@DavidCarlisle Does picture do anything to spaces?

That is, I have

  \hbox_gset:Nn \g_cfr_cefndir_box {
    \begin{picture}(0,0)
      \unitlength=1mm
      \put(-25.4,25.4){
        \file_input:V \g_cfr_cefndirclawrblaen_tl
      }
    \end{picture}
  }

which nearly works. But it is not aligned at the left or the paper. Instead, there's a white gap, as if something is introducing a space. But even commenting the line endings doesn't get rid of it.

@DavidCarlisle ^^

@marmot You could disallow an empty domain. Logically dubious, but potentially existentially satisfying.

@marmot ^^^ I think I should open a prof. van Duck's account :):):)

All of you folks are really weird for being non-existent, I have to give you that.

@HaraldHanche-Olsen :)

@CarLaTeX Actually, I have had some issues with non-existence myself: ↓↓

@HaraldHanche-Olsen LOL

@CarLaTeX The stackexchange software was really unhappy with the title of that question. But it did grumpily allow it.

@HaraldHanche-Olsen Oooh

@cfr is that within an expl3 context where white space is ignored?

A new package by @LoopSpace is on the works!

github.com/loopspace/savenodes

Too broad?

@Johannes_B is "don't use vim" an acceptable answer?

@DavidCarlisle No.

@cfr also why -25.4 rather than (say) -26

@marmot No, my Italian is too weak to be @CarLaTeX ...

@ChristianHupfer Non puoi essere CarLaTeX, perché lo sono.

@ChristianHupfer Sono italiano dela Foresta Nera. :)

Do we have a suitable dupe for that one?

@DavidCarlisle perché lo sono io (in that case, the subject is mandatory)

@PauloCereda della: I don't know why foreigners hate double letters :):):)

@CarLaTeX Sei un falso, cosa sai dell'italiano?

Off-topic?

@CarLaTeX Sonno ittallianno della Foresstta Nera.

" Only the true Messiah denies His divinity." - Life of Brian

@DavidCarlisle I have practiced it for decades!

@PauloCereda Only if your are spleeping (sonno) you can write that :):):)

@CarLaTeX you probably grew up in texas speaking Spanish and have only taken on this fake "CarLaTeX" Italian identity recently to try to claim authority on the matter of pizza toppings.

@DavidCarlisle I'm Italian, I have a witness (@egreg)

@DavidCarlisle Do you want to answer that? tex.stackexchange.com/questions/365834/…

@DavidCarlisle ay chihuahua

I am a duck

@CarLaTeX poor choice of reference, no one believes him

@DavidCarlisle :P

@DavidCarlisle isn't the design services bloke?

@PauloCereda only if you believe me

@DavidCarlisle ^^ @egreg exists. :)

However, I see no pizza!

@PauloCereda yes I think "It must be true, I saw it on twitter" is what they teach at school these days.

@DavidCarlisle Exactly! Historic facts are only real if they are properly tweeted. :)

@PauloCereda No sono Italiano di Napoli :-P

@CarLaTeX: Mi chiamo Christian ... ;-)

@ChristianHupfer "Non conosgeve parliamente, signor devo me parlo sono Italiano di Napoil quando il habitare de Milano."

@PauloCereda Was ist das deutschen Wort for Mittelschmerz ...

^^ that's terrible Italian, and MP script is all wrong :)

@ChristianHupfer ooh complicated

Ich bin ein Erpel und ein mineralwasser ist sehr gut, dankeschon!

@PauloCereda I am sure for British blokes and ladies it was Italian enough :D

@ChristianHupfer LOL

@ChristianHupfer actually not: Did you mean: Non conoscevi parliament, signor devo me parlo sono Irallano di Napoil quando il habitare de Milano

@DavidCarlisle Google Translate "corrected" it, didn't it? :D

@JosephWright This seems to be fixed? tex.stackexchange.com/questions/332110/csquotes-warning Close as such?

@PauloCereda What would have done Google for MP about 50 years ago ;-)

@ChristianHupfer :D

@Johannes_B I guess post an answer saying as much

@JosephWright You are the maintainer, please write an answer :-)

@PauloCereda It is very useful in the first lesson to learn the word "spoon" in Italian ;-)

@Johannes_B I've not changed anything!

@PauloCereda: You prefer the MP sketch in which Ron Obvious is trying to dig a tunnel to Java ;-)

@ChristianHupfer LOL

@JosephWright The question is from 2016, you are the maintainer since 2015. If you didn't change it, it wasn't an issue at the time?

Notice how @JosephWright appeared when a MP sketch was mentioned...

Coincidence? I. THINK. NOT.

