TeX, LaTeX and Friends

Atex

00:01

@mbert Thaaaaank you very much - That was the cherry on top for me!:)

Btw isn't there a way to upvote/like the responses in here? Feels quite rude that I can't acknowledge the time and effort you guys put into helping a newbie like me

cfr

@Atex you can give comments in here stars. that's more than you can do for comments on the main site ;).

David Carlisle

00:24

@Atex packages are more or less never deleted, so what you see is code written over many decades by hundreds of contributors sometimes Improving on existing packages, sometimes in ignorance of existing code, as almost all code is copyright the original author making some combined do everything package is often a technical and legal challenge,

2

cfr

06:54

@JosephWright the fonts in dist are trickier than the map file. pdftex.map is relatively easy, though I need more variables than I would ideally like. you can just make a minimal one in checkinit_hook() and drop it into testdir. but, the dist tree aside, l3build doesn't change the trees searched, correct? so it isn't just liable to pull enc files, tfms etc. from texmf-dist, it will also find them in texmf-local, ~/texmf, /usr/share/texmf, right?

cfr

07:05

@JosephWright is there a reason localdir doesn't get cleared? if you do things in the wrong order, it is easy to produce code which fails when built in a different order. this turns out to be why I needed l3build doc before l3build check. this seems odd when everything else gets cleared?

Joseph Wright

08:04

@cfr Yes, souds right

@cfr Er, we can't clear out other than in selected places as you have the unpack step for dependencies first

Paulo Cereda

09:44

Sunday quack

David Carlisle

@PauloCereda Sunday Breakfast

CarLaTeX

10:04

@DavidCarlisle You're mean

David Carlisle

Jun 29, 2017 at 16:15, by Paulo Cereda

@DavidCarlisle you are not mean :)

Paulo Cereda

@DavidCarlisle oh no

Skillmon

10:50

@Atex I was contemplating exactly that in the past, then wrapstuff came around and I stopped contemplating it :) Though I personally don't write many documents and don't need it anymore.

samcarter

11:01

@Atex There is a simple solution: instead of asking in chat, write a proper question on a Q&A site. If you get an answer there, you can upvote and/or accept it and it will help future users with the same problem to find answers.

Jasper Habicht

11:15

@Atex Wrapstuff is the project to redo all these wrapping stuff ... but in TeX it is a bit tricky to do this, so this project will probably still need some time for further development =)

I think the package author is happy to receive help from other contributors: github.com/qinglee/wrapstuff

Atex

14:17

@cfr Do you get a notification when I do it? I've "upvoted" (or, I guess you could call it "starred") your message along with other messages which I found particular helpful/important:)

@DavidCarlisle ooooh, this really explains a lot! There's been instances where I've wondered why there aren't packages for certain features when the possibility of executing it in LaTeX seems do-able. There're some really skillful coders, too. I guess the legal stuff is what hinders/obstructs progression in some departments of LaTeX, unfortunately. On the other hand, it's also understandable that programmers want the credit for their code/effort protected.

David Carlisle

@Atex I don't think stars generate a notification but as @samcarter says, anything that generates a response worthy of acknowledgement should probably have been asked as a question not in chat.

@Atex but would you expect all python prograns, or all fortran programs implementing similar features to be consolidated? why assume that would happen for tex?

@cfr can't you set TEXMFROOT so that it disables search of the entire texmf-dist tree and then you have full control?

Atex

14:32

I actually thought about doing that at first, but idk if it'd generate the expected result. I attempted something similar with this question: https://tex.stackexchange.com/questions/730023/tabular-vs-tblr-environment
I got some useful information from some of you people here in the comments of that post, but no one made an in-depth answer explaining the differences like this post, for example:
https://tex.stackexchange.com/a/66095/318574
Thus, I thought that perhaps I discovered this platform (stackexchange) too late. Either that, or the experts don't notice these kinds of posts (which I do…

@Skillmon Yeah, I also tend to ask questions regarding features in LaTeX which I may not use at the moment, but knowing that the solution exists just in case I might need it 1 day eases my mind! Then I don't have to ask when I need it and wait for an answer, because in that case, I did my homework;)

@JasperHabicht Will be exciting to follow its development. Though I'm nowhere near qualified enough to contribute to the development of the package, I'd definitely recommend improvement on the english documentation of the package first. If more people understand the package and thus can use it for more purposes (and correctly), perhaps more people will also report actual issues they encounter, thus alerting the developers about potential limitations/optimizations?

