« first day (1200 days earlier)      last day (2614 days later) » 

@PeterGrill Good job, but I still prefer mine. ;-)
@egreg :-)
@PeterGrill steal a few more and you may beat Stefan in joining egreg's 100k club
@PeterGrill The question is obscure. There should be something else by the same OP, but I didn't try finding it: what I remember is that it was a bad idea to begin with.
12:39 AM
@DavidCarlisle :-) Yeah, kind of miss the old days when I almost kept up with you for a few days.
@egreg Question seems ok to me, not sure why the OP requite an xstring solution though.
12:53 AM
Hi there, I asked this question (tex.stackexchange.com/q/159896/6870) concerning the use of hobby paths within the pgf Basic Layer today. Now that I thought about it I would also like to know whether it is possible to use hobby paths in the definition of a new decoration. But at the heart of the latter question lies the same subject as in my current question (to my knowledge I can only use Basic Layer commands to declare a new decoration).
Should I ask it nontheless or maybe just edit my current question (and maybe question title) in order to reflect this important use case?
1:28 AM
Yes, I am here. What is the question?
8 hours later…
9:11 AM
I thought I would bounce a question by you guys. I know big community wiki lists of similar answers is a bit complicated. However, I found myself looking at my ToC the other day and thinking: "is there nothing nicer around?". I have looked a bit and found a few question implementing specific wishes for specific ToC layouts but I am yet to find a big list of nice ToCs.
Would you say it is okey to ask for such a thing? Maybe there already exists something that I hav missed?
I mean I am sure people must have done nice solutions and it would be nice to be able to browse through such
2 hours later…
11:27 AM
any one can give me a hint on this problem? I simply have a latex document, where I have lots of images in it. In number of places, some images get cut of at the bottom of the page instead of Latex starting a new page to locate the full image.
For example, I have this

but image2.png in the above example, do not show completely in the pdf. It gets cut off, since it is displayed at the bottom of the page below the first image, and there is not enough space to display it fully. I though Latex will take care of all of this?
11:40 AM
@Nasser Maybe the bounding box information is wrong. Try adding \fbox around the \includegraphics to see if it's the case.
@egreg These are .png images not eps? Actually, I have the actual image sizes in there (Scientific word figures the image size and adds that). I did not show that above since I do not think it is relevent. The actual code is:
followed below that is another include, similar to the above for the second image. But many times, Latex cuts off the second image.
@Nasser My guess is that the information is computed wrongly.
@Nasser There should be no need to add natheight and natwidth information.
@egreg hummm.... This will be a major bug in SW if this was the case.
@Nasser Possibly a bug in the PNG file. Try the \fbox method.
May be I should make a small example then. But will try the \fbox method now....
@egreg I've added \fbox and same problem. I do not know if I did it correct. Here is a screen shot and the code

Here is the output, showing the second image is cut off
The second image above is much larger than that. It should go to next page, but it does not.
Only if I add \clearpage manually it will. But I have so many images, and I can't keep manually managing this.
I will try now to remove the explicit image size information added by SW to see what happens
I get the same exact result using this code:

12:01 PM
@Nasser the fbox won't change the position it just lets you see the bounding box, is the box just going round the top bit of the image (stick a clearpage in first so you see the whole image, if so the bounding box in the file is wrong
@DavidCarlisle but this is a PNG image? do png images have bouding boxes? I did resize these images earlier today using PAINT.NET to make the image sizes smaller. So may be PAINT.NET has a bug?
I will have to make a MWE then....
In theory, Latex should have moved the second image to top of second page, right?
@Nasser so I noticed aftewards, but then natheight=4.808600in, might be wrong hard to tell from here
@Nasser latex will do what you tell it to do,
@DavidCarlisle Does write my thesis for me work?
specifying height and width is also usually wrong (unless you also use keepaspectratio, otherwise you will distort th efigure
@PauloCereda yes
@DavidCarlisle woohoo
12:04 PM
I thought when I include an image, if it does not fit in what is left in the page, Latex will move the image to the next page. Do I have to tell Latex this?