@PauloCereda That's the glue of the universe that connects the Initiated ;-)

@Johannes_B I left a comment, it can probably be closed as dup unless clarified

@DavidCarlisle Alright, thanks.

@CarLaTeX Exactly. And I’m witness that Prof. van Duck exists, too.

@egreg you don't have lamstex 1 do you? (you being old enough:-)

@DavidCarlisle ooh

@PauloCereda somewhere I have a sun DAT tape that probably has a copy, but I wonder if I can find the tape, or a reader

@DavidCarlisle ooh

@DavidCarlisle The historical archive probably has it

@UlrikeFischer l3draw now has ellipse support ... still need to sort 'path usage'

@egreg And how do we know that you exist? Because @CarLaTeX said so? I sense a circularity …

@JosephWright I don't think so; LamsTeX was paid software; only version 2 has been released for free (not the manual).

@egreg Ah

@egreg That would make it more tricky

@JosephWright Meanwhile I rewrote the Soviet digit code without pict2e.

@JosephWright doesn't go back far enough

@egreg is Joseph's code robust enough for me to try aircraft design?

@egreg Ooh

@DavidCarlisle Goofness

@DavidCarlisle Broadly: the only issue at the moment is you'll have to 'hack' in the use of the path and stroking it: that should be sorted in a week or so

@JosephWright maybe I'll work on utf8 instead (but actually working on work, first)

@DavidCarlisle Good plan

@ChristianHupfer Clearly you're Italian, since your sentences are correct :);):)

@egreg Indeed! Quack!

@HaraldHanche-Olsen There is a video about % which proves that @egreg exists

@CarLaTeX fake news. As any phule nose, what follows % does not exist. I have a TeX file here which proves it.

@CarLaTeX And don't you dare change the \catcode`\%.

@HaraldHanche-Olsen @DavidCarlisle remembers it well :)

@CarLaTeX Now there's a carnivore that I might trust. If he exists.

@egreg I think you'll be able to use l3draw with proper path usage before the end of the month :) (I'll likely move it to l3experimental at that stage)

I've checked my l3draw progress. In tex/generic/pgf/basiclayer there are about 5500 lines once you skip the blank and comment lines. I have around 1300 lines of DocStripped code: allowing for the stuff I've deliberately skipped, I have probably got about 1/3 of the way through.

@DavidCarlisle github.com/latex3/latex3/commit/…

@CarLaTeX One sentence is correct ;-) My Spanish is better. Yo llamo Christiano y soy Aleman de Selva Negra ;-)

@egreg Perhaps you should add this as a comment to the relevant question?

@ChristianHupfer :)

@DavidCarlisle not at a pc, I'll see if I can remember to add it to my memoir todo list

@daleif thanks

Does anybody understand this question? latex.org/forum/viewtopic.php?p=104887#p104887

@Johannes_B I guess (s)he wants the middle parts of the first two equations a bit more to the right, such that they are centered at the center of "EQUATION" below.

Happily Heiko stole one of my ticks yesterday ;-).

@JosephWright I can't load the current l3draw version. I get an error about an unexpected comma at line 786 (\draw_xvec_set:n { 1cm , 0cm }). Should this work? Does it need a more current expl3?

@UlrikeFischer Yes, you need a full check-out

@UlrikeFischer Bruno has added tuple support to the FPU, I've also added \dim_set_inline:nnnn, etc.

@UlrikeFischer Install l3kernel as well as l3trial/l3draw

@JosephWright Ah. What is the build.lua call to build everything and put in a expl3texmf e.g. parallel to the git folder?

@UlrikeFischer texlua build.lua install --texmfhome <root>

@UlrikeFischer You'll want the latest check-in: I've hopefully got basic path usage working (so @DavidCarlisle can draw planes!)

@JosephWright It's installing (but seems to need some time ;-))

@JosephWright I did try, but...

\documentclass{article}

\usepackage{l3draw}

\begin{document}



\ExplSyntaxOn

        \draw_begin:
        \draw_path_moveto:n { 0 , 0 }
        \draw_path_lineto:n { 1 , 1 }
        \draw_path_use:n { stroke }
        \draw_end:


\ExplSyntaxOff

\end{document}

@JosephWright is ^^ supposed to work or is it user error

(/usr/local/texlive/2017/texmf-dist/tex/latex/base/article.cls
Document Class: article 2014/09/29 v1.4h Standard LaTeX document class
(/usr/local/texlive/2017/texmf-dist/tex/latex/base/size10.clo))
(/home/davidc/texmf/tex/latex/l3trial/l3draw/l3draw.sty
(/usr/local/texlive/2017/texmf-dist/tex/latex/l3kernel/expl3.sty
(/usr/local/texlive/2017/texmf-dist/tex/latex/l3kernel/expl3-code.tex)
(/usr/local/texlive/2017/texmf-dist/tex/latex/l3kernel/l3pdfmode.def))
! Undefined control sequence.
<argument> \LaTeX3 error:

@DavidCarlisle Did you update expl3?

