TeX, LaTeX and Friends - 2019-07-13

« first day (3177 days earlier) ← previous day next day → last day (1747 days later) »

12:00 AM

same as anything else gets compressed i imagine

well yeah i mean what compression format (or whatever the word is)

it's Flate compression

i'm not sure it's relevant to context

yeah pdf supports paths

so pdf is a viable output format for graphics stuff

12:27 AM

of course

sometimes that's all i use it for

can i paste code here

like a few lines

yeah

it should ask you if you want to maintain spacing or something similar

how

i don't know how it determines it but when i pasted some metafont code awhile ago it added another button next to "send"

/Times-Roman findfont
20 scalefont
setfont
newpath
200 200 moveto % cursor to x(200), y(200)
200 0 rlineto % go right
0 200 rlineto % go up
-200 0 rlineto % go left
0 -200 rlineto % go down
235 295 moveto
(HELLO WORLD) show
closepath
stroke
showpage

square.ps

7 hours later…

7:22 AM

Is it bad style to call a pgfkey directly when a key is passed as an argument? It seems more conventional to create a macro that is called.

For example, here is the top level macro call:

\fileloop[cappos={(7,0)}, cap=FOOBAR, blankpage]{keys.pdf}

But when passing blankpage through, I run ` \pgfkeys{/pdfloop/blankpage}`

The corresponding code is blankpage/.code = {\addblankpage}, where addblankpage is a predefined macro.

@FaheemMitha Why would that be bad style? The point of pgf keys is that you can apply all sorts of nice things to the styles such as /.list if you have a command with an argument and you need to loop over arguments, and much more.

@marmot I see. So that kind of usage is ok? And actually, I do need to loop over arguments. Do you have a simple example on hand?

Never mind. Doesn't look like it would be useful in my case.

8:10 AM

Sure:
\documentclass{article}
\usepackage{pgffor}
\pgfkeys{dc/.code={David is a #1.\par}}
\begin{document}
\pgfkeys{dc/.list={duck,koala,bird,goose}}
\end{document}

@FaheemMitha ^^^

9:03 AM

@marmot Thank you.

What about different values per iteration of the loop? Could one nest the values?

9:17 AM

@FaheemMitha What do you mean? (Yes, pgf supports arrays, if that's what you mean.)

@marmot So we have a macro with multiple key-value arguments. I want to call this macro multiple times, but each time, the arguments can be different, including possibly a different set of optional arguments.

@FaheemMitha Do you have a preferred syntax/an example?

@marmot Um. Nested list of lists? Yes, I'm working on an example, but it's a mess.

@FaheemMitha I can only guess.
\documentclass{article}
\usepackage{pgffor}
\pgfkeys{dc/.code={\pgfkeys{/dcs/.cd,#1}%
\pgfkeysvalueof{/dcs/name} is a \pgfkeysvalueof{/dcs/adj} \pgfkeysvalueof{/dcs/obj}.\par},
/dcs/.cd,name/.initial=David,adj/.initial=supportive,obj/.initial=duck}
\begin{document}
\pgfkeys{dc}
\begingroup\pgfkeys{dc={name=Ulrike,adj=sneaky,obj=bear}}\endgroup
\medskip
\pgfkeys{dc/.list={,{name=Ulrike,adj=sneaky,obj=bear},
{name=Enrico,adj=pizza,obj=lover}}}
\end{document}

@marmot Ok. I'll look at this once I've figured out what is going on with my current code. Thanks.

I just wish I'd got some more sleep.

DIs the macro \pages already defined, perchance?

I'm getting very strange errors when I try to use it.

9:56 AM

Yes, apparently it was a problem. Though I thought that the whole poing of that family thing was to avoid clashes.

10:19 AM

If I have a line like:

blankpage/.code = {\addblankpage}

it's not supposed to actually execute \addblankpage, correct? Only when the key is executed?

@marmot: Hi, mr. marmot! Regarding your meta question about addressing overlap, it is, of course, doable to implement such feature. However, in my humble opinion, it would also be nearly impossible to reach a reasonable usage point.

From a classical standpoint, this is a typical text similarity problem. We could take the easy route through lexical analysis, or go ballistic with the semantic counterpart. Since we are talking about code, I'd favour the first strategy. There are several metrics out there, like Jaccard. It's important to mention that these analysis are very time consuming.