@PauloCereda Of course, your supervisor may not give you a degree for the 12 days of Christmas, but that's a different issue.
@DavidCarlisle Oh. :(
@Nasser You told latex what size the image was, I suspect you lied about that and told latex it was small enough to fit. But can only guess from here.
But I then removed the image sizes? (assuming SW was wrong). I tried :

And it still chopped off the second image
12:08 PM
@Nasser It might, or it might not. If there is no flexibility on the page (unliklely between two center environments) and you have flushbottom in effect then just with line breaking sticking stuff in the margin, you may have set unachievable constraints, in which case it will complain a lot in the log file and let the page overrun
I really do not understand Latex at all
@Nasser You know the drill by now, see if you can reproduce just using \rule instead of \includegraphics, and post a complete document as a question on site.
In other news, I'll tell you guys, pdftk is a really handy tool!
@DavidCarlisle Yes, I think I need to make a MWE. But this will take me some time. but will do that.
:13734870 $ pdftk
bash: pdftk: command not found
@PauloCereda what would it have done had it been there?
12:12 PM
@DavidCarlisle :) It's a tool for handling PDF files, like merging files, rotating pages, extracting content, lots of stuff. :)
I used it a lot during the conference. :)
2 hours later…
2:41 PM
From the cnltx manual:
> \LaTeX{} tools and documenting facilities the CN way

@cgnieder I like your manuals
@Johannes_B Germans. <3
@PauloCereda Nice people
@Johannes_B thanks :)
@PauloCereda :)
@cgnieder Just wondering: Does KOMA provide a simple if-statement, to test if a KOMA class is used?

Something like `\ifKOMA{}{}`
@Johannes_B you can test if \KOMAScriptVersion is defined. IIRC it is only defined by the KOMA-Script classes...
2:54 PM
@cgnieder Thank you
Good idea
3:11 PM
@Johannes_B doesn't work:
\ifcsname KOMAScriptVersion\endcsname
  no \KOMAScript

@Johannes_B This seems to do it:
\ifcsname KOMAClassName\endcsname
  no \KOMAScript
My new hero
Thanks again
3:27 PM
@Johannes_B :) I'm surprised there are even new questions to be asked ;)
@What do you mean?
@Johannes_B so many questions have already been asked in TeX.sx ... funny that there are new ones every day that haven't been asked already
@cgnieder Ah, now i understand.
@cgnieder I wonder if one day will come in the far future, when every question asked on TeX.SX is marked as [duplicate] :-)
@Johannes_B Ah, but then LaTeX3 will be published and new questions will pop up :)
Hopefully @egreg will be still around
3:34 PM
@Johannes_B Possibly. ;-)
level 2:
Q: Drawing unknown functions

AchifaifaI need to draw a graph with three functions in it, and I need to fill certain areas between them. However, while two of the functions are defined in the problem ($y=x^2$ and $y=\frac{1}{2}x^2$), the third one is not (Finding it is actually the problem), so I have no idea about the function I have...

