« first day (3788 days earlier)      last day (1162 days later) » 

2:33 AM
Hi
How do I make the following answer a community wiki
0
A: List of all modules in ConTeXt

Jairo A. del Riometaducks module Available on ConTeXt Garden and GitHub for ConTeXt MkIV and ConTeXt LMTX. Documentation In Spanish (GitHub) and English (GitHub) Short description A fork of TikZducks rewritten using MetaPost/Metafun to work without dependencies on PGF and integrating with ConTeXt machinery bette...

I wanted to add an answer, but I forgot to click on the CW check
I see it was done automatically. Sorry for the noise
 
@JairoA.delRio Noto que disfrutas de líneas laaaaaaaaaaaaaaaaaaaargas al escribir duckmentación...estás siguiendo el estilo de arara (@PauloCereda)?
2
 
3:07 AM
@PabloGonzálezL Lo hice más bien para divertirme un poco y a mi novia le ha encantado. I'm only honoring the stereotype of "verbose Latin American" but my duckumentation isn't as colorful as @PauloCereda
@PabloGonzálezL I'm now writing a module for big integers using a native Lua library, but it will be terser as usually documentation should be more "technical" in nature. I want my first module to be a warm exception :)
 
 
2 hours later…
5:03 AM
@JairoA.delRio Now that's a great project...Duckumenting takes longer than creating :D
 
@PabloGonzálezL Indeed! I had the module upload ready back in February, but my gf, who isn't an actual TeX user, needed to see options to work with coloring and stuff, so I delayed updates until now.
@PabloGonzálezL As for the bigint thing, I think I could port it to LuaLaTeX, but it depends on time and user needs (after all, there are bigint packages using pure TeX, right?)
 
 
2 hours later…
7:08 AM
@UlrikeFischer that is a wonderful idea! I know a guy (ejem) that used to name the local file the same as the url (to not forget the source)...
...although I think that linking local files is risky at minimum.
 
@PabloGonzálezL I took another look and the vocabulary is mixed throughout interface3.pdf. In the headings they refer to them as packages, whereas in the text they are called module mostly. Still, I use my rule of thumb from above :)
 
 
1 hour later…
8:21 AM
@yo' ooh
 
@barbarabeeton re the third instance of fleqn: yeah, but I figured it's just there to prevent anyone passing fleqn to the document class lipics itself:

```
\def\lipics@opterrshort{Option "\CurrentOption" not supported}
\def\lipics@opterrlong{The option "\CurrentOption" from article.cls is not supported by lipics.cls.}

\DeclareOption{fleqn}{\@latexerr{\lipics@opterrshort}{\lipics@opterrlong}}
```
At least that's how it looks like to me as a LaTeX inexperienced :-)
 
@PauloCereda Fr...
 
@DavidCarlisle oh no
 
@Dr.ManuelKuehner For blurring text, you might also checkout this: tex.stackexchange.com/a/558671
 
8:51 AM
Ah, just discovered you have seen this already.
 
 
2 hours later…
10:22 AM
ooh
 
@JairoA.delRio regarding the tipa question: you are completly overdoing. The main problem is the wrong assumption that "the short commands must be necessarily enclosed in \textipa". \textipa does nothing more then switching to T3 encoding and this is simply unneeded with lualatex. So one can remove the command or define it to do nothing at all.
 
10:48 AM
@JosephWright it looks like a backend problem, and we should report it on the dvipdfmx list. But it looks a bit worrying that the color stacks are separated from the main color. This could pop up in more places.
 
11:00 AM
@PauloCereda anyway everyone knows pi day is 22nd July.
 
@DavidCarlisle oh
 
@PauloCereda more accurate than your feeble decimal approximations
 
@DavidCarlisle sorry
 
Oct 5 '19 at 10:51, by David Carlisle
@PauloCereda don't worry it'll soon be dinner time
 
@DavidCarlisle oh no
 
11:05 AM
@JosephWright I left a comment, I guess we need to make it work for now either commenting that out or guarding by a version number or ..
 