@PauloCereda Thanks! What do you mean by "reach a reasonable usage point"? (Right now I wrote an answer on a cuboid (which the OP called cube). Wouldn't it be great if other answers on a cuboid would be automatically listed and linked to my answer? It would certainly help me to avoid reinventing the wheel, and others to have a survey of alternatives without the usual comments "related:....".)

@marmot I mean that it's very challenging to come up with a good solution. :)

For starters, alpha reduction is something that would not be caught so easily. Read this as "variable renaming". :)

10:35 AM

@PauloCereda Yes, but "good" is a relative term. A not-so-complete and/or not-so-ideal list is IMHO still better than nothing. Of course, I am aware of the fact that the search tools on our site are have considerable room for improvement, but still.

@marmot Agreed. :) But to check similarity between your answer and the others, you need to calculate your answer a with every answer b in the site, at runtime. You cannot cache this.

@PauloCereda Yes, sure, this is a tough question. Wait, wasn't there something like a page rank used by some almost unknown search engine? ;-)

@marmot They have a better computer than I do, I guess. :) As I said, there are different approaches, e.g, using machine learning to come up with some patterns.

@PauloCereda Probably it is easier to train a machine to answer 90% of the questions that come up here. ;-)

@marmot I'd favour this idea. :)

10:41 AM

@PauloCereda We could also train a machine to ask questions. Always clear and always with MWE. ;-)

Thinking of this: do we currently get questions by machines? I think there is strong evidence for that.

@marmot In 2010, I wrote a bot that would use Yahoo! Answers to get answers in order to help it improve the underlying ontology. :)

@PauloCereda Yes, and there is a LaTeX editor in which there is a button: Get TikZ code from TeX-SE! You only need to draw a sketch and the bot will do the rest.

@marmot :set dpc-mode on

Hello to everybody and buongiorno.

@StevenB.Segletes Hi Steven I have a problem with your code of this question. tex.stackexchange.com/questions/337356/…

\let\svitem\item
\newcommand\difbox[1]{\stackengine{0pt}{\color{gray!30}\rule{5ex}{1.15ex}}{%
    \color{cyan!60!black}$\mkern1mu\makeballs{#1}$}{O}{c}{F}{F}{L}}
\def\makeballs#1{\ifnum#1>0\relax{\bullet}%
  \expandafter\makeballs\the\numexpr#1-1\relax\fi}
\newenvironment{benumerate}
{\renewcommand\item[1][1]{\def\difficulty{##1}\svitem}%
  \def\labelenumi{\smash{\stackunder[1pt]{\color{cyan!60!black}%
  \bfseries\sffamily\theenumi}{\difbox{\difficulty}}}}%
  \enumerate}{\endenumerate}

\begin{benumerate}
\item[1] quando $\theta_0$ è zero, la traiettoria non è più una spirale intorno all'asse $z$; si tratta di una linea retta lungo l'asse $z$, per la quale $r=0$ in coordinate cilindriche.
\item[2] Se $\theta_0=\pi/2$, allora $\wbar{B} \perp \bar{u}_0$ e il moto della particella si riduce ad un semplice moto circolare di raggio $r=p_0/(qB)$ così come mostrato in figura \eqref{fig:subim55}.
\end{benumerate}

I have used \usepackage{stackengine} package but I have this error:

! Undefined control sequence.
\stackengine #1#2#3#4#5#6#7#8->\def \@STRT
                                           {}\if T\@strutlongstacks \if L#8\...
l.8473 \item
            [1] quando $\theta_0$ è zero, la traiettoria non è più una sp...

But I have another enumerate

%----------------------------------------------------------------------------------
%	New enumerate: fancy style enumerate
%----------------------------------------------------------------------------------

\let\svitem\item
\let\thebang\relax
\colorlet{bangcolor}{Peach}
\newenvironment{bangenumerate}
{%
  \fboxsep=1.4pt\relax%
  \renewcommand\item[2][\relax]{%
    \ifx!##2%
      \def\thebang{\makebox[0pt][l]{\kern-1pt\textcolor{bangcolor}{%
        \raisebox{1pt}{$\scriptstyle\blacktriangleright$}}}}%

@StevenB.Segletes Is is the problem: double different enumerate? Excuse me very much for the disturbe.