@DavidCarlisle I generally agree and think it's better to post questions for a greater benefit for all LaTeX users. It does require quite a bit more time and effort to construct a proper question, however. If the only reply to all of that is gonna be a comment(s), I just thought I might save myself some time and effort and post a question in here instead where it's even easier to go more in-depth with the comments since it's in a chat format. I'm still willing to turn it into a question if you...

... guys think it's worth it, even if the answer(s) to the question are just comments.

Atex

15:03

@DavidCarlisle Good point, idk why it didn't strike me earlier that LaTeX is a language of its own, just like Python. I guess that's because I've always associated code with python-like code in VS Code. Using LaTeX with the Overleaf Editor must be the reason I subconsciously viewed LaTeX as an open software where anyone could freely add/improve existing code so long as they had the qualifications, because I've seen numerous different names when reading package documentations

Jasper Habicht

15:49

@Atex Well, I tried translating the documentation into English a few years ago. But since English is not my primary language there might well be some parts that require revision. But I tried to be as close to the original as possible. Maybe things have improved overtime, so a new or optimised translation is needed. You could of course contribute to the package by optimising the documentation as well :)

samcarter

16:03

@Atex You could take all the information you received in comments and write an answer :P

David Carlisle

16:15

@Atex latex packages anyone can (and does) write them just like anyone can write a python package and put it on a server so pip finds it. you don't expect anyone to consolodate every similar python package into some uber-package. On the other hand not so many people have write access to the definition of the python language itself, similarly for latex, the core latex-team maintain the core latex implementation and language definition

Jasper Habicht

@Atex Oh, there is only a Chinese PDF at CTAN. Hm, the English doc is there, but it is hidden inside the .dtx file ... You can find it here: github.com/jasperhabicht/wrapstuff =)

... But as I said, it is probably not up-to-date

David Carlisle

@JasperHabicht @Atex the english doc is in texlive but not first choice so you need texdoc -l wrapstuff then choose 2

Jasper Habicht

@DavidCarlisle Oh, I see. So it is there =D Good to know how to access it!

And strange that CTAN does not offer it ...

David Carlisle

@JasperHabicht 当然，你也可以阅读中文文本

@JasperHabicht ctan just offer what the author uploads, texlive/texdoc show whatever the texdoc team can find to show,

Jasper Habicht

@DavidCarlisle 那当然啊！ I just saw that there is a link in CTAN pointing to the English version of the package which has an English doc

ctan.org/pkg/wrapstuff-doc-en

But it is someone else's translation obviously. Not mine ...

=D whatever

David Carlisle

16:29

@JasperHabicht not mine either:-)

Jasper Habicht

At least ... there are some options

cfr

16:45

@JosephWright I guess I was curious why things are setup that way ....

@DavidCarlisle is that the way to do it?

(can I?)

@DavidCarlisle and do something else for the other trees, you mean?

Atex

17:00

@JasperHabicht I think the only 2 options are yours and the translator of this document you referenced. I must say I do like the index added, but I don't think that's what's really needed (yet) when the documentation is so little in english atm. I do appreciate that you've included 1 additional example in the end, and that you've boxed the code, however. That definitely helps on the readability - Keep it up!:) All there's left now is to include a complete english documentation... or learn chinese;)

David Carlisle

@cfr I suspect so if you really want to stop the standard tree being used at all, rather than disabling each type of search individually, of course it then means you need to do more to set up a working system

Atex

@JasperHabicht That's a fair point. I do think you overestimate my capabilities if you think my contribution would be beneficial, however. I can barely use packages with predefined syntax, and I definitely can't understand how some of these advanced commands have been created. I can happily cheer on from the sideline and give feedback as a "casual user", however:)