@DavidCarlisle Depending on the time of the day, March 14th can be more accurate
 
@samcarter_waiting_for_summer only by abusing time syntax
 
@DavidCarlisle True, but well...
> print (math.abs(3.14-math.pi)*100)
0.1592653589793
> print (math.abs(22/7-math.pi)*100)
0.12644892673497
(learning LUA)
 
@Rmano but once you are expert at Lua, you will be able to calculate that .13 is less than .16
 
11:20 AM
@Skillmon Yes, that's what confused me :(
 
@DavidCarlisle for that I said "that's true..."...
I normally teach my student to do "ballpark" calculation of filters by saying that $\pi \approx \sqrt{10} \approx \frac{10}{\pi}$ which is even more approximate but works a treat for cut frequency and Bode diagrams...
 
ooh pie
 
@Rmano I went through my school life believing that it is a universal law that all circular arcs had a radius a multiple of 7 so that calculations were feasible.
 
11:29 AM
@PauloCereda round ducks are for foie gras? (/duck)
 
@Rmano OH NO
@yo' from a friend of mine: :) twitter.com/rvf0068/status/1371197685676531712
 
@yo' see the starred comments to the right :-)
 
yo'
@DavidCarlisle ah :) that's not seen in the mobile view... So I missed that
 
@PauloCereda :)
 
11:36 AM
@PauloCereda ohhh :)
 
@samcarter_waiting_for_summer ooh
 
@DavidCarlisle I don't think that using the time syntax is any more abuse than using decimals in numbers
 
11:52 AM
@JosephWright @UlrikeFischer I'm thinking of dropping the "use xetex if fontspec appears" rulerorunlatex.js (which means adding % !TEX xelatex to example in lesson-14) and better documenting the runlatex-sk.js version to use in forums where there is no control over the site markdown and you have a more forgiving version that has more heuristics. the xelatex switch was added just to get xelatex at all before I added the !TEX comment support for direct control over the engine)
 
@DavidCarlisle sounds sensible.
 
12:11 PM
@DavidCarlisle Sounds good to me
 
12:24 PM
@UlrikeFischer That's why I don't usually touch LaTeX questions. AFAIK the issue was printing Unicode strings and \textipa does some switches as defined in tipa.sty
 
@JairoA.delRio I know what \textipa does and I know that it is unneeded, that's why I removed it.
 
@UlrikeFischer I understood the OP wanted something like that working with Unicode characters.
@UlrikeFischer Oh... I already see how does it work. Thanks for the remark
 
@DavidCarlisle sorry I added a few things on the site's js, also for experimenting
 
@JairoA.delRio no he asked about \textipa and the shorthands.
 
@StefanKottwitz no problem, good to have experiments tested on a real site:-) Blame @UlrikeFischer as she asked me how come that pstricks example worked, and I said it needed a !TEX latex comment but it worked wthout so I could tell you'd extended the logic:-)
 
12:34 PM
@DavidCarlisle should we better use latex or xelatex for documents with pstricks?
 
Pre-test providing useful :)
 
@DavidCarlisle >1300 users compiling >3000 documents on a Sunday, nice
 
@DavidCarlisle @UlrikeFischer How would one maintain those shorthands local (in LaTeX)? Maybe I overthinking, but \! and other spacing commands are being redefined
 
@StefanKottwitz do you have any existing stats for user numbers at your forums, combined?
@StefanKottwitz xelatex can only support a subset of pstricks functionality
@JairoA.delRio haven't been following this thread no idea what \! is being used here:-)
 
@DavidCarlisle I never made statistics of forum usage beyond the number of topics/posts/members, so I cannot tell the actual usage such as page hits or visitors
 
12:42 PM
@JairoA.delRio yes, tipa has imho an option which avoids the redefinitions. One can also like siunitx define things only locally. But this safety stuff is unrelated to the question how you define it at all.
 
@StefanKottwitz I think we can safely assume that is more than at learnlatex:-)
 
@DavidCarlisle \textipa lets \:, \! and other work as switches (for non-pulmonic consonants, retroflexes, etc.). @UlrikeFischer I was asking because it would allow to redefine \textipa so that spacing commands aren't affected
 
12:58 PM
@JairoA.delRio you can store them first under a safe name and then map locally to \:
\documentclass{article}
\usepackage{fontspec}
\setmainfont{CharisSIL} % ctan.org/pkg/charissil
\newcommand{\textrtailn}{ɳ}
\DeclareTextComposite{\tipacolon}{TU}{n}{"0273}

\begin{document}

Hello world!
ɳ {\let\:\tipacolon \:n} a\:b \textrtailn
\end{document}
 
@wtsnjp Thanks for the issue report
 
@UlrikeFischer Oh, nice. I like more the Lua approach (it's simpler to do things without unwanted layers of complexity), but I suppose this is the right LaTeX way.
 