11:11 AM

Ok, here is some behavior that I wasn't exactly expecting.

\documentclass[12pt]{scrartcl}
\usepackage{pgfkeys}
\newcommand{\barxx}[1]{barxx #1}
\begin{document}

\pgfkeys
{
  /foo/.is family, /foo,
  default/.style = {barcode=,},
  barcode/.code = {\barxx{bb} \barxx{cc}},
}

\newcommand\xx[1][]
{
  \pgfkeys{/foo, default}%
}

\xx
\end{document}

It seems that default calls the code in /barcode key, even though I didn't intend for this.

11:25 AM

Without testing: shouldn't /foo, be /foo/.cd,?

12:01 PM

@marmot In which part?

Inside the \newcommand declaration?

In both. The following runs through:
\documentclass[12pt]{scrartcl}
\usepackage{pgfkeys}
\newcommand{\barxx}[1]{barxx #1}
\begin{document}

\pgfkeys
{
/foo/.is family, /foo/.cd,
default/.style = {barcode=,},
barcode/.code = {\barxx{bb} \barxx{cc}},
}

\newcommand\xx[1][]
{
\pgfkeys{/foo/.cd, default}%
}

\xx
\end{document}
Whether or not it is something useful I don't know.

@FaheemMitha ^^^^ (For some reason the reply button does not seem to work...)

@marmot Are you on a phone?

@FaheemMitha No.

Hmm. What browser/OS?

Hmmh, now it worked. MacOS, Chrome and marmot claws.

12:06 PM

@marmot oopsie, das ist nich gut.

@PauloCereda Well, it could be good.

@marmot ooh das ist sehr gut!

Ich bin ein Erpel

Ein Bier, bitte!

Hallo, heute arbeit ich!

@PauloCereda Phht pfft phhpht (<- that's marmot language)

Ich kann mit einem Eierlöffel Fledermäuse töten!

@marmot ooh

@PauloCereda "heute arbeite ich nicht" klingt besser ;)

12:08 PM

@TeXnician ooh I managed to understand it!

@PauloCereda Ducks are cruel…

@TeXnician oh nein

@TeXnician Depends on your preferences. If not working is the exception, yes.

@PauloCereda M-x not-mean-mode ;)

@TeXnician This sentence is from John Cleese. :)

12:15 PM

@marmot Are marmot claws part of the OS? :-)

@FaheemMitha No, but they are the most convincing reason why the reply button did not work. ;-)

@PauloCereda we just discussed that we need a tikzlings bat

what's the fascination with ducks and marmots

12:36 PM

Shakespeare on tikzlings: „O wonder! How many goodly creatures are there here!“

(The Tempest, V, iv, 184/85)

2

@UlrikeFischer A little girl yesterday at the lake: "Mama, Frau Ente!"

@tjt263 You must be new here.

3

@UlrikeFischer Ooh, Shakespeare.

(I thought I should say it, since Paulo didn't.)

@UlrikeFischer Also:

> O brave new world / That has such people in't!

@tjt263 To know all the story, read section 2 here: tug.org/TUGboat/tb39-1/tb121duck-tikz.pdf

12:55 PM

@CarLaTeX ooh entes

@FaheemMitha ooh

1:33 PM

On re-reading Ryan's pgfkeys answer (tex.stackexchange.com/a/34318/3406), I'm puzzled why he chose to set defaults without using the .default handler. Which, after all, is the reason it exists.

@FaheemMitha Ask him.

@PauloCereda I think he's left.

> Last seen Jul 4 at 19:33

But apparently still around occasionally.

Anyway, I'm going to use .default, and Ryan's answer be damned.

1 hour later…

2:42 PM

@UlrikeFischer From the fact that you can still answer TikZ questions I take that you did not yet finish the honey liquor? Could you please send it to me?

@CarLaTeX Oh, she recognized you even without car? ;-)

user image

@DavidCarlisle Oh, a commercial for detergent!

@marmot Perhaps because prof. van Duck was with me. Children love him :)

@CarLaTeX Wait, Prof. van Duck is Frau Ente?