@DavidCarlisle That's the error I just asked Joseph about, with a newer expl3 it is gone.

@JosephWright well I git pulled the whole lot, would texlua build.lua install in l3draw install the latest expl3? Package: expl3 2017/12/16 L3 programming layer (loader) hmm it lied when it said installing dependencies? will check...

@DavidCarlisle Works here but at the low level dimensionless stuff is in points, so you might want some units or something like \draw_transform:nnnnn { 1cm } { 0 } { 0 } { 1cm } { 0 , 0 }

@DavidCarlisle If you want the dep in your local tree, you've got to do that by hand (the message about dependencies is purely for unpacking)

@DavidCarlisle I need to get the 'mark local installs' stuff done I guess

@JosephWright clearly you've been slacking:-)

@DavidCarlisle :)

@DavidCarlisle On a serious note, the code is just about usable for simple drawings now, so I might do the docs on the existing functions then move to l3experimental (so it will ship in the next release)

@PauloCereda (can't find the message to reply directly to it ...) Yes, looks that way. Needs a bit of robustness testing first.

@JosephWright ^^^^

@DavidCarlisle In pgf at the low level each point has to be given as e.g. \pgfpathmoveto{\pgfpoint{0}{1}} so it's 'natural' to use the base vector approach, \pgfpathmoveto{\pgfpointxy{0}{1}} (the base vectors as 1cm long as-standard)

@UlrikeFischer Cool

@UlrikeFischer I'm trying to get all the bits and pieces done, but the order I do them in will depend on demands from testers :)

@JosephWright Is this alternative to PGF or on top of? When should I implement hobby's algorithm for it?

@LoopSpace It's a pure expl3 implementation using pgf as a strong model (though there are things I'm doing differently)

@LoopSpace If you are working at a low-level, the interfaces can be swapped out for l3draw ones, though you might want to take advantage of the expandable FPU (which can now do tuples directly)

@LoopSpace When I document the interfaces, I'll be noting all of the places it's different from pgf

@LoopSpace At present this is all very experimental, but the core idea 'the team need a more powerful graphics system than picture mode' is I think reasonable (@DavidCarlisle!)

@JosephWright finally: real application for your work

\documentclass{article}
\usepackage{l3draw}
\addtolength\oddsidemargin{-100pt}
\begin{document}
\iffalse
\begin{picture}(200,100)
\put(30,40){\line(1,0){150}}
\put(30,40){\line(0,1){60}}
\put(30,100){\line(1,0){20}}
\put(50,100){\line(1,-4){10}}
\put(60,60){\line(1,0){100}}
\put(160,60){\line(1,-1){20}}
\put(100,50){\line(0,-1){80}}
\put(130,50){\line(0,-1){80}}
\put(100,-30){\line(1,0){30}}
\put(100,61){\line(0,1){49}}
\put(130,61){\line(0,1){49}}
\put(100,110){\line(1,0){30}}
\end{picture}
\fi

@UlrikeFischer ooh a duck body

@JosephWright @DavidCarlisle : one with tikz one with l3draw ;-)

@UlrikeFischer when @JosephWright goes through the whole site giving l3draw answers to every tikz-pgf question, his rep is going to pass @egreg's

@UlrikeFischer Oooh: I've not done colour just yet, I guess you've just used base \color support :)

@DavidCarlisle, @UlrikeFischer So at least for testing it looks good to go to CTAN :)

@JosephWright Yes I added \driver_draw_fill_cmyk:nnnn {0} {0}{1}{0} (I didn't try to use complicated colors, only base one).

@UlrikeFischer Ah

@UlrikeFischer Stroke/fill colouring is relatively easy to do

@DavidCarlisle ... or also @egreg will start to answer TikZ questions with l3draw...

@UlrikeFischer I'll need to get l3color-extras sorted too I suppose, and move that across from the trials (it does mixins as in xcolor)

@JosephWright ^^^ It growing ...

@UlrikeFischer Ooh

@UlrikeFischer Some code might be nice for the manual

@UlrikeFischer don't show that to @PauloCereda or he'll never get his thesis finished.

@JosephWright: One is l3draw ;-). How can I set the default unit? Currently I have to set everywhere cm as unit.