3:59 PM
@Johannes_B most of them are already duplicates of questions asked on c.t.t last century (apart from tikz ones, which can be safely ignored:-)
4:30 PM
@cgnieder Well, many new questions are duplicates, just it's silently ignored
My recent question is basically a duplicate, but the other ones are so incomplete that it is actually not
I have in mind obviously something more general than what people before me had
4:46 PM
@egreg your l3 code looks nice but the spacing is still wrong:-)
@DavidCarlisle By the kerning between "t" and "s"
@egreg by the different italic corrections of t and s oh perhaps not, there is a ) to save you (which means my answer isn't as right as it could have been, oh well)
5:08 PM
@DavidCarlisle Fixed in a different way, that needs measuring only one box.
@tohecz Almost a "do it for me without me telling you what I want" :-)
@DavidCarlisle And I got the tick. :P
@egreg (and a vote from me:-( double disaster
@DavidCarlisle I was already capped, so only one disaster
@egreg it still goes to badges doesn't it? and anyway it's the pain of voting not the effect on your total score that hurts
5:17 PM
@DavidCarlisle :)
@DavidCarlisle In I have 38 upvoted answers, 46 in . So bronze badge for both, but far from silver.
By the way, I'm 13 answers short of a gold badge in
@tohecz I know :)
6:03 PM
@egreg I may actually beat you too something :-)
6:15 PM
@PauloCereda What's this?
@MarcoDaniel Attack on Titan.
7:08 PM
@PeterGrill Yep, quite :)
btw, got a new mouse today, and I tell you, the flywheel is soooo gooood :D
7:36 PM
it feels strange, to be at home and after dinner already at 8.30pm
So, Jeff came back in today, and says he is "Warming up to siunitx", especially after he found it can do tables. @JosephWright
It seems our template is based off something called csthesis
@tohecz: a friend of mine lent me a book on interactive theorem proving -- I'm on page 1 and it's already boring. :)
@PauloCereda We've had a PhD seminar today aboud "Incorporating OCaml into Coq" :)
@tohecz You are kidding, right? :)
@PauloCereda no, I'm not
7:49 PM
Coq can generate OCaml code from your proof. :)
@tohecz Oh my.
"Un compilateur d'OCaml vers Coq" à
> 11h. Je vous communique la salle dès que je l'ai.
> I will present a compiler to translate OCaml programs to Coq.
> OCaml is a functional programming language and Coq a proof language we develop
> at PPS. The aim is to have a tool to prove properties about OCaml programs and
> get a better understanding of links between proofs and programs.
> Coq already includes a purely functional programming language thanks to the
> Curry-Howard correspondence. The main challenge is to handle effects
@tohecz Ah. :)
Damn CH isomorphism!
@PauloCereda but I mean, someone who needs to penetrate himself from behind to be able to get the information that user pressed enter? Sorry
@JosephWright MATLAB questions are off-topic on Mathematica by definition :)
@rm-rf Drat: must have got wrong target :-(
7:54 PM
:13741823 That's for questions about interfacing between the two (with the primary emphasis on the Mathematica part) :) You probably want to send it to Stack Overflow.
@rm-rf Was meant to go to math-SE
@JosephWright these two are confusing
@JosephWright Would still be off-topic there... I don't blame you though. We get at least two off-topic math post on our site per day. To complicate things, mathematics is actually called mathematica in some European languages.
@rm-rf I thought there was a matlab tag on the math site
@rm-rf yep, I can confirm, just with "k" instead of "c" for us :)
8:02 PM
@JosephWright There might be a tag, but that's only half the story... I'm sure Mathematics might be OK with a MATLAB question on plotting a function or minimizing something, but "out of memory" errors are most definitely SO material.
@tohecz :)
@JosephWright: send Matlab questions to @percusse. :)
@tohecz Heh, it's an unfortunate choice of name, but the funny part is that a lot of users are confused about whether to answer or to close it, because those questions can certainly be answered using Mathematica :)
@rm-rf yep. Well, calling mathematical software by the word "mathematics", just in Latin, is a bit ridiculous. Scientific journals at least add "acta", "communicatio", "studia" or so.
Apparently, Steve Jobs suggested the name...
Question: Is there a reason not to load both floatrow and caption then set the caption width for each table to FBwidth using caption, rather then the extra code from floatrow?
8:13 PM
@rm-rf well, it's not the only ridiculous name of a software out there. To name some other: android, TeX, C#, sage, ...
The first thing I learnt about python is that the reference site is python. org, not python.com :D
Doesn't work, nevermind
@rm-rf well, makes sense, after all, it's not com merce :)
@tohecz Have you visited python.com?
Oh.. .it has changed now. No wonder it didn't seem to make much sense the last time I said it to someone else...
@rm-rf well, looks like some scam almost
@rm-rf btw, nice to see that TeX.SX is not the only site that has a xkcd-like drawing thread :)
8:21 PM
@tohecz From what I remember, it used to be a porn site, prominently featuring male "pythons"... I've had several embarrassments when I'd quickly type in .com instead of .org with someone watching :P
@rm-rf oh pomograhy!
@tohecz Our site is where the craziness started :) Most others (R, MATLAB, etc.) just ported some of the solutions from that thread.
@rm-rf well, the TeX's solution is probably different, since we have a very crazy (and powerful) drawing library
@rm-rf Yep, yours is 5 hours older :)
@rm-rf well, after all, I think that those question, in all their foolishness, are what makes a community :) and comments like this one getting 80 upvotes:
I've been holding the mouse over the plot for ages now and I have yet to see the punchline displayed. — David Carlisle Oct 1 '12 at 14:07
@tohecz Some people are incredibly talented... and crazy (with loads of free time)!
You can actually play the simulation of minecraft smoothly if you have a copy of Mathematica installed.
@rm-rf are you funny? no way I ever had mathematica installed
but the guy is crazy
I used to be a maple person when I needed it. Since last year, it's Sage for me
8:33 PM
yeah yeah, use whatever rocks your boat... as long as it gets the job done :)
yep, that's it :) I used to be skeptical about sage at the beginning, but I understood that the project is not bad
It has improved quite a bit over the years. And it's a perfectly good (and free) alternative to most problems students would encounter in their coursework. If they already know python, it's an even smaller reason to switch.
@rm-rf yep. Acutally, the python behind, first I thought it's bad because I didn't like python. And then I discovered yield
@tohecz I found this question and its answers useful (I still haven't grokked it completely)
@rm-rf well, this is the answer that finally made be understand it:
def f123():
    yield 1
    yield 2
    yield 3