@marmot oh nein

2:48 PM

@marmot no typical English Weekend... There were no marmots but there was ferret racing which is more or less like stretched marmots

2

@marmot No, he is a drake, of course. Frau Ente was swimming on the lake

@DavidCarlisle It reminds me very much of "Der Weisse Riese". You are sure that your information is correct?

@CarLaTeX When we were kids in Italy we always had Sorg-Ente.

@DavidCarlisle LOL

@DavidCarlisle When is the boat race?

@DavidCarlisle did you wear a fancy hat?

@PauloCereda no Morris dancing a very ancient tradition:

thebrackleymorrismen.org.uk/gallery-5

@PauloCereda next year

@DavidCarlisle oh

2:55 PM

The Boat Race 2019

The Boat Race 2019 took place on 7 April 2019. Held annually, The Boat Race is a side-by-side rowing race between crews from the universities of Oxford and Cambridge along a 4.2-mile (6.8 km) tidal stretch of the River Thames in south-west London. This was the 74th women's race, and the 165th men's race, and for the fourth time in the history of the event, the men's, women's and both reserves' races were all held on the Tideway on the same day. The women's race was the first event of the day, and saw Cambridge lead from the start, eventually winning by a considerable margin. This is their third...

@marmot Acqua di sorgente?

@CarLaTeX Si.

@marmot good!

@CarLaTeX We then had Sorg-Hahn as a synonym for "Aqua minerale frizzante".

@marmot does the sorghahn flow from the tap?

3:01 PM

@CarLaTeX No, not frizzante.

3:14 PM

@DavidCarlisle and do they wear those uniforms? I don't know why but they recall "A Clockwork Orange" to me, lol

3:36 PM

Football Badgers : animated music video : MrWeebl

FOOTY FOOTY

1 hour later…

4:36 PM

user image

@marmot ^^^ look what happened while we were away: They have been at it!

hey

does anybody know how can I remove Unpublished working draft. Not for distribution. from the ACM Template ?

Hey!

Whats the matter if I read in the info-bar '-2 User was deleted'?

@Richard looking at the source it seems you only get that if you use the [authordraft=true] option

@cis someone who once voted for you left the site and all their votes were deleted (happens quite often)

4:54 PM

@DavidCarlisle Ah, I understand.

@DavidCarlisle I don't have that in the sample-authordraft.tex

@DavidCarlisle I have found it in the acmart.cls and it fixed the problem. Thank you very much

@Richard well you provided no example, I was just looking through acmart.cls for "not for redistribution" it is in 3 places, as far as I can see all guarded by an if that is set by that option

Yes thank you

do you also know how can I remove the Woodstock ’18, June 03–05, 2018, Woodstock, bla bla at the end and also in the ACM Reference Format

@Richard if I grep for woodstock i see it's only used in the example files as \acmConference[WOODSTOCK'97]{ACM Woodstock so presumably your real document specifies a different conference

Yeah I think that is it. How can I remove that ?

5:04 PM

@Richard I have never used this class, it's only in the example file isn't it? simply do not put \acmConference[WOODSTOCK'97]{ACM Woodstock} in your real document

Ok. Yes this removes it after the Abstract but it is still shown at the end of the document under a horizontal line

$ grep -il woodstock acmart-master/*
acmart-master/sample-sigchi.tex
acmart-master/sample-sigchi-a.tex
acmart-master/sample-sigconf.tex
acmart-master/sample-sigconf-authordraft.tex
acmart-master/sample-sigplan.tex

@Richard the string woodstock only appears in sample files not in files that you should be using in a real document

@DavidCarlisle should I add an issue for the graphics+luatex problem in develop? Or do you already have a fix?

@UlrikeFischer probably best to have an issue even if I get a fix in this weekend I need to match what I did before to the current dev branch status and I'm not fully on top of the state of play there (again)

@DavidCarlisle I will add it asap.

5:13 PM

how could I use the formating and the stuff from the ACM Template in a new document without having of their examples

@Richard the examples are surely just examples they are not something you actually use directly, just \documentclass[sigchi, review]{acmart} copying literally the example line \acmConference[WOODSTOCK'97]{ACM Woodstock conference} is like copying the line \author{Ben Trovato} it makes no sense unless you are Ben Trovato.

5:55 PM

What's the recommended approach in \pgfkeys if you have a line like:

\node [font=\bfseries, align=center] at \cappos{\cap};

but you want a switch to just turn the line off (that's just a caption), in addition to controlling the values of \cappos (caption position), and \cap (caption string)?