@JairoA.delRio well the lua way has the disadvantage that it is lua only. We have to consider other engines too.
 
@UlrikeFischer Yes, it makes sense, (especially) for XeLaTeX, to do your way :)
 
 
1 hour later…
2:09 PM
@JairoA.delRio You can do \usepackage[safe]{tipa} and the commands will not be redefined.
 
2:27 PM
@egreg I was asking for other reasons (grouping to be specific). Ulrike's example made things clear to me
Btw, is it my impression or are braces preferred over \begingroup and \endgroup in LaTeX? The latter are rather common in ConTeXt definitions
 
@JairoA.delRio I think because {...} is documented for LaTeX whereas \begingroup ... \endgroup is TeX primitive syntax not in Lamport's book
 
yo'
@JairoA.delRio I wonder, shouldn't it be \startgroup...\stopgroup in ConTeXt?
 
@JairoA.delRio just easier to type?
 
@JairoA.delRio in examples like the one above I use braces as it is shorter, but in code I normally prefer begingroup/endgroup or the expl3 variants as it is more readable.
 
@yo' lol, nice question. In fact, we have \start and \stop, just like that, which aren't the same but do a similar work.
 
yo'
2:42 PM
@JairoA.delRio :)
 
@samcarter_waiting_for_summer Considering ConTeXt code sacrifices ease of writing for readability, it is rather an argument not to use braces (in ConTeXt, of course). Sometimes LaTeX sty files are really difficult to read because of shorthands.
@JosephWright I didn't know about that (and I haven't read the book either). So, is it because of custom?
 
@PhelypeOleinik thanks for the test, and it is rather a pain isn't it? Unicode is far away in some cases ...
 
3:23 PM
@JairoA.delRio I think it's that way in the older LaTeX sources ... let me check
 
@UlrikeFischer You're welcome :)
@UlrikeFischer It is! Nothing that works everywhere. But I think for Linux I'd bet on Okular and Evince (the rest are less used)
@UlrikeFischer Hm... Also I have a quite old distro, so maybe things changed along the way
 
3:53 PM
@PauloCereda -- huh?
 
@barbarabeeton no idea. :)
@barbarabeeton ooh
 
@PauloCereda -- German strings together actual existing words so if you+ separate them again, you will know what is being talked about. But compare this situation with French, that I experienced at a TeX conference near Paris. There was simultaneous translation, which was mostly very nice, but two words that sounded almost the same confused one translator. "colombe" is not the same as "colonne", and talking about two doves with respect to laying out a page is preposterous. I laughed my head off.
 
@barbarabeeton oh my!
@samcarter_waiting_for_summer oh no twitter.com/hyxpk/status/1371491389553438721
@DavidCarlisle ^^ YOU ARE FAMOUS
 
@PauloCereda hmmm
@JosephWright ^^
 
@DavidCarlisle ooh
@JosephWright ooh
 
4:08 PM
@PauloCereda ooh, the bear looks very happy
 
@PauloCereda Hey!
 
@wtsnjp ooh hello :)
 
@JosephWright Thanks for transferring the issue to the backend!
 
@samcarter_waiting_for_summer see the duck on top of the bed wearing a pi medal. :)
 
@wtsnjp あなたの活動を追跡するためにパウロを非難する
@PauloCereda you are blamed ^
 
4:12 PM
@DavidCarlisle quack
@DavidCarlisle ooh I am famous oh no
 
@PauloCereda Nice detail!
 
@PauloCereda DUCKS WATCHING YOU (ME)
 
@wtsnjp :D
 
@PauloCereda -- Oh, my! A camouflaged armadillo!
 
@barbarabeeton siesta time :)
 
 
2 hours later…
6:24 PM
I'd like for inparaenum to always have italic i), ii), iii), ...
With `\usepackage{paralist}` I tried

1. `\setdefaultenum{\em i \rm )}{}{}{}`
2. `\setdefaultenum{\textit{i})}{}{}{}`
1 has the downside that in italic environments (e.g. in definition environments of my class) the i) becomes roman
2 doesn't work as i is embedded in a TeX command
 
6:40 PM
@ComFreek \rm hasn't been defined by default in latex since 1993...
 