That's a fair point, but I'd only be able to relay what you guys said in the comments. I don't know what the thought process which led you guys to your conclusions was, and so I won't be able to provide sufficient depth to the answer (although it might instigate someone to correct me/elaborate?). With this copyright stuff as well, idk if I'm allowed to share the screenshot of tabularray in Overleaf which you showed me earlier.
On a sidenote, I had completely forgotten about self-answers. I think I'll attempt 1 regarding center-aligned wrapped text, however

Ulrike Fischer

17:18

@Atex try it anyway. You will learn a lot answering questions - for example how much time it can cost to formulate a good answer and search for references and examples.

cfr

17:29

@UlrikeFischer :)

Atex

@UlrikeFischer Good point. I'm currently trying to create a question for wrapping text around center-aligned figures, and I've already faced challenges when thinking of answering it. It seems like the wrapstuff package's \wrapstuff environment only works if the text that's supposed to wrap fills out the gaps to the left and right of the figure (if you wrap a tikzpicture environment in my experience, at least). Another issue is that it seems like you can't do a forced linebreak for...

any environment. It just removes all of the text prior to the linebreak or gives you an error. Although this package is really good otherwise, it certainly seems to have some limitations upon further investigation and testing

David Carlisle

17:50

:6659978 they sound like questions you could ask, not problems that would prevent you asking a question, make an example that shows one issue and ask how to solve it

mickep

Also, it is not a big deal with questions. It is not that someone will be angry ...

Atex

@DavidCarlisle Yeah, I'm pretty good at asking questions... that's the only thing I do. I do like the challenge that @UlrikeFischer proposed about trying to answer, however. Only problem is that even more questions occurred as I tried to hypothesize an answer so I'll have to resolve them before making the post about wrapping text around center-aligned figures

But don't worry, I'm already working on formulating this question, and it'll probably be uploaded in a little bit;)

David Carlisle

18:06

@Atex good to set yourself a challenge:

Jan 2, 2013 at 16:46, by David Carlisle

@FaheemMitha The new years resolution is to get a tikz gold badge by answering tikz question with tex primitives, picture mode, or if really pushed pstricks, doing quite well so far this year.

Ulrike Fischer

@DavidCarlisle and the next challenge is to get the socratic badge?

David Carlisle

@UlrikeFischer I would do but I never managed to find the ask a question button

Jasper Habicht

@Atex The Chinese doc is not much more elaborate … so learning Chinese doesn’t really help you much here, I guess :)

At least it wasn’t when I last checked

It is still not. A few of the larger text blocks are just placeholder text with no meaning … sorry, all that could help you is studying the code

Ulrike Fischer

18:27

@DavidCarlisle that's why it is a challenge. With the question button every one can do it.

Atex

@JasperHabicht oooh, I suppose this makes it that much more difficult to find the correct solution to my newfound problem with the package: tex.stackexchange.com/questions/730527/…

samcarter

@DavidCarlisle As a natural linguist, you of course don't have lang = en in your texdoc.cnf :)

David Carlisle

18:44

@samcarter Ich halte mir alle Optionen offen

samcarter

@DavidCarlisle sehr klug!

Joseph Wright

19:13

@cfr Oh, I remember - it's always been assumed that the first thing you do is l3build clean before anything else, particularly for a ctan run

cfr

19:31

@JosephWright I ... er ... didn't ....

@JosephWright shouldn't the ctan target do that, in that case, rather than relying on people to know/remember?

samcarter

@cfr (workaround: I have a comment at the start of my BUILD.lua file with a list of all the steps to run...)

cfr

@samcarter hmm ... and you remember to read your build.lua?

samcarter

@cfr yes, because I can't remember the necessary steps :P

Richard

Hello everyone! I really don't want to derail the conversation in the chat, but I was wondering if it would be possible to load a lua library on Overleaf. Does anyone have some experience with that?

cfr

@samcarter so I need either a better memory or a worse one. I think I'm going to use l3buildchroot for ctan, but I'd like the utility code to do something sensible if somebody else tries to use it.