I thought possibly a .style, but I'm not sure.

> Microsoft Office 365: Now Illegal In Many Schools in Germany

@DavidCarlisle emacs -nw for a decent reference. :)

6:24 PM

@UlrikeFischer I put a start of a fix in graphics-def

@DavidCarlisle I will look in a minute. In the mean time you perhaps like this xelatex problem (not with latex-dev, and probably a pgf problem):

\documentclass{article}
\usepackage{graphicx}
\usepackage{tikz}
\begin{document}
\newsavebox\mybox
\sbox\mybox{\includegraphics[scale=0.5]{example-image.pdf}}
\fbox{\usebox\mybox}

\begin{tikzpicture}
 \node{\fbox{\usebox\mybox}};
\end{tikzpicture}
\end{document}

will look, I suppose I will need @JosephWright's git rev-parse --abbrev-ref HEAD > branch.tmp") for setting tds-root in the graphics-def build.lua...

@UlrikeFischer i ran it in dev (as that's what I have at the moment) and it worked as far as I can see, two images, teh second one slightly to the right, same as in pdflatex?

user image

@DavidCarlisle I get this with xelatex ^^^

user image

@UlrikeFischer I got ^^ let me to a tlmgr update so I get a bleeding edge tikz...

@DavidCarlisle if you don't get this, can you sent me the log?

6:35 PM

@UlrikeFischer yes tlmgr is still updating, looks like I was a few days back

@DavidCarlisle with your fix I get not error anymore but no images either. The place is empty - I had this problem already yesterday when I tried to change the commands to remove the quote.

How much longer do I have to tolerate tirades of insults, which do not even get deleted?

@UlrikeFischer the luatex one?

@DavidCarlisle yes. does it work for your?

@UlrikeFischer I'll check in a bit...

@UlrikeFischer oh xetex one just broke, only these changed so looks like a pgf regression?

[ 1/38, ??:??/??:??] update: adtrees [356k] (39438 -> 51618) ... done
[ 2/38, 00:02/11:16] update: asmeconf [390k] (50967 -> 51583) ... done
[ 3/38, 00:04/10:45] update: asmejour [386k] (51233 -> 51567) ... done
[ 4/38, 00:06/10:37] update: babel-greek [1030k] (42010 -> 51628) ... done
[ 5/38, 00:09/08:20] update: bath-bst [759k] (47507 -> 51595) ... done
[ 6/38, 00:12/08:14] update: beebe [635k] (51544 -> 51603) ... done
[ 7/38, 00:15/08:27] update: biblatex-bath [462k] (48943 -> 51599) ... done

6:42 PM

@DavidCarlisle that would be a relief ;-). I will check with the older pgf from the backup and then make a bug report.

yesterday, by David Carlisle

@PauloCereda as long as you have someone to blame, and it isn't you, than it's OK if the software breaks.

@DavidCarlisle ooh

@DavidCarlisle not if one invest time to ensure that one isn't to blame ;-(

@DavidCarlisle it is pgf ...

@UlrikeFischer you are so demanding, I presumably just checked there was no error from luatex, and didn't look at the output.... fixed now...

@DavidCarlisle ooh

6:58 PM

@DavidCarlisle better ;-) But @JosephWright should probably change the build.lua so that the development branch of graphics-def installs in latex-dev too.

@Skillmon hi mr. rabbit!

@PauloCereda hi, mr. duck

In \pgfkeys, is it possible to get keys to execute in a particular order? That doesn't seem like a very TeX-like thing to do. And I don't see anything for something like this in the manual. I'm thinking something like /baz=c, /foo=a, /bar=b, but \baz should not execute till both /foo and /bar do.

@Skillmon how are you? :)

@PauloCereda good, visiting my parents this weekend. The presentation got a 1.3.

@PauloCereda how are you?

7:13 PM

@Skillmon cool! :) What was the max score?

@Skillmon In a hurry, but fine. :)

@FaheemMitha Almost sure this is not possible, and IMHO it actually defeats the purpose of having K/V pairs in the first place...

@PauloCereda Defeats? How?

@PauloCereda 1.0 is the best mark, 1.3 is pretty close. But I know no one who got a 1.0.