Haha, I copied \em i \rm ) from my advisor who was very well alive even before the 90s :D
 
@ComFreek well \rm is defined (in article clss, not in general) as a legacy command emulating latex2.09 so it always forces upright roman not following any existing italic or other shape. If you try it with other classes, eg knoa script it is simply not defined at all.
@ComFreek what do you mean by (2) above?
 
@samcarter_waiting_for_summer youtube.com/c/GreatBitArcade/videos this YouTube channel made lots of show openings with Animal Crossing!
 
@ComFreek oh is this enumerate package syntax for setting the label? \bgroup\itshape i\egroup probably works.
 
7:00 PM
@DavidCarlisle Wonderful, that works. Thank you!
 
@ComFreek brilliant package, enumerate I wonder who wrote it.
 
I thought it was you?
 
@ComFreek :-)
 
A big “Thank you” goes to David Carlisle,
because there wouldn’t be any optional arguments for the enumerated environments without the code from his enumerate package.
(from paralist docs)
:)
 
@ComFreek enumitem is actually better these days (and supports the same syntax) enumerate was probably the first package I wrote. a long time ago....
 
7:03 PM
?
 
7:21 PM
@StefanKottwitz I'm thinking if pstricks should force defaulting to xelatex then https://texwelt.de/fragen/123/wie-kann-ich-mit-luatex-grafiken-erzeugen ought to default to luatex I may extend my runlatex-sk.js version a bit.. to look for some likely luatex signs if there is no !TEX comment.
 
yes, that would be good. And btw. if a latex engine with postprocessing to pdf is ok then seeing pstricks can default to latex of cause
and I guess there's a good way to replace the many
if(t.match(/.../)){
editors[nd].insert("\\usepackage{...}\n");
}
by a loop
if(t.match(/expression/)){
editors[nd].insert("\\usepackage{package}\n");
}
iterating over a data structure like [ ("tikz|pgf", "tikz"), ("fancy", "fancyhdr"), ...)
 
@StefanKottwitz A (maybe irrelevant) comment on a fragment of the code I noticed yesterday: it seemed that checking for fontspec triggers xetex compilation? Is there a separate and prior check for lua specific things, since luatex also uses fontspec.
 
some already wrote me they are very happy with snippet completion since it's comfort and even helps building a real MWE for an answer
 
@StefanKottwitz I think this is the more robust solution for pstricks because xetex doesn't always work with it.
 
@AlanMunn comments welcome! to help incomplete snippets making compilable for readers and helpers it's a wild guess what may be needed to have a good chance, but the more specific, the better
 
7:33 PM
@StefanKottwitz No, I meant on a snippet of the checking code that @DavidCarlisle posted.
 
@AlanMunn sure!
- engine
- class
- packages
frame to beamer, \addsec to scrartcl, default article, ...
 
@StefanKottwitz By engine check you mean looking for a magic comment?
@StefanKottwitz The code snippet I'm referring to is:
21 hours ago, by David Carlisle
} else if((t.indexOf("fontspec") !== -1) || (t.indexOf("pstricks")!==-1)) {
engv="xelatex";
}
 
@AlanMunn David already did magic comments. It's looking for hints in the document code for selecting the engine if no magic comment was given
 
@StefanKottwitz Ah, ok. So it will find things like directlua etc.
 
@AlanMunn exactly!
 
7:38 PM
@StefanKottwitz yes I can clean that up a bit, meanwhile luatex defaulting works with
    } else if ((t.indexOf("\\usepackage{lua") !== -1) || (t.indexOf("\\directlua") !== -1) ){
	engv="lualatex";
    } else if((t.indexOf("fontspec") !== -1) || (t.indexOf("pstricks")!==-1)) {
	engv="xelatex";
 
so, if no engine is specified, it may use an engine based on matching keywords instead of just defaulting to pdflatex
 
but I'll chenge pstricks to use latex not xelatex (now that works)
 
@DavidCarlisle yes, that's better!
 
