@StefanKottwitz I've got one of those too. I did all my lecture notes for the past several weeks on it. My handwriting is not much to write home about, though. But I love the tablet. It's way nicer to write on than the ipad using the apple pencil.
@Plergux Way back, I was at a Nordic math meeting in Reykjavík. A colleague of mine, who isn't the kind of person to use proper raingear, had forgotten his umbrella at home. He was trawling all the stores in Reykjavík looking for one, with no success. I guess umbrellas are mostly pretty useless there.
@HaraldHanche-Olsen I may lend it to my kids because they don't like writing so much (they love reading though) perhaps they have fun with it and so do practicing.
My only concern: it looks so cool with the calligraphic pen, that normal pencil writing may become more boring.
@Plergux That was my experience too. On my last day of my first visit to Reykjavík, the rain was not only horizontal, but the raindrops were huge. Each drop hurt when they hit my face, and they did so at an alarming rat.
@StefanKottwitz I actually use the fineliner when I write mathematics. But yes, the calligraphic pen is cool. I play with it from time to time.
@HaraldHanche-Olsen yes, electrical engineer here. Until today (while I'm teaching students) I still don't understand why we use j instead of i. They say so that you don't confuse it with a current, but j is also the current density, no one cares that you could confuse those two, it seems.
@HaraldHanche-Olsen Yes, it can get pretty nasty. And it's worse than hail (which can hurt pretty bad too) because it soaks you. I always feel a lot colder when it rains than when it snows.
@Skillmon :-) electr(onic) engineer also --- yes, the reason is not to confuse with currents (current density is rarely used unless you are doing semiconductor physics, really). But in the end is fashion --- for example, in Italy we use j for the imaginary unit, but v (and not u) for voltage, and no one has a problem to mix this with velocity...
@Rmano It's probably the same as with language. You read it from the context. Like "I took a bow" versus "she's wearing a bow" versus "He used his bow and arrows" :p
@PauloCereda what's annoying, too, is when you don't see the rain, but still get wet and cold when you're outside as the droplets are just really small and evenly dispersed.
@Skillmon Oh, I missed that. And I wouldn't have guessed what it's for, as I haven't encountered that convention before. Reminds of the UK, where I have seen people use a tilde below a letter to indicate a vector. On the blackboard.
@StefanKottwitz No, I haven't had the need, luckily. Been teaching a really small class this fall, and we had a big auditorium, so no problem with physical classroom teaching.
@HaraldHanche-Olsen maybe it's close to that. Where I grew up we rarely had this kind of weather, but here in Bayreuth the last three winters where just that. Close to no snow, but wet and cold air for weeks.
@Plergux like a drizzle maybe, but maybe a bit less rainy and more just wet (depends on what you call a drizzle). I don't know how to explain this (as a matter of fact, I don't even know how to describe this in German).
@Skillmon In my first contact with phasor notation, in high school, my teacher advocated for using ʝ (unicode 0x29d, even if at that time unicode did not exist) --- it's difficult to have complex numbers and linguistic confused, I think (but I can be wrong).
@Skillmon Heh heh heh... Well, we have a word in Icelandic, "úði" which is the same word that we use for the kind of moisture that comes out of a spray can. Kind of "misty" but more wet. :p (and I thought Germans had words for everything :p At least that's what the internet says. :p)
@Skillmon Heh heh heh... Well when you have a secret institution to tell you what words are and are not French. :| (though the general public probably disagree :p)
@Plergux well, if you consider those strange terms used in chemistry to describe organic stuff, the number of words in French is a bit bigger (infinite)
@Plergux Obviously the same word. Google translate translates úði into mist, or “tåke” in Norwegian. But unlike some, I don't necessarily accept what google translate says.
@Plergux It's "sirimiri" in Spanish (although only non-native knows that, I suppose), which is really "zirimiri" in Euskera --- it's a quite typical rain in the northern coast of Spain, so it must have a name. es.thefreedictionary.com/sirimiri
@HaraldHanche-Olsen No, me neither. :p When your native language is fairly underrepresented you learn pretty quickly not to trust these kinds of things. :p
@PauloCereda well, amplifiers at 12 V don't explode, modulo the occasional elcos mounted upside down. But still covid protocol is strict here - so mask + shield
@JosephWright I checked the new colorstack special in xelatex and it seems to work fine and quite similar to pdflatex/lualatex. I think one should add some abstraction for pdflatex/lualatex/xelatex to expl3, but I'm unsure if this should be \pdf_colorstackXXX or if this should go into l3color?
@UlrikeFischer I was going to say we extend the colour part now, then look at non-color aspects sepearately, but I see @MarcelKrüger has different ideas :)
@MarcelKrüger yes I though similar (it is used in the transparent package to add /TRX gs)). Perhaps one even shouldn't name the abstraction color but graphicstate or something like this.
