
@egreg That was an example I know. I can't understand why \lowercase{endgroup!} works. Of course you changed the lowercase part of ! but why it works after endgroup
 
@MarcoDaniel The group ends when \endgroup is executed, which is after the case change
 
@MarcoDaniel Because what TeX sees after the conversion is \endgroup\def\firstA{A}; as Joseph says (and I said before), nothing has been executed or expanded yet.
It's a very nifty trick. Plain TeX uses a similar one, but without the \begingroup and \endgroup trickery, in the definition of \if@ (line 271), which should be followed by if both with category code 12. It might be
\begingroup\uccode1=i \uccode2=f \uppercase{\endgroup\def\if@12}{}
 
@JosephWright @egreg I believe I understand. Please let me try an rowdy explanation ;-). First you begin a group. Inside you set the lower part ! to my input. Now you call \lowercase with the argument \endgroup!. lowercase doesn't expand or execute anything. It takes every token and provides the lowerpart of them. During this process everything is saved in a "savebox". After the switching the box will be called. \endgroup+lower part of input but without a global changing of !
 
10:17 PM
@MarcoDaniel There's no box
 
@egreg It for my mind
Or works `lowercase` like a sequence
1. token1 -->lowerpart token1 --> print token1
2. token2 -->lowerpart token2 --> print token2
 
@MarcoDaniel It's more like a "temporary \def", I believe.
 
@egreg the "temporary def" is like "savebox" ;-) -- In this case I define a command \foo which includes \temporarydef\foo{\endgroup+lower part of!}
 
The gotta be kidding: ubuntuvibes.com/2012/02/…
 
10:44 PM
"Hi, I'm trying to install TeX Live, but no luck so far. I'm using [ RebeccaBlackOS | Hannah Montana Linux | Justin Bieber Linux ]"
 
@PauloCereda Did you try "Chuck Norris Linux"? Probably the installation of TeX Live is performed before you think to download the distribution.
 
@egreg LOL! :)
 
Now off for some minutes: the radio is broadcasting Mozart's "Exsultate! Jubilate!"
 
@egreg Cool! :)
 
@egreg: Thanks for all your explanation. I think I understood. Bye
 
10:58 PM
@PauloCereda Mozart wrote it when he was 17 for a famous castrato of his time. That's why it's very difficult for a female soprano: it has many passages with grave notes. I once heard it sang by a boy of the Wiener Sängerknaben: very good, I'd say.
A movie inspired by this castrato was made some years ago; they reconstructed his singing by pasting together female and male voices.
I think that the answer to this one is "you can draw lines in any color you want, so long as it's black" (from H. Ford). :)
0
Q: Changing color of \dir in xypic

Chris GranadeI'm trying to use \dir to draw custom frames in an XYpic drawing, but have been unable to figure out how to change the color of the line drawn by each command. For instance, to draw a triangle frame, I'm using the code \POS ="i","i"+UL;"i"+DL **\dir{-};"i"+R **\dir{-};"i"+UL **\dir{-},"i" inside ...

 
@egreg Ah interesting! Mozart was fantastic, specially due to his ability to alternate grave and high notes. :)
@egreg haha :P
 
11:18 PM
@PauloCereda Actually it's possible to color lines!
 
@egreg Wow! :)
 

« first day (466 days earlier)      last day (4460 days later) »