@PauloCereda why are you in a hurry? And is asking this counter productive?

@Skillmon ooh!

@Skillmon ooh I am burning DVD's as we speak. :)

@PauloCereda do you know of wow'en Wilson?

wowenwilsonquiz.com

user image

@Skillmon ^^ not my crowning moment. :)

7:22 PM

@PauloCereda got two, too

@Skillmon ooh ooh

7:41 PM

@PauloCereda Just did random clicking and got 4…

@TeXnician ooh

@TeXnician Your random is better than mine. :)

Which brings me to one of the best quotes ever in the chat room:

Aug 23 '12 at 14:09, by David Carlisle

@AndrewStacey nah nah na nah nah: my pool is bigger than your pool

:D

@PauloCereda Or you just walked along the wrong path in the probability tree ;)

@TeXnician LOLOLOL

@PauloCereda Maybe you shouldn't try to reach the duck pond while playing a quiz :D

@TeXnician Challenge accepted! :)

user image

@TeXnician ^^ I just clicked the first option every time!

7:46 PM

@PauloCereda 2 is too much, I managed 1/10 ;-)

@PauloCereda Well done. I knew that using a single random seed had its advantages :D

@TeXnician EXACTLY! :D

def random():
    return 1

@PauloCereda decided by a fair dice roll?

RFC 1149.5 specifies 4 as the standard IEEE-vetted random number.

8:07 PM

@marmot what happened?

2 hours ago, by David Carlisle

will look, I suppose I will need @JosephWright's git rev-parse --abbrev-ref HEAD > branch.tmp") for setting tds-root in the graphics-def build.lua...

@UlrikeFischer yes, I should probably set that up

@UlrikeFischer did you see what had changed in pgf to translate the graphics?

@DavidCarlisle I didn't try to check. I could probably make a tracing and compare with a newer version, why?

8:30 PM

@DavidCarlisle if I should guess I would say this commit: github.com/pgf-tikz/pgf/commit/…

8:57 PM

@MarcelKrüger do you know how one can retrieve the first value set with kpse.set_progname. E.g. after \directlua{kpse.set_program_name("luahbtex","lualatex")} i can get lualatex with \directlua{print("progname: ",kpse.var_value("progname"))} but find nothing for luahbtex.

9:39 PM

@UlrikeFischer I think this means it's not used

/* Engine support is a bit of a problem, because we do not want
 * to interfere with the normal format discovery of |luatex|.
 * Current approach: run |os.setenv()| if you have to.
 */

static int set_program_name(lua_State * L)
{
    const char *exe_name = luaL_checkstring(L, 1);
    const char *prog_name = luaL_optstring(L, 2, exe_name);
    if (!program_name_set) {
        kpse_set_program_name(exe_name, prog_name);
        program_name_set = 1;
    } else {
        kpse_reset_program_name(prog_name);

@DavidCarlisle hm, I'm not sure if I get it. Why does kpse_set_program_name use other arguments than kpse_reset_program_name? And what does the two functions do? (Background is that I'm trying to figure out what to do with this lualibs code:

if luatexbase and luatexbase.find_file then
  find_file = luatexbase.find_file
else
  kpse.set_program_name"luatex"
  find_file = kpsefind_file
end

luatexbase.find_file seems not exist and kpse.set_program_name"luatex" interferes a bit with the lualatex-dev project.

so I wonder if I can simply change this to find_file=kpsefind_file or if some checks are needed (and which one ...).

10:03 PM

@UlrikeFischer I was trying to trace the code back but it's in the kpse/luatex boundary C code and it's too late at night for me to understand it it seems, I would change lualibs just to use kpse directly anyway

10:14 PM

@DavidCarlisle it does use kpse, but it sets the wrong progname for lualatex-dev, so the tex-ini file must overwrite it.

@UlrikeFischer, @DavidCarlisle Looks like we are getting the Windows UTF-8 change in TL 'real soon now'

@JosephWright yes I saw the message.

« first day (3177 days earlier) ← previous day next day → last day (1747 days later) »

Transcript for

Jul '1913

$TeX - LaTeX$ TeX, LaTeX and Friends

\begin{chat}

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

$TeX - LaTeX$

site design / logo © 2024 Stack Exchange Inc; legal

mobile