@Richard I thought derailing conversations was the point of using the chat room?

5

samcarter

19:50

@cfr yes :)

@Richard If you don't get an answer here, you could ask the overleaf support. (there is some overlap between these groups, but I currently don't see any of them in the room)

Richard

@cfr It seemed a bit inappropriate to ask directly and jump in a running conversation from my point of view, but maybe you are right :)

David Carlisle

@Richard not sure what you mean, but if th eLua code is in your project you can load it into luatex with \directlua{require('foobar'} as usual

@Richard they were only talking about tex, you didn't disturb any important cricket discussions.

samcarter

I bet the easybook author doesn't have any problems remembering all the l3build steps with their daily releases :P

Richard

@DavidCarlisle So I only have to load the library files in the Overleaf folder of the project and then proceed as usual?

David Carlisle

@Richard yes same as if you have a latex package that isn't in texlive just add it to the project and \usepackage{foobar}

Richard

19:57

@DavidCarlisle Ok, thank you sir

cfr

20:10

@samcarter what is it with easybook? I keep seeing references to it, but no idea what it is about.

David Carlisle

@cfr it's some basic book class but they have automated a ctan upload on every commit, so they add a comma and push it out to every texlive user on the planet...

Skillmon

@Atex to be more precise: I was considering writing that package when I already was in the phase of not writing many documents anymore :)

cfr

@DavidCarlisle oooh ... isn't that likely to break things, even aside from the issue of ctan/texlive/mixtex/user resources? but I probably shouldn't generalise from my own case ...

David Carlisle

@cfr well I haven't looked at what self tests they do, but clearly they push all the time

samcarter

@DavidCarlisle ... and now I'm wondering if there are any off-planet texlive installations? Does the ISS have texlive installed?

cfr

20:26

@DavidCarlisle no github actions I could see. gitee won't let me see without registering.

David Carlisle

20:48

@cfr yes I looked a while ago and I think I gave up at same place

@samcarter they probably use overleaf

mickep

@cfr Maybe the number of users is ... low ?

cfr

@mickep high chatter-rate though, if this room is anything to judge by.

Joseph Wright

@cfr Well yes, we could adjust - but if running manually I always do clean first as a matter of policy

cfr

@JosephWright I don't trust myself to remember my policies unless I script them.

that's why I wrote uniquify(). it's also why I almost forgot to write it.

maybe I should just rewrite the ctan target locally to

print("Paid!\n")

@JosephWright if you're used to gnu make, this seems counterintuitive ....

Joseph Wright

@cfr huh? All recipies I've ever used are make clean ...

cfr

21:06

@JosephWright make clean doesn't generally remove things which are built, so make clean ; make doesn't usually make everything from scratch. so unless you have an unusual recipe or use the souped up clean, you aren't rebuilding things which don't need to be rebuilt, are you?

Joseph Wright

@cfr I've no idea ;) Just what I've always done

@cfr Anyway, I'll see what I can do to adjust - it's an oversight

cfr

@JosephWright generally, make clean removes intermediate files. if you want to really start over, you need to make a target which resets everything and removes the build products.

at least, in recipes I've seen.

@JosephWright I often do make clean after a successful build because I don't need the intermediate files for diagnosis.

cfr

21:22

@JosephWright actually, I think you're right if you've not put your final package somewhere else. so normally I do run make clean (or the equivalent), but I think it would delete the package if it was in the build directory. however, it doesn't remove everything, even so. so you aren't rebuilding from scratch because it will keep some things, usually, even if they're in the build directories.

@JosephWright maybe?

cfr

21:41

... or maybe ... the UK TUG workshop Makefile descendant I have has make clean remove

aux auxlock bbl bcf blg ilg ind lof log lot glg glo gls mw nav out snm run.xml synctex.gz backup bkup toc xdy for aux.copy mw.mw mmz mmz.log acn acr alg ent hd vrb

so make clean doesn't remove final build products ... [pretty sure I wouldn't have changed the purpose of clean.]

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends