« first day (3710 days earlier)      last day (130 days later) » 

7:55 AM
quack :)
8:33 AM
user image
@MoneySetsYouFree lol
9:26 AM
@PauloCereda All the ducks
@PauloCereda You'd like the weather here: lots of rain
@JosephWright ooh
@JosephWright I can relate. We've been through heavy rains every day... :(
@PauloCereda Oh crap
@JosephWright it's bad, but we are summer, so it's expected to happen...
@UlrikeFischer @samcarter_prepared_for_xmas @Plergux The Great Tikzlings Christmas Extravaganza 2020 has 237 views so far!
@JosephWright ^^ LL does not work on IE 9
9:44 AM
@PauloCereda :)
@PauloCereda Where are you testing that???!!!
@JosephWright :)
@JosephWright ooh it's a secret :)
@PauloCereda Speaking of LL, the design work is 95% done
@JosephWright huzzah!
@JosephWright deadline will be met, finally :)
@PauloCereda Yay :D
@MoneySetsYouFree :D
@JosephWright Görünü┼če göre misyon Türkiye ile ilgili: learnlatex.netlify.app/en/mission :)
10:01 AM
@samcarter_prepared_for_xmas github.com/JairoAdelRio/context-metaducks
@PauloCereda A mail from CTAN was sent explaining that ConTeXt modules are mirrored from ConTeXt Garden. Thus, when TL2021 arrives the little module will be available, I guess
@JairoA.delRio yay, congrats! I should learn ConTeXt at some point, perhaps this will be the driving force. :)
@PauloCereda You'll probably have realized this intention when L3 is finished ;)
@TeXnician ooh L3
@PauloCereda Well, when LaTeX is not mandatory, ConTeXt is very nice. It has Lua for programming, MetaPost for graphics and TeX for typesetting. :) :) :)
10:16 AM
@JairoA.delRio ooh :)
Just for the lulz, I tried to graph a strange attractor and results were...
10:31 AM

function mp.mira_attractor(x, y, iter)
local a = 0.4995
local f = function (n)
return a*n + 2*(1-a)*(n/(1+n*n))^2
for i = 1, iter do
local w = x
x = y + f(w)
y = -0.9998*w + f(x)
mp.fprint("drawdot (%f, %f);", x, y)
i = i + 1


picture nice;
nice := lua.mp.mira_attractor(12,0,20000) xsized 10cm;

draw nice withpen pencircle scaled 1 withcolor darkred;
Does markup work here?
@Diaa I am not sure to understand you, but is this what you need?
\NewDocumentCommand \parentcommand {s m m m}
@JairoA.delRio the three backticks notation doesn't work, unfortunately... you need to indent the entire code by 4 spaces. You can paste the code and click the fixed font button...
@PauloCereda Nice, thanks.
@PauloCereda Oh, that's why this curious button shows up...
@JairoA.delRio ooh :)
@JairoA.delRio an attractive attractor
10:45 AM
@Rmano He, he. I was playing with ConTeXt features just to see if it was going to crash at some point. It didn't until very large values. And that's nice
@TeXnician Although it looks interesting, does anyone find L3 a sane way of doing things? I'm scared of it tbh
user image
@PauloCereda Very cute <3 That kind of pictures remind me of all the "wholesome" groups on Facebook I'm joined in.
@JairoA.delRio Well, it's as scaring as the ConTeXt internals. So I guess there's no difference on the programming layer. Otherwise, I think the concept is quite good, especially separating layers. The good thing about it is that one can simply wait and see ;)
@JairoA.delRio with L3 you mean expl3? Why should you be scared?
11:09 AM
@UlrikeFischer Conditionals, recursion, string handling, tables (or similar structures), scope (locals and globals)... Those are things a bit difficult to do using only TeX IMHO, but maybe it's rather a question of taste.
@TeXnician Ohhh... Comments on source files are quite interesting. Hans usually includes some nice goodies. Try:
\expression sqrt(17)*sin(pi/4)+3^(1.5)\relax
@JairoA.delRio Oh my :D
@Rmano I just released PGF 3.1.8a which should contain all the fixes you need.
@HenriMenke Thanks!
@Rmano I haven't tested any of this for edge at this point. I expect there to be some regressions. I hope I can iron the all out easily, but if it causes too much breakage I think I have to revert.
@TeXnician It's not exactly cutting-edge technology, but there are many other helpful resources if one takes some time to read source files. I found a Poisson point pattern algorithm and a binding to Metapost. Hans comment them as "fun" :D
11:22 AM
@JairoA.delRio Yes, some goodies are really nice. I always love to see such comments on the mailing list.
@HenriMenke I'll try to use it these days (I have to prepare a paper and a couple exams) and see if something breaks. Anyway, I agree that given the kind of change, if it starts breaking things the better option is to revert (even if it is maybe conceptually correct).
@Rmano The problem is really that PGF potentially has millions of users and that carries quite some responsibility for their documents. Some people who report bugs and then get angry when I don't fix them seem to not understand this (or they just don't want to).
@PauloCereda :)
@HenriMenke Hard work
@HenriMenke I understand fully. I have a default option for voltages in circuitikz that's simply buggy. I have a big fat warning (that no one ever read) in, but I do not dare changing the default...
@PauloCereda Yeah!
@PauloCereda ohh!
@JairoA.delRio May I steal the coffee mug?
11:38 AM
@samcarter_prepared_for_xmas Yes, please! It was actually for my girlfriend, but I decided to include it in the module as a Christmas gift.
I'd like to simplify the update procedure for MikTeX. Is there any harm in just running