@DavidCarlisle and we may use declared strings again for button texts, inserted code comments (support for English, German, French)
I had issues with french accents in the code, what editor would you recommend? vim, pico, joe, all had issues (did not try emacs to not break the server)
 
7:43 PM
@StefanKottwitz yes I can but the buttons array back but only define it of not defined then you can simply define it before loading runlatex.js with the language texts?
 
@DavidCarlisle separating visible texts would be excellent for multi-language support
perhaps even in a separate file
since you improve the code all the time I have to frequently update a dozen sites ;-)
 
@StefanKottwitz they ought to work in any of them? or you can use hex escapes so é is \xE9
 
@DavidCarlisle would hex escapes work in the ace editor?
 
@StefanKottwitz stop putting in feature requests then:-)
 
@DavidCarlisle Our French friends featured you on page 28/29 here: gutenberg.eu.org/IMG/pdf/lettre42.pdf
 
7:48 PM
@StefanKottwitz that is javascript so in teh runlatex.js "\xE9" is the one-character string e-acute
 
@DavidCarlisle I learn javascript now :-)
 
@StefanKottwitz seems I write javascript more accurately than English
 
8:26 PM
user image
7
 
@AlexG Thanks anyway!
 
@Werner And comma compliant to boot.
 
8:53 PM
@Werner someone else who appreciates proper comma positioning:-)
 
@UlrikeFischer I kinda think I'd rather cheat here, and autoload mparhack instead of marking it as emulated. It clearly isn't as mparhack tries to solve a problem in the kernel regarding whether one can trust \if@firstcolumn, which you cannot. As far as I can see mparhack "fixes" this and that is more or less what it does, so might as well just load it instead of reinventing the wheel (might be nice to have it fixed in the kernel though, if fixable)
How does one get 4 points exactly, then I'll hit a palindrome
 
@StefanKottwitz localization done, pstricks using latex done, defaulting to luatex for \directlua or \usepackage{lua.... view source to see localized texts (just added ?? here) davidcarlisle.github.io/latexcgi/test-sk
@daleif answer a question badly and get 1 upvote and 3 downvotes
 
@DavidCarlisle hmm, might not be worth it just for a palindrome
plus the comma is in the way
 
@daleif or you could downvote three of egreg's answers.
@StefanKottwitz oh I was going to add a class attribute to added buttons to match your version.
 
@AlanMunn I must have missed the conversation on the comma separation. :-|
 
9:07 PM
@Werner @DavidCarlisle Certain people are sticklers.
 
@daleif We've had a stream of just-do-it-for-me questions, many of which show no effort at all. They provide opportunities for odd rep increments.
2
@AlanMunn :)
At least it's not 526.625k. :-|
@egreg: One of your recent interactions reminded me of marmot. I'm sure he's back on this site every now and again, just under a different userXXXXX name.
 
@StefanKottwitz llbutton class added allowing subtle styling at davidcarlisle.github.io/latexcgi/test-sk
 
9:28 PM
@Werner I guess so.
 
9:53 PM
@Ulrike Is there a way to define new link types using the new hyperref hooks? I would like to define a link type for the SetOCGState action type.
 
@AlexG what is the PDF code of such a link?
 
@UlrikeFischer It differs from other links by the content of the Action dictionary /A<</S/SetOCGState ...>>
I would like to be able to create links with hyperref look and feel
The colour could be inherited from internal (GoTo) Links
 
10:15 PM
Too many similarly sounding package names for the layperson :(
Is there any site listing package evolutions?
 
@ComFreek Simply assume that @DavidCarlisle is the author of whatever package you are looking at - chances are high that this will be correct
 
@ComFreek not really (other than ctan whose catalogue lists related packages for each package (for some definition of related) ctan.org/recommendations/enumerate
 
wow
"reverse enumerate"
for real? hilarious
 
@samcarter_waiting_for_summer did you know that I wrote beamer?
 
how about tikz?
 
10:19 PM
@DavidCarlisle Yeah, finally somebody to blame :)
 
@ComFreek NASA Need that all the time: 10,9,8,7,6,5,4,3,2,1
@ComFreek no that's why picture mode is so much better!
 
Oh, I thought the missing 3 was a joke which I didn't get
 
@ComFreek mission aborted?
 