@JosephWright Looks exciting. At some point, I'll have to have a play with it. Do I need to checkout the latest code from github or will a TeXLive update fetch it?

@samcarter @UlrikeFischer @JosephWright @CarLaTeX: So there will be a expl3ducks package soon? ;-)

@ChristianHupfer LOL

@LoopSpace github at the moment

@UlrikeFischer this I think:

German is beautifuller!

I wonder, if there's need-space in TeX, is there needs-pace in Lilypond? @PauloCereda @egreg

@yo' That would be applicable to some conductors.

@cfr I was playing around with prooftrees but many/most? of your sample proofs in the docs aren't compilable without some extra packages and some macro definitions. I manged to figure them out from the doc source, but it would helpful if you could provide a minimal preamble in the docs that would make the examples compile. (Not with each example, but somewhere.)

@yo' There's a famous model of syntactic processing called the Sausage Machine. Nothing to do with German, though. sciencedirect.com/science/article/pii/0010027778900021

@DavidCarlisle Yes, this works. But two \draw_transform:nnnnn seem not quite to do what I expect them to do, the (rotated) eyes disappear -- ah one need \draw_transform_concat:nnnnn for the second transformation.

@egreg lol

@egreg :)

@DavidCarlisle Yes.

@DavidCarlisle Because in the minimal case, regardless of page layout, I seemed to need to compensate for the standard 1" top and left margins. That is, (-1",1") seemed to be to top left corner of the paper. But I'm not keen on inches, so I actually used (-25.4,25.4) with \unitlength set to 1mm.

@DavidCarlisle I just need a way to align the image absolutely with the paper, without knowing the paper size. The dimensions of the image are determined by the paper size, so that the image is precisely the size of the paper. That works fine. With TikZ overlay, remember picture, I can align the image using an absolute position on the paper. That works fine. But if I use picture, it gets shifted right and I can't figure out what is causing that ...

@cfr Ok so only .00002pt diff so presumably not what you are seeing, so question is where are you placing this?

because I thought \put would put it at an absolute position.

@cfr you might prefer eso-pic

@DavidCarlisle But doesn't that also write labels to the .aux?

@cfr It does but you need care placing the 0 sized picture environment to avoid \lineskip moving it by a pt for example

@cfr no idea:-) (I wouldn't have expected it to)

@DavidCarlisle It is placed by a macro (one of two macros) which put it into a titlepage environment at different points in the document.

\NewDocumentCommand\packetpart { o }{% if a different image is used, won't the copyright statement be wrong?
  \clearpage\cleartoevenpage[\thispagestyle{empty}]
  \begin{titlingpage}
    \AddThispageHook{
      \box_use:N \g_cfr_cefndirdau_box
    }
	\IfNoValueTF{#1}{\mbox{ }}{
	  \begin{center}
		\vspace*{\fill}
		\efbox[backgroundcolor=bcol1!25!white]{\includegraphics[width=.95\textwidth]{#1}}\bigskip
		\vspace*{\fill}\vspace*{\fill}\\
	  \end{center}}
  \end{titlingpage}
  \int_gincr:N \g_cfr_packetpart_int

@DavidCarlisle ^^ Not sure how much sense this makes in isolation. I tested the change for the second of the two titlepage environments. So

\AddThispageHook{
      \box_use:N \g_cfr_cefndir_box
    }

was placing the picture. And the contents of \g_cfr_cefndir_box was set using

@cfr got to dash, I'd need to look at what the addthispagehook does, but you have a scary amount of white space there. a mwe would help:-)

  \hbox_gset:Nn \g_cfr_cefndir_box {
    \begin{picture}(0,0)
      \unitlength=1mm
      \put(-25.4,25.4){
        \file_input:V \g_cfr_cefndirclawrblaen_tl
      }
    \end{picture}
  }

in \AtBeginDocument.

@DavidCarlisle ?

@DavidCarlisle I know. I've so far failed to come up with a sensible MWE.

@UlrikeFischer Probably best to use the vector-based approach, then set the base vectors as required

@LoopSpace GitHub at the moment, clearing with the team for it to go to CTAN

@JosephWright Anything particular to test/play with?

@LoopSpace At present, all I can say is 'checkout GitHub, install l3kernel and l3draw, and have a look at l3draw.dtx for some minimal interface clues'. I'll probably pause doing code to get some docs written: things now do work for basic stuff, so testing is probably where I need to go.

@LoopSpace If the team are happy, I can have a CTAN release in a couple of days

@JosephWright Right, so if I'm going to play around with it, what would be most useful to you for me to try out?

