@DavidCarlisle @DavidCarlisle I don't think that's it, the \ignorespaces does work. I also tested only adding \noindent to the hook and that causes the same issue. Looks like I'll have to somehow reproduce this in an MWE..
@PauloCereda well, that would be a cowsay-thingy most likely (that would be easier to script and doesn't require as much resources as it directly returns the characters you could print in a typewriter font). I can write a cow-file for the duck (but I think you could do that as well, cow-files have a really simple syntax).
ubuntu@ip-172-31-12-203:~$ latex-dev \\stop
This is pdfTeX, Version 3.14159265-2.6-1.40.21 (TeX Live 2020) (preloaded format=latex-dev)
restricted \write18 enabled.
entering extended mode
LaTeX2e <2021-05-01> pre-release-0 (develop 2020-10-29 branch)
L3 programming layer <2020-10-27> xparse <2020-03-03>
No pages of output.
Transcript written on texput.log.
ubuntu@ip-172-31-12-203:~$
@PauloCereda I have avoided flat thingy and I keep trying to work out if it's safe to use apt to uninstall snap to free up some space but I haven't dared yet
@PauloCereda Yeah, I was messing around with a document I'm making for my son and realised there was no sheep in \ducksay :p So I quickly remedied that. :p
@PauloCereda Don't really have time for chat anymore - but it's half term this week so I have a smidgeon more time than usual. And by not being here yesterday I've caused more hassle than if I had been with the pgfpages & geometry question ....
@DavidCarlisle Hmm, I don't think that works for me. After a good night's sleep, my problem isn't so mysterious any more, though. I still don't know how to best solve it, but I made it a proper question.
@PauloCereda Yup. It's one of those no-win situations. I don't have time to keep up with everything in the LaTeX eco-system, and I certainly don't expect the LaTeX team to notify package maintainers when there's an update that might break something.
@AndrewStacey well we did try: we notified quite a number of package maintainers over the summer. And I was just writing an issue for pgfmorepages (we realized this morning, that pgfmorepages doesn't load pgfpages, but rewrites it) when I saw your answer.
@PauloCereda and now the next step: is it "der Bär und der Schokolade", "der Bär und die Schokolade", "die Bär und der Schokolode" or "die Bär und die Schokolade"?
@UlrikeFischer There have been a few times where I've been caught out by updates to LaTeX, or L3, breaking one or other of my packages. I absolutely do not expect the team to grep through the entirety of CTAN looking for potential breaks. But on the other hand, I find it a little frustrating when the breakage is pointed out here (as that is often where it is) and there's a comment or other about how "That isn't how it ought to have been done" (ctd)
(ctd) without at least some acknowledgement that at the time, that might well have been the only way it could have been done.
So I have some really horrible code in the hobby package which implements an array structure in L3. That is now totally redundant and should be replaced because there is now \seq_item:Nn for accessing a specific term in a sequence. Unfortunately, when I wrote hobby then that wasn't there, so to have something that worked I had to implement it.
It absolutely wouldn't be how I would do it today, and I recognise and applaud the LaTeX team's goal of streamlining the inner workings to get rid of a large number of hacks that have built up over time. But until we get to the end goal, some of those hacks are going to need to be maintained and kept at least sort-of working until they can be gracefully replaced.
So I absolutely didn't mean my "answer" to be taken as either a solution or a criticism of the amazing work that you are all doing. And my 18th Century joke was meant as acknowledgement of that since a common complaint about TeX is that it is stuck in the 1970s and never been updated - I hoped that 18th C would be far enough back to make it obviously an appreciating joke.
But it was meant as "If you are about to send a paper off to the journal and need a fix for this right now, here's something that will work." as quite often that's the situation an asker is in.
@AndrewStacey certainly. But this is not the point here. You could see from my comment that we planed to fix it. You had enough time to delve into the code, to find a fix, upload it to your github and write a long answer. Why didn't you find the time to ask us a short question if this will clash with our fix and what our timeline is?
Nah, arara is awesome because there's awesome people behind it, specially this guy here: @TeXnician. We wouldn't be here if it weren't him. So listen up, ladies and gentlemen, boys and girls, ducks and mallards, bears and penguins, even ostriches: an awesome community is made of friends. The Island of TeX exists because of @TeXnician! <3
@UlrikeFischer In retrospect, that would have been a good idea and I'm trying to acknowledge that! But I didn't plan to do what I did. I had a look at the code to see what was going on to try to understand the question. I then played around with it a bit to see if I could use the existing hooks to fix it. After realising that I couldn't, I thought it at least worth writing up what I'd found out. I thought that my italic comment at the start made it clear that this was not a long-term solution (ctd)
I also didn't want to get caught in the crossfire between L3 and pgf. To be honest, I felt that several of the comments on the question had been of a tone that I didn't want to step between (from both sides).