for item in f123():
    print item
but still, I hade that you can't slice a generator with the [:n] notation. I hope that there're no real obstructions and that it'll come one day
8:45 PM
@tohecz Yes, that explains what it does and I get that part. I haven't programmed enough in python to encounter a practical use for yield. I've read that it's a memory efficient way of doing things rather than using return, but haven't seen it in action.
@rm-rf well, for me, a generator is a good symbolic dynamical system. In each step, it moves forward and returns the symbol
so I use it quite
Yeah, that's a good way of putting it. You can only move forward and access it once.
@rm-rf yep
but in general python way of using iterators is so great. In almost any other language, the following would be either an unreadable code of a for cycle:
xmin = min( x for i,x in MyTwoDimensionalArray )
@PauloCereda Obligatory link abandonmatlab.wordpress.com
@percusse Ha, I've read that and it's awesome! (and true)
8:52 PM
@rm-rf :) I have the (robust) control toolbox versions of those things.
For the record, Mathematica is not my friend either :P
@percusse That's ok. I have plenty of things to bad-mouth about it too :)
In fact ;
Oct 2 '12 at 23:41, by percusse
@PauloCereda MATLAB is the antagonist, Mathematica is the murderer that you only get to see in the end. Maple is the guy who gets killed in the first 5 minutes. Octave is the girl to be saved.
Heh, people often assume that since I'm a user on Mathematica, I'll spring to its defense at the drop of a hat. I'll probably be the first to join in on any valid criticism :D
@rm-rf just clearing my name hehe.
@rm-rf now that's nice to hear:) people should be reasonably critic about whatever they do. (Ask anyone let's say 20k+ on this site whether they think TeX-core is good as a programming language haha)
8:56 PM
I switched to Python so I am free from negative thoughts now.
@percusse well, python is nice, but only to some extent :)
@percusse <3
@percusse I don't think that highly of Octave either... it's almost exactly like MATLAB but without the only parts that make MATLAB bearable/usable (toolboxes).
@rm-rf octave is the free maltab, right?
8:58 PM
@tohecz It's the first prototype of MATLAB before it is sold to Mathworks that was accidentally open-source licensed
The other licensed version took off as MATLAB.
If only I can remember the person's name
I have friends working for Mathworks so I heard the stories gazillions of times over beers
Octave is definitely not my cup of tea.
@percusse You were being serious? If so, I don't think that's true. MATLAB was developed in the late 70s-early 80s, while Octave came much later (90s?)
@PauloCereda Well, Matlab has at least one feature that is missing in python, and it is nargout
9:00 PM
@rm-rf Yes that's about right iirc.
@percusse: where's that Chaplin-like video about Python and Matlab? :)
@tohecz That's not even a challange for Python
@rm-rf However, the open-source version is picked up much later.
@tohecz What if you emulate that via metaprogramming? :)
Damn, there goes my TCS hat again.
@percusse no, but the code needs two libraries and it's a mess of very strange code, at least for me
9:02 PM
@percusse YES!
it's metropolis
shame on you. Go iterate over dictionaries with unknown structure :P
In [1]: def foo(a, b, c):
   ...:     return 'quack';