`mpm --update && mpm --update --admin && mpm --update && mpm --update --admin && mpm --update-db && mpm --update-db --admin`?
@JairoA.delRio :) thank you! A nice Christmas gift!
12:00 PM
oh, this broke my installation "major issue: out of sync..."
Why is updating MikTeX always an endeavour of >= 20min?
12:39 PM
@ComFreek If MikTeX is not behaving nicely, there is always the option to use texlive instead :)
12:52 PM
user image
1:13 PM
Does anybody know why doesn't MiKTeX include a ConTeXt distribution as TeX Live does?
1:26 PM
@JairoA.delRio imho because it is too much work to get and keep it working. It was there at some time, but context has a tendency to change things.
@UlrikeFischer It makes sense. ConTeXt standalone distribution is already quite unstable
1:47 PM
@JairoA.delRio well the problem is not so much that things change, but that context doesn't actually cares if it is in miktex or not and if it is compatible with it or not. So why should the maintainer spent time to keep it working?
@UlrikeFischer Mmm... I get it
@PauloCereda @samcarter_prepared_for_xmas Duck walk: twitter.com/overleaf/status/1343192437104246785
2 hours later…
3:43 PM
@yo' Gorgeous! What a clever duck to find the way!
@samcarter_prepared_for_xmas :)
@yo' Did you use path decorations or something else?
4:09 PM
@samcarter_prepared_for_xmas yep, pathmorphing + rounded corners for the line and pathreplacing for the ducks.
@samcarter_prepared_for_xmas (the code is in the tweet)
4:29 PM
@HenriMenke unfortunately, I have found a regression (which is probably my fault in circuitikz, but alas --- just to comment it here; I will open an issue if I am lost and can't find the reason). The following snippet:
\usepackage[siunitx, RPvoltages]{circuitikz}

\pgfcircversion{} on \pgfversion
    \draw (0,0) to [
        ] (3,0);
compiles correctly on circuitikz 1.2.6 over Tikz 3.1.5b, but fails on 3.1.8a with
! Package tikz Error: Giving up on this path. Did you forget a semicolon?.

See the tikz package documentation for explanation.
Type  H <return>  for immediate help.

l.12         ] (3,0)
! Undefined control sequence.
\pgfutil@next ...flinewidth =\tikzscope@linewidth
l.12         ] (3,0);