@LoopSpace Hmm

@LoopSpace I guess there are two aspects: does stuff actually work (without 'surprise') and do the interfaces make sense

@LoopSpace In particular, do the function names 'work': I've stuck close to pgf but not always the same

@JosephWright Okey-dokey, I'll have a play around. It actually shouldn't be too hard to make it work with hobby due to the amazing design interface.

@LoopSpace Cool

@LoopSpace If it goes well, perhaps a pull request at some stage ;)

@JosephWright Actually, the hobby adaptation would be more suitable in the hobby package. But my spath3 package might have some stuff that could be more tightly integrated, and then it's only a short hop to knots and calligraphy.

@LoopSpace Cool

Just realised that two of the best Sci-Fi quotes are: "I'm afraid I can't do that, Dave" and "Everybody's dead, Dave". In short, don't go into space if you're called "Dave".

@LoopSpace Speaking of, I noticed that in braids you're using one of the things @JosephWright hasn't implemented yet. (\pgfpathcurvebetweentime)

@LoopSpace :)

@TorbjørnT. braids needs some serious updating. The list of "Must try implementing that one day" for braids is getting ridiculously long.

@TorbjørnT. I'm trying to prioritise: a few of the pgf functions are pretty esoteric

@TorbjørnT. As I complete pgf files I'm making notes on stuff I've either skipped 'for the moment' or are minded not to do: I suspect eventually we'll cover almost everything

Name that show: "I'm stuck with a man who'd lose a battle of wits with a stuffed iguana." "What's an iguana?"

@LoopSpace Red Dwarf?

@JosephWright Sure, and you should. braids was actually the only thing that showed up in the grep I did.

@TorbjørnT. :)

@JosephWright Yup. Given the anniversary, figured it was time to introduce it to the family.

@TorbjørnT. I'm checking each time I hit a 'complex' looking function

@LoopSpace :)

@LoopSpace 'They're all dead, Dave' had given the game away :)

@JosephWright Ah, should have figured.

@LoopSpace I put the code of the l3duck here github.com/u-fischer/duck-extravaganza/blob/master/… if you want to check.

@JosephWright I did once get a quote in one of my papers (arxiv.org/pdf/1301.5493.pdf - also notable for my choice of symbol for \qed)

@UlrikeFischer Ate some yesterday ...

@UlrikeFischer You'll want \draw_path_use_clear:n for the latest commits, and don't need \fp_eval:n in the arguments for the transform cm (or anything else, really)

@LoopSpace Cool

@JosephWright \draw_path_use_clear:n instead of every \draw_path_use:nor only some?

@UlrikeFischer All of them :)

@UlrikeFischer Names still somewhat fluid

@UlrikeFischer I changed it as it matches e.g. \box_use_clear:N. Feedback on the name welcome

@LoopSpace QED for “quick, eat duck”?

@JosephWright Oh you changed the name. I will have to update expl3 first. How do I make a build that also install documentation?

@JosephWright is there a newpath I thought I'd need that for the wings but path_use seemed to clear or do you mean you have changed that?

@UlrikeFischer Currently it doesn't: probably I should add an option for that

@DavidCarlisle I've changed it as use doesn't normally also clear stuff in expl3

@DavidCarlisle You can do \draw_path_use:n { stroke , clear } or \draw_path_use_clear:n {stroke } at present

@egreg Or 'Quod Erat Duck'? @PauloCereda might prefer 'Cogito Ergo Duck'?

@JosephWright Ok makes sense (although I'm more used to the postscript model where using it doesn't clear the path and there is an explicit newpath to scratch the current path)

@DavidCarlisle That's what I've changed to :)

@JosephWright yes I guess so:-)

@LoopSpace Frank says 'OK' to me moving the code to the release part of LaTeX3: expect it on CTAN very soon

@JosephWright I found a clone of the L3 github repository on my computer so I just pulled that and have all the code to play with. I'm thinking of a new way of doing knots, so I might just play with l3draw as part of it.

@LoopSpace :)

@ChristianHupfer A expl3ducks package would be a good reason to finally learn expl3 syntax :)

@samcarter texdoc expl3ducks reveals the truth: "Every time the expl3ducks package is used, 3 ducks are exploded." I think @PauloCereda might have something to say about that package. (Mind you, if someone writes the equivalent expl3kittens package then I'd put \usepackage{expl3kittens} in every document and every answer on TeX-SX.)

@LoopSpace Oh no! This should teach everybody to read the documentation carefully when using a package :) Although I cannot approve your expl3kittens plans, I know the perfect user to write this package: tex.stackexchange.com/a/387146/36296

@Werner nice round number:-)

@Werner Congratulations!