In [2]: import inspect

In [3]: inspect.getargspec(foo)
Out[3]: ArgSpec(args=['a', 'b', 'c'], varargs=None, keywords=None, defaults=None)
@tohecz: ^^
def expecting():
    """Return how many values the caller is expecting"""
    f = inspect.currentframe()
    f = f.f_back.f_back
    c = f.f_code
    i = f.f_lasti
    bytecode = c.co_code
    instruction = ord(bytecode[i + 3])
    if instruction == dis.opmap['UNPACK_SEQUENCE']:
        howmany = ord(bytecode[i + 4])
        return howmany
    elif instruction == dis.opmap['POP_TOP']:
        return 0
    return 1
@rm-rf Ah! Cleve Moler.
@PauloCereda nargout, not nargin ;)
9:05 PM
@tohecz \usepackage[nargin=2cm]{geometry} Oh wait. :)
So he writes MATLAB. Then Jack something rewrites it in C and commercializes it.
So that initial MATLAB becomes Octave.
@tohecz: now I see what you want. :)
@PauloCereda I want this:
@percusse I still don't think that's the case... I think Octave started on its own, as an alternative to MATLAB.
def f() : return 1,2

a = f() # should contain 1 and not (1,2)
a,b = f() # should be as is, a=1 and b=2
9:07 PM
Moler and Little together formed MathWorks to commercialize MATLAB.
@tohecz Got it. :) Perhaps some trickery might make it work. :)
@PauloCereda well, the code above, if you run it inside a function, returns the number of expected values
@rm-rf Next time I'll listen more carefully. But I'm almost sure that this boring story relates those two things.
@percusse Wack-a-Moler? :)
Nevertheless I think both are terrible so no damage is done.
9:10 PM
Hm guacamole.
@tohecz Got it. :)
@percusse this, quite
@tohecz: I need to be more open-minded, I'll be back to some Python coding. :)
@PauloCereda yeah, well, Python seems to be good. I was actually going to remove the last for cycle from my function, and make it a complete one-liner full of for in if max(in for if else)
9:12 PM
@tohecz Good call. :)
@PauloCereda No star notation?
[*args] ?
@percusse ooh pointers? :) Beats me. :)
@PauloCereda no, I actually don't think so. My python code shouldn't be more messy than my C code :)
Q: What are the main differences between AutoLaTeX and LaTeXMk?

IngoI stumbled upon AutoLaTeX today and it seems to be very similar to LaTeXMk indeed. Can you summarise the main differences, which might help me choose between the two tools?

@percusse Heh, it's ok, the mangled up stories are always better. Cleve Moler wrote MATLAB in the 70s and open sourced it. Stephen Wolfram copied it, improved the syntax and created Mathematica with Jack Little. Meanwhile, a small group of masochistic programmers craved for the rawness of MATLAB's syntax with the usefulness of a blunt machete and created Octave without any toolboxes. =)
9:13 PM
@tohecz LOL
Maybe I should've thrown a Knuth in there somewhere...
@rm-rf Don't touch Knuth ;p He's made a programming language one-of-its-type :)
@rm-rf Meanwhile, Maple was sitting in the dark corner. "Hey guys, can I play?", said Maple. Then a voice echoed, "I wanna play too". "Shut up, Scilab, no one loves you". :)
Skynet is actually an out of control TeX macro.
I can't decide which is more masocistic — programming Skynet in TeX or CSS?
@rm-rf :)
9:20 PM
@rm-rf I think if I find those LA/LINPACK people again, I can get the real story.
@rm-rf what is Skynet?
Skynet is a fictional, self-aware artificial intelligence system which features centrally in the Terminator franchise and serves as the franchise's main antagonist. Rarely depicted visually in any of the Terminator media, Skynet's operations are almost exclusively performed by war-machines, cyborgs (usually a Terminator), and other computer systems, with the goal of exterminating the human race. Origin and nature Skynet was a computer system developed for the U.S. military by the defense firm Cyberdyne Systems. Skynet was first built as a "Global Digital Defense Network" and given comm...
oh ok :D
@tohecz We are safe for now. :)
@PauloCereda btw, should do a bit of work on htlatex
9:24 PM
@tohecz Uh-oh.
@PauloCereda well, we have started a brave project, unfortunately, we don't have a better tool than htlatex for processing latex files
9:39 PM
@PauloCereda Is this involving pointers?
A: Find out into how many values a return value will be unpacked

Martijn PietersThere is little magic about how Python does this. Simply put, if you use more than one target name on the left-hand side, the right-hand expression must return a sequence of matching length. Functions that return more than one value really just return one tuple. That is a standard Python struct...