Your comment on the question mentioned a package called firstaid. To my mind, this also doesn't speak of a long-term solution but of a quick fix while a proper solution gets found. So, clearly wrongly, I misread what the LaTeX teams intentions were with respect to this. There didn't seem to be any urgency in the comments.
My intention was just to say "While something proper gets sorted out, here's something that will at least work.".
@AndrewStacey I think there's a need to communicate clearly about the cases where something gets added to firstaid: the intention was to avoid individual authors needing to issue urgent fixes ('the team break stuff, we fix it'), but at least here that's not worked out. At least in part that stems from \shipout being particularly hard to handle
@AndrewStacey I know that you didn't plan it. But I wished people would stop for minute and consider if it is really needed to publish in a hurry a short term solution with potential to do harm in the future. This hacks have a tendency to be used regardless how many warnings you add before them. And in a year some question pops up "I'm using the hack from this answer in this local class and now it breaks ...". And then it is probably not you who will have to handle this.
@UlrikeFischer And so I've deleted the actual code. But this has always been an issue on this site. Someone comes here asking a question because they need an answer right there and then. Sometimes there's an immediate answer that solves the problem for that person and allows them to move on, but which isn't a good long-term thing to have on the site. That's the tension we deal with when answering.
@AndrewStacey yes I know. And if the OP had written anything about time pressure, I would have helped with some temporary fix myself yesterday. But there is nothing in the comments or the question that indicates that the fix is needed urgently.
@UlrikeFischer, @AndrewStacey As I said, I think there needs to be some thought about communicating the ideas used in firstaid: we all want users to be unaffected by changes, I think we also all see the need for the changes, it's a question of managing different people's effort
@AndrewStacey hi Andrew, Chris thought I was sounding rude/aggressive when asking about the 18th centry, so if I was my apologies ... wasn't intended as such ... more being concerned that you expressed your believe that we made a big step backwards.
@AndrewStacey somehting else ... I was hunting for an email address of you and ended up finding stacey@math.ntnu.no in the module repository list ... guess that needs an update, doesn't it?
@FrankMittelbach Hi Frank, thanks for stepping by. I didn't want to further clutter up the comments there so didn't respond further there. I was actually aiming for a compliment to the team! It was intended as a riff on the oft-heard complaint that (La)TeX is stuck in the past and I wanted to say that you were bringing it more up to date, so the intention was to imply that original-LaTeX was even older than 18th C.
So, again, my apologies for it not coming out right. You (all) have my utmost admiration for what you are doing with overhauling the LaTeX code and I don't envy the balancing act that you are doing with it between the needs of the past and the future.
@FrankMittelbach Yup, that's old! The redirections should still get to me from that, but better to use loopspace@mathforge.org. I do change it when I notice it - where did you find that?
@AndrewStacey thanks for the compliment then. Yeah I got it that it could be read both ways ... it is just that with us moving again (and I think on the whole in a good direction I rather sensitive about statements like "breaking tons of package" (not by you) or ... in your case implicit compliments that can be read as a complaint if you are already feeling that way.
@AndrewStacey it does not go back to you, it came back to me as a failure. And it is in the prefix registry for expl3 prefixes
@FrankMittelbach Ah, they've obviously finally realised that I no longer work there! Sorry to be an ignoramus about this, but I vaguely remember someone saying something about expl3 prefixes and that I should be added to them in the right places, but I don't think it was me that did it so I don't know where to go to correct it.
@FrankMittelbach Yes, use that one. That's linked to a domain that I've registered so isn't dependent (or linked to) any employment I may or may not have! (Though if I end up without employment at some point I may not be able to keep up the payments ... but we'll let that worry sit in the corner and not bother us for the time being!)
@AndrewStacey actually it is gone from there ... and the entries are empty -- seems these days we try to put github or other repositories in ... perhaps you can take a look and descide what should be there. it is in latex3/l3kernel/l3prefixes.csv
@Plergux you have to hit the fixed font button (it blobs up next to upload when you paste chunks of lines, if it doesn't add a space at the end and delete it, then it should be there)
@Plergux is moving the bubble-tail closer to the sheep's head (well, actually moving the sheep more left) fine for you, or should I include it as is? How should I acknowledge your creation? A footnote in the documentation, a comment in the sources, something more standing out?