I am too non-US-centric to being able to pull off a NASA joke now :-)
How come you like German words, though? :D
 
@DavidCarlisle About your latex online thingy: Is there a magic comment or something similar to switch off tex for a given code block (for example for code blocks that show log files or other non-tex code)?
 
10:30 PM
@samcarter_waiting_for_summer in the original version as used at learnlatex the run latex buttons are not added if tehre is no \documentclass and the whole editor insertion is skipped just leaving an html <pre> if the markdown has a {:no-edit:} pragma. ... But currently in the version @StefanKottwitz and I am working on for tje forums the button gets added everywhere. Forum users don't have access to the markdown so there are fewer options...
... I had wondered about a comment or perhaps if you use forum supports the ```latex type syntax highlight block like this site, we coudl skip the processing if anything other than tex is specified...
 
@DavidCarlisle Yes, I understand there is not much which can be done about the buttons etc, I'm just wondering if it makes sense to wrap a .log file in an article class document and then trying to compile that ... maybe with something like % !TeX Program = none could just tell the program to do nothing?
 
@samcarter_waiting_for_summer yes tex none could be made to work but actually it would be better if you coudl access some out of code control, if for example you wanted a fragment of bash script or a latexmk config, a % comment wouldn't be a great insert. I just looked at texwelt but I didn't see anything equivalent to this markup allowed here
```bash
export TEXINPUTS=somewhere
zzz
exit
@samcarter_waiting_for_summer that adds a specific class to the generated <pre> for teh code block which can easily be detected by the javascript.
@samcarter_waiting_for_summer does topanswers have anything similar?
 
@DavidCarlisle I don't think code fences are supported at texnique (they seem to be interpreted as inline code)
@DavidCarlisle no, topanswers just uses an existing library for code highlighting (which supports code fences :) ) [I would have to look up which one]
 
@samcarter_waiting_for_summer ues I was just experimenting a bit at texwelt (I don't have an account but it lets you draft a post anyway)
@samcarter_waiting_for_summer put a feature request in to @StefanKottwitz :-)
 
10:51 PM
@DavidCarlisle I though I try the famous person first :) (@PauloCereda said so, it must be true)
 
@samcarter_waiting_for_summer never trust a duck
 
@DavidCarlisle It is after dinner time - if a duck is still around, it must be a clever duck
 
@samcarter_waiting_for_summer fortunately it is alledged that the world is round, there may still be time in Brazil.
 
The markdown base syntax doesn't have code fences but indentation, that was the base for TeXwelt and TeXnique: daringfireball.net/projects/markdown/syntax#precode
I can be implemented of cause.
 
@StefanKottwitz so there is no way for the end user to affect the css class? we could have a %!TEX none comment ooh it just occurs to me we could not insist on the % so you could put # !TEX none in a bash file and skip the runlatex processing? (@samcarter_waiting_for_summer)
 
10:58 PM
@DavidCarlisle On behalf of @PauloCereda: Oh no!
 
@DavidCarlisle a magic comment %!TEX none sounds good as it's consistent (the "text" engine) and not another syntax feature
 
@StefanKottwitz hmm I'll see what I can do....
 
@DavidCarlisle or just %!TEX text (editor, highlighting, just no buttons)
 
@StefanKottwitz yes that's what I do at learnlatex for no \documentclass, (to get no editor at all requires direct access to the markdown to add {:no-edit:}. but text and none magic comments might work even though they are processed at a different time (before adding the buttons/editor not when the button is clicked) but the user won't care about the processing difference
 
@DavidCarlisle Thanks a lot for looking into this (the duck was right :) )!
 
11:05 PM
@AlexG well GoTo links are a bit of a problem, as their subtype is set in the backend. I already wondered if one should change this. With a look and feel of URI it is rather simple:
\RequirePackage{pdfmanagement-testphase}
\DeclareDocumentMetadata{uncompress}

\documentclass[12pt]{article}
\usepackage{hyperref}

\begin{document}
\section{abc}
\url{xyz.de}

\ExplSyntaxOn
\pdfannot_link:nnn { URI }
 {
  /A
   <<
   /Type/Action
   /S/SetOCGState
   >>
  }{link text }
\ExplSyntaxOff

\end{document}
 

« first day (3788 days earlier)      last day (1162 days later) »