@Rmano not sure if it helps in your case or is an alternative for you, but the IoT containers might help.

@Skillmon I was thinking about that... Is there somewhere a tutorial for dummies on how to use them (command line...)? I never used containers...

@Rmano If it is context you want, just installing the zip from pragma is very quick. If you do not want context, then I have no clue.

@mickep guess he needs both, LaTeX and context for testing.

@Rmano you can take a look at this tugboat article, iirc there's an introduction contained: tug.org/TUGboat/tb44-2/tb137island-docker.pdf

rollback :(

@DavidCarlisle oh no

@Skillmon thank. I'll have a look.

@mickep I want to check that the context mini-example still compiles, in a LTS environment. In check and home with -dev. I will fire a VM at home to see what's happening...

@mickep luatex users causing trouble again

@DavidCarlisle they are the wurst

@mickep What kind of Wurst are they? Bratwurst?

@samcarter Or bockwurst

@mickep yummy :)

With all the TL2007 on the main site, I feel like accidental having fallen in a time machine :)

@samcarter old stale ones circa 2015, seem to have turned poisonous

@DavidCarlisle it would have been easier if people just could stick to latex+dvips

@mickep too modern, what's wrong with dvitops

@DavidCarlisle maybe one could treat them like these Chinese century eggs?

@DavidCarlisle my husband says "rollback" didn't work in the Cold War either

@UlrikeFischer well that's OK then.

@CarLaTeX Mangia l'ananas sulla pizza solo se il mese inizia con la lettera maiuscola

@DavidCarlisle Exactly, you can eat pineapple pizza only where the months are written in capital letters :D

@CarLaTeX Makes sense now that the Romans didn't need lowercase letters as pineapples weren't yet in Europe

@Petəíŕd The Spring Wizard Code Golf - Didn't hear about that yet, but I am rather new to the whole Stack Exchange-thingie. I asked because the programming paradigm of the typesetting language/macro language TeX with its tokenization and separation between expansion and subsequent stages of processing is a bit different from everything else I encountred in the matter of programming so far. But i am definitely not skilled enough to be a code golfer.

@Cattleya I did a few in TeX, depending on the task at hand it isn't too bad (but obviously can't compete with the golfing languages).

@Cattleya it's macro expansion paradigm so not so different from other macro processing systems notably the C pre-processor #define.... #ifdef ...#endif

@Skillmon Where can Ifind them? ;-) I definitely cannot yet really judge on whether TeX is good or bad. There is so much, I am still in the stage of gaining an overwiew.

@Cattleya good or bad at what? at typesetting or as a programming language?

@Cattleya the answers I wrote in CG are listed here: codegolf.stackexchange.com/users/77083/…

@DavidCarlisle Maybe, but I were not yet in the situation of doing much C-programming, yet. There was an introductionary course about accelerated c++ and c# for beginners and that's what I encountered so far in the matter of programming in flavours of C.

@DavidCarlisle I cannot yet judge on both. How can I know what good typesetting is? Without learnng about typesetting. first. ;-)

@samcarter :D

@Cattleya well to put my question another way, If you decided not to use tex, would your alternative be a different prgramming language (such as python or C or ..) or a different typesetter (typst or sile or Word or Open Office or ..)

@Skillmon Thanks. Already the first one, the biohazard symbol is interesting.

@DavidCarlisle I guess I will need both when highschool is done and university begins. I don't know typst, sorry. And word is so much about clicking around without knowing what is going on under the hood. My impression is that with TeX and friends getting information on what the programs themselves do is much easier.

@Cattleya yes sure but deciding that question first affects how you assess tex. tex programming is undoubtedly a bit weird (but if you do it most days for 40 years you get used to it) but for most tex users that is not of any concern as they don't program in tex they just use it to typeset documents.

@DavidCarlisle One problem I have, e.g., with word is that I cannot easily look up how the program works in case it does not do what I wish it to do. So the look of my word files often is sort of a compromise when deciding whether to waste more time on fiddling around. I did not say TeX programming is weird. It's different from what I encountered so far, and thus delving into the matter required more focussing.

If things go as I hope, then in a few months I will study mathematics and, if I have what it takes, pursue a university career. In this case, it probably won't hurt me to be familiar with typesetting tools and be able to program some things in TeX. ;-)

@DavidCarlisle I saw some papers which looked like people who wrote them in TeX didn't know about peculiarities of TeX like the \endlinechar-mechanism causing the need to put % behind { or } sometimes to avoid unwanted horizontal spaces. So I like to know enough about TeX to be able to avoid such things.

@cis you can pay them

@mickep If I knew that this is really a vector graphic, I would be more motivated.

Currenntly I am fascinated by Convert control sequence with a variable number of undelimited parameters into a token list.