@Skillmon I think making it more compact (bubble tail and sheep closer) would be better. :) As for acknowledgement, I don't know. Whatever is least work for you :p
@Plergux that would be comment in the sources (where nobody ever sees this :)
@Plergux but seriously, I think I should somehow make acknowledgements more standing out (well, you're the first one directly contributing an animal, so the need just wasn't there yet).
@Skillmon O.O really?! Yes, my user name is fine :) A comment in the source? Honestly, I have no idea :p If I was adding stuff from someone else into something I'd done I'd probably put a comment.
@DavidCarlisle Are there any specific holes in MathJax w.r.t LaTeX mathmode? Is there a description anywhere of the stuff it can't render?
@DavidCarlisle I'm trying to convince someone that although anything that MathJax can render can be converted into actual LaTeX, in the other direction it's not the case even for math.
@AlanMunn there is a list it can: docs.mathjax.org/en/latest/input/tex/macros/index.html as it can't really understand tex it basically can't do any package unless someone writes a javascript emulation of that package, but the emulation of the commands it does emulate is usually quite good
@Canageek in a table I'd go for \tstyle. The \text... variants are for short portions of text (read: just a few words), not for an environment in which everything should be using this. A big difference is that the \text... macros are defined short, so they can't span across paragraphs.
@Canageek the scope of the group. So I'd use something like \begingroup\textt\begin{tabular}...\end{tabular}\endgroup (I prefer to use the easier to spot and more prominent \begingroup...\endgroup construct over putting those things in braces, but that is more or less the same as {\textt\begin{tabular}...\end{tabular}}).
@Canageek If this should apply to every tabular in your document, you could use \AddToHook{env/tabular/begin}{\tstyle}.
@Canageek Well, adding it to tabular should definitely work (tabular doesn't reset any fonts, iirc). The "problem" is that floats (like figure) reset the fonts locally so that the place where you used that float in the code doesn't have any effect on the float itself. That's why \centering doesn't work.
@PhelypeOleinik :)
@Canageek and that's also why @UlrikeFischer proposed patching \@floatboxreset, as that is the macro which does all the font and formatting restores to the document's standards. So after everything was normalised, we can again change it, e.g., by adding \centering. She does that in the env/figure/begin hook because this way it only affects figures and no other floats.
@PhelypeOleinik admittedly the situation is a bit more complex. It prefers being shouted at "PDF" over "png", etc. And unlike the other engines, it supports artificial intelligence.
@PhelypeOleinik actually, it isn't. I think what the others are doing is strange. The priority shouldn't be casing dependent (well, as casing independent as possible), imho.
@UlrikeFischer the graphics .def for the different engines.
I'm reading the tutorial on LaTeX by overleaf, and in their section Adding Images they have some example code:
\documentclass{article}
\usepackage{graphicx}
\graphicspath{ {images/} }
\begin{document}
The universe is immense and it seems to be homogeneous,
in a large scale, everywhere we look ...
@UlrikeFischer it was made a few months ago, iirc, and I remember objecting it, because just because one of the big OSes has a serious shortcoming, LaTeX shouldn't share that shortcoming. Windows users deserve no sympathy :)
@Skillmon it wasn't made for windows users. We have already case insensitivity. It was made for the poor linux users, which constantly get code from me with \includegraphics{example-image-A} which then errors ...
@Canageek For data files? \pgfplotsset{/table/search path={<comma separated list of paths>}}, tex.stackexchange.com/a/286886 (in case you hadn't found that yet ..)
@Skillmon but even then, I'm not sure if it would be right to move PDF before png, it means an (always unnecessary on windows, often unnecessary on linux) additional file look up, before png is tried.
@PauloCereda -- Every time we see a sign "Covid testing" (and there are a lot of them around here), we misread it "Corvid testing". (Must be a lot of old crows.)
@LaTeXereXeTaL you don't have to. Except for @DavidCarlisle we're all friendly. @DavidCarlisle eats ducks, so he isn't so friendly. But considering you're a purple square, I think you should be safe.
@LaTeXereXeTaL and being here is the fastest way of loosing weeks of your life to coding in an esoteric macro expansion language. So stay here, sit on the couch and make yourself at home :)
@Skillmon no I was more thinking of if I have a degree and phd in mathematics but spent a few years in computer science post docs (writing latex2e, mostly:-) do I "come from" computer science or is that somewhere I just passed along the way.