I suspect is my fault because it works on circuitikz 1.0 and tikz 3.1.8a, so probably I have done something pesky in the middle (will start bisecting), but well, for now, let's see...
4:44 PM
...and in the bad commit I commented:
I should have done separate commits, I know.
:-( now I know...
@yo' Ups, I've been so enchained by the nice images that I did not notice the link to the code :)
@samcarter_prepared_for_xmas no worries :)
Hmmm... I really can't see what's happening. If I remove just one of the attributes (all of them are nodes added in the to path macro) it works. As soon as the number of nodes exceeds 3, it explodes. @HenriMenke, any idea on where am I doing wrong? Seems kind of a maximum expansion limit or something like that...
5:19 PM
I am facing a bug of biber (this one: github.com/plk/biber/issues/332), how can I update my biber to the latest dev version?
I downloaded a ZIP from here: sourceforge.net/projects/biblatex-biber/files/biblatex-biber/…. After extracting, I have got a "biber.exe" file. Where do I place it?
@ComFreek Are you using TeX Live or MikTeX?
I have now replaced biber.exe in C:\Users\ComFreek\AppData\Local\Temp\par-6e726f7578\cache-b32225defac3568d1e921af5a7245ca571b236aa\biber.exe
sounds dirty
let's see if it works
@ComFreek That's almost certainly wrong. But I know nothing about MikTeX, sorry.
it worked, heh
I would have desired a cleaner solution using MikTeX's package manager somehow
@ComFreek no, you should either replace the standard biber.exe in miktex/bin, or rename it e.g. to biber-dev.exe and put it somewhere in your path, and then call biber-dev.exe
5:26 PM
@ComFreek That's impossible. Please standby while we adjust Reality configuration.
@StephanLehmke :)
@UlrikeFischer You mean C:\Program Files\MiKTeX 2.9\miktex\bin\x64\biber.exe?
@ComFreek yes, but as I said you can also rename it, and/or put it somewhere else in the path (if it is found first there).
I find changing my PATH too hacky, sorry :D
alright, replacing it in the PATH I've posted last (also?) works. I'll leave it at that
Thank you all
@ComFreek sorry? you did install miktex and this adapts the path too. So why don't you want to add some other folder it?
5:32 PM
because uninstalling miktex will remove its path automatically, while it won't remove anything I add manually to PATH
adding something manually is a memory leak for me :-)
@ComFreek well you just add manually a biber.exe to Temp folder.
5:50 PM
@HenriMenke (as suspected) my fault, although I still am not sure why. I'll post a link to the patch solving the problem (you never know)... as soon as I finish my checks.
\printbibliography gives me "[Abc18]: title of paper, .... blah blub"

I want to have that in the middle of text, too. I tried `\fullcite{...}`, but that doesn't give me `[Abc18]`
Any ideas?
E.g. in the image here (tex.stackexchange.com/a/135950/38074) you see that \fullcite is missing the "[1]"
6:12 PM
@MarcelKrüger travis is telling us to migrate?
(I know, one does not read the doc :) ) The doc of source2e seems to have quite a few unresolved cross references.
@samcarter_prepared_for_xmas I think this will be resolved with the next update.
@UlrikeFischer Would be fine with me, but I'm not sure how to do it.
@MarcelKr├╝ger we could ask @JosephWright ...
6:30 PM
@UlrikeFischer I think I found it, it should be migrating now.
@HenriMenke that was the needed fix: github.com/circuitikz/circuitikz/pull/476/commits/… (I am not sure why, so it's more a workaround than a fix)
6:53 PM
@Rmano I have reverted the offending commit in PGF. I didn't think that this change made sense in the first place.
@Rmano 3.1.8b is currently building
@Rmano Anyway, I recommend that you use expandable conditionals where possible, i.e. instead of \ifpgfcirc@has@i ...\fi use \pgfcirc@if@has@i{...}{} which is defined as
Ah, ok. I will anyway let the workaround --- I have a sensation that it was my fault
@HenriMenke Thanks! will study this and apply as soon as I can...
@Rmano You can easily automate this definition of conditional as I did here in pgf-pie: github.com/pgf-tikz/pgf-pie/blob/…
@Rmano Another unrelated suggestion of mine would be to restructure the circuitikz repo into TDS form.
@HenriMenke wow, black magic... :-) thanks again
git mv tex tex-old
mkdir -p tex/latex/
git mv tex-old/ tex/latex/circuitikz
@Rmano The same for doc.
@Rmano Having the repo in TDS form allows using TeX Live usertrees without modification.
@HenriMenke Yes, I should do that before or later. Will simplify releases a lot.
...now I have to fix the hastened fix I did :-(
7:17 PM
@Rmano git revert?
7:38 PM
@HenriMenke No, I have changed things that I want to keep, so I'll push a new release with your nice conditionals and the extra groups, that are not dangerous but they still help (BTW, I checnked with 3.1.8b and that fixed everything)
...and then I'll learn to be a bit more patient and to make smaller commits ;-)
1 hour later…
8:59 PM
@HenriMenke BTW --- it seems that the fix for 3.1.8b is not related with the to path thing, is that correct?

« first day (3710 days earlier)      last day (130 days later) »