On other news, check out the last of the features and the comment from knockerccd.
Both are unfortunate I'm afraid :P
@percusse Looks like a pointer to me. :) But I guess * is overload in the second case to have another semantics. Good question.
@PauloCereda Isn't it the same for def lalala(*args): usage. Then you can use stuff like sum(args) no?
TeXnet will take control over all typesetting systems in the world. No more lines with more spaces than letters on them. No more equations that pop out from the page and punch your eyes.
@percusse Yes, pointer. :)
@PauloCereda Is that the notorious pointer deal?
9:47 PM
@percusse Yep. :)
man I was expecting like go and retrieve address 0x234e3268ff34b
always less exciting than their names. That's why CS is boring :P
@percusse: on a second thought, might be an array:
In [1]: def foo(*a):
   ...:     return len(a)

In [2]: foo(1,2,3)
Out[2]: 3

In [3]: foo(1,2,3,4,5,6,7,8,9,1000)
Out[3]: 10
It's a wildcard for unknown number of arguments.
whatever that number might be
@percusse Here:
In [11]: type(a)
Out[11]: tuple
It's a tuple. :)
n-tuple to be geeky
9:51 PM
@percusse LOL
@percusse Or a recursion over the ordered pair. :P
@PauloCereda bunch of stuff
10:34 PM
Why I will continue to use Latex. Here is a result of wasting 2 hrs trying Adobe InDesign with MathMagic to typing mathematics in inDesign
It has no sense of line spacing. So math gets meshed into each others.
Amazing software. We are so lucky to have Latex.
11:07 PM
@Nasser lol :D
btw, tell me please, in the abstract, do you see plain Xm(q) or is the m as a superscript, in the following page: kmlinux.fjfi.cvut.cz/~hejdato1/temp/S20/cubic/cubic.html ?
@tohecz I see a superscript
@egreg cool. Which browser and in which version do you use please, Sir? :)
@tohecz Safari 7
@egreg Noted. Thanks
Basically, that's first successful try with htlatex, and the figures were manually redone, but it'll come
@tohecz In the unnumbered equation before eq. 2.1 the spacing is wrong.
11:17 PM
@egreg yeah, these things will be there. Is it this one?
@tohecz Yes
@egreg ok, thanks. good to know. But in the end, we're not sure how we want to draw the equations, whether rely on MathML support or on MathJaX
or (3rd option) on SVG
Well, time to go to bed, see you tomorrow!
@tohecz sadly I'd say to use mathjax (or what I do at work which is native mathml for firefox and mathjax for everyone else)
@DavidCarlisle well, somehow the point is that this part is not my part. Once I have everything in (X)HTML/MathML/SVG, it's someone else in the project who cares
we aim at wide accessibility with reasonable quality, and with good layout even on small tablets
@tohecz well unless they have total control over their readership (eg a classroom where they install the browser) the current situation is that pure mathml doesn't work in IE or Chrome, so between them that's a lot of non-readers unless you use mathjax as a polyfill
11:32 PM
@DavidCarlisle well, the original proposed workflow was to export math verbatim (whence my recent question on the site), but soon I realized that \DeclarePairedDelimiterX and whatever stuff people use would be impossible to handle.
@tohecz yes well we skip the conversion from tex problems by just writing the documents in xml to start with:-)
@DavidCarlisle I get your point, but we basically aim at wide mathematical society which knows LaTeX well and is unwilling to learn anything new
@tohecz I'm not sure I'd recommend it yet to a general audience, we have a very close set of authors (the software developers) and we have full time documentation (human) editors, and of course we have emacs:-)
@DavidCarlisle and it's your job or hobby? :)
11:55 PM
@tohecz job:-)
@DavidCarlisle ah ok :)
@tohecz latex is a hobby:-)
@DavidCarlisle so you develop something? Damn I have no clue what all the people here do when they're not academics!

« first day (1200 days earlier)      last day (2614 days later) »