I guess under the precondition of all parameterss of the macro being undelimited, you can get info about parameters (amount of parameters and parameter character used with the parameter, usually #, but who knows for sure...) by looking at the \meaning and taking the things between the first : and the first -> for a list of character-pairs, the 1st component of a pair being the parameter character, the 2nd one being the parameter-number.

To be on the safe side, I'm currently trying to get an exact list of what the output of \meaning looks like for which cases. Does such a list already exist somewhere? Or do I have to puzzle it together from things like "TeX the program"? If it doesn't exist yet and I might do the work - would it be interesting for others too?

@Cattleya you can't really get that information.

@DavidCarlisle What information do you mean? Do you mean a list of what the output of \meaning looks like in which cases - this is the work which I consider trying to do in case it is interesting to others as well? Do you mean concluding from \meaning to a parameter text?

I know I cannot really conclude from the meaning to the parameter text. For example if in the \meaning I find ->->-> I don't know if the last argument is delimited by -> or ->-> or the replacement text starts with -> or ->->. That's why the limitation to undelimited arguments.

@Cattleya you can't in classic tex determine how many parameters as macro has. Even if you know it has undelimited arguments you can't distinguish in meaning#1 denoting the first argument from #1 being literal part of the replacement text. You can do things like applying some known arguments then seeing if the #1 gets replaced, but only if you know it only has delimited arguments and \foo123456789 won't give a syntax error.

@Cattleya see tex.stackexchange.com/q/305806/2388

@DavidCarlisle Isn't it the case that, when looking at the \meaning with a macro where arguments are undelimited, a #1 belonging to the parameter text in any case is behind the first : delivered by \meaning and before the first -> delivered by \meaning while a #1 belonging to the replacement text is behind the first ->? So that in the special case of allarguments being undelimited, one can take the result of \meaning apart for concluding about parameters?

@Cattleya undelimited arguments are no fun try

\documentclass{article}

\begin{document}

\catcode`>\catcode`&

\def\foo>1->2{aaa}

\texttt{\meaning\foo}

\end{document}

@Cattleya athough even if you omit the - so it has two undelimited arguments it's a bit tricky to handle

@Cattleya sorry meant # in the first line not &

@DavidCarlisle In your example the 1st paramete is delimited by - and thus the precondition of all arguments being undelimited is not fulfilled. Probably there is a misunderstanding. I'm not after detecting whether arguments are undelimited. I think about handling things under the precondition that all parameters and - thanks for pointing this out - the control sequence token itself are undelimited and you don't need to check that.

I think in this special case you can safely assume that things between \meaning's first : and \meaning's first -> denote parameter text in a form of character-pairs. The first character of a pair being the parameter-character in use, could also be >, the second one being the parameter number.

@Cattleya while it makes it easier to answer the question a constraint of "restrict to undelimited arguments" makes any code pointless in practice, as it's hard to think of any case where you know a macro has undelimited arguments but don't know how many there are or you have access to the source

@Cattleya yes as I said restricting to the undelimited case is not interesting

@DavidCarlisle I agree, but the challenge in Convert control sequence with a variable number of undelimited parameters into a token list seems not to be about practise any more... ;-)

If you look at it apart from all of this, perhaps a list of what the output of \meaning looks like in which cases could also be useful in other contexts? Or am I overestimating that?

In case I get to writing an answer to that challenge, I consider mentioning that practical use is limited. ;)

@Cattleya it would be very useful in many cases to be able to interrogate a user supplied control sequence name and determine what it is and take some suitable action, but that's a lot harder than it ought to be (and any answer to a constrained version of the problem like that might be useful for getting a few points but isn't actually useful) see for example the main loop in bm.sty that has to interrogate each token in a math expression and decide how to make it bold...

@DavidCarlisle Do I understand correctly: Your point of view is: Knowing what the result of \meaning looks like in which cases is not nearly as interesting and not nearly as non-trivial as the problem of deducing the exact nature of a token whose \meaning is displayed? If that is your point, I definitely agree. ;-)

(But stil I did not yet find a precise description/list of what tokens \meaning itself delivers in what cases. Do you think such a list is not that useful as having such a list doesn't help solving the non-trivial problem? ;-) )

@Cattleya yes a list of meaning types would be useful (it gets extended in luatex) that's what bm is doing in \xdef\meaning@{\meaning#1}\expandafter\bm@mchar@test\meaning@ code in latex hooks has to do similar things. The real missing feature though is access to the underlying definition object. ...

...The fact that to get any information at all you have to use \meaning which generates a string and throws away all information about the structure of the command, is a sign that something is not good, even if you can work around the issues often enough that no one has really suggested any tex extension to provide better support.

@Cattleya

@Cattleya that ended up being the unravel package

@DavidCarlisle I understand. Or more cautiously: I feel like I understand. I still have to think about everything a bit. For today I would like to thank you for the help and the friendly and pleasant conversation in which I received this help. I don't want to seem too ungrateful, but unfortunately I have to interrupt the chat at this point - my younger siblings are tugging at me and demanding my attention. ;-)