@JosephWright Transparency (which can be seen as kind of color related) and most graphic state operators in PDF in general. (I recently e.g. looked into using it for text knockout settings)
@MarcelKrüger But for dvips one has to do that via PostScript anyway, so I'd expect us to manage it 'inside' an interface for transparency or (perhaps) graphics state without exposing any kind of stack
@JosephWright For an internal interface it's fine, I was mostly thinking that a public interface with this kind of functionality should not be l3color specific.
@JosephWright well in my (new) transparent code I have things like ` \let\TRP@pdfcolorstackinit\tex_pdfcolorstackinit:D \let\TRP@pdfcolorstack\tex_pdfcolorstack:D` which I would like to get rid off ... And there is pdfcol and pdfcolfoot which uses it too. dvips is naturally difficult, but it could be a no-op for now.
@JosephWright we could naturally also move the transparency code into l3color and keep it internal, then transparent would be only a stub defining the \texttransparent command.
@yo' with the newest latex format you could use \RequirePackage{structuredlog}, along with a longer line length it helps a lot to make logs more readable.
@JosephWright I mean the logs that the user can access, i.e. what ConTeXt writes into the .log file after a run. And there I get logs without forced line breaks at a certain width.
@TeXnician Ah, yes, the ConTeXt one: I guess as I'm a developer I want the Real ones (which mean finding the format, running the engine manually, etc.)
@JosephWright Well, the ConTeXt one gives me all information necessary (one can enable more tracing directives etc.) and ConTeXt pre-structuring it is a plus for me. But I understand that at a lower level one might have different expectations. Just wanted to mention that user logs are one of the other aspects I actually like when using ConTeXt :)
@yo' Philip Lehman wrote a package called logreq to go with biblatex, to write structured data in the log; he hoped it would be used more generally, but that never happened
@TeXnician I guess that the issue I have is that it doesn't keep strictly to the way TeX formats things, so it's harder for me to compare outputs. But that's a pretty unusual case
@JosephWright And it would do what exactly? Would it prevent the previous line to have exactly 80 characters, or any hardwrap completely so that latexmk could stop looking for wrapped lines?
I suppose you know better than I do what all issues are there with log parsing :)
@JosephWright Oh, of course. But if I had to compare the information given in open source > level 1, order 1, name '/opt/luametatex/texmf-context/tex/context/base/mkxl/cont-new.mkxl' to standard TeX output syntax I would be lost in all the parentheses to look for anyway (especially if it's not level 1, order 1) :D
@JosephWright you need to parse all input files in the log to check for file dependencies; that's the reason why you ever care about long lines. So it something helps with this and can be safely used for all documents without breaking anything, I'm all ears and eyes!!!
@TeXnician I guess I'd say the issue in both LaTeX and ConTeXt is there should be fewer files being opened
@TeXnician ConTeXt has the 'issue' that it doesn't attempt to run the source files together; really it could just have two big files, one .tex and one .lua
@JosephWright Actually, I can live with many files being opened (except when on a HDD with slow access times) as long as I am able to recognize at which point the process stopped. Is there some technical background you would argue for fewer files?
@TeXnician File opening is relatively slow, so keeping the number down is important, plus as you notice they don't half clutter the log (plus I tend to want to avoid version-clash, which having everthing in one file helps with)
@yo' You can change all the big ones
@TeXnician I'm going to guess you don't use Windows
@yo' overleaf could tell texlive they want 160 chars in the log, then texlive could change it, and then I wouldn't have the problem that my diffs differ ...
@yo' You could ask on the TL list; other than people doing automated testing, it's probably quite a safe change. Of course, people like @HenriMenke might have a view (as would the team, though a one-time change is quite manageable)
@UlrikeFischer OK, so safe for us, but what about others? As @Skillmon says, I guess one would be wary about editor and script parsing for highlighting issues
@JosephWright The thing is, we in general want as little manual modifications as possible. (Already now we're patching the TeX Live 2020 image the 5th time for minor issues. Anything that's nonstandard requires extra work and is more likely to break something.)
So if this were a thing to put in texmf.cnf or a similar place, that would be great. If it requires a larger change somewhere, we probably better live with people complaining tha BibTeX doesn't work.
@JosephWright well I don't know, imho the short lines are more of a problem, as file pathes are simply longer today. And didn't Karl asked about better log layout sometime ago somewhere anyway?
@Skillmon Lab is not so bad (we have 50% capacity limit, one student here and the other at home connected via teams mainly doing simulation and the microcontroller SW part). But the shield is awkward, and the echo effect in my ears make 2 hours of that really tiring.
@JosephWright I don't really rely on the log format (yet) other than grepping for There were undefined references, etc.
@yo' Just remove all \n from the log, then you have it in a single line and can just regex match files, or you use the -recorder option which gives you a nice list in the .fls file. See also gist.github.com/hmenke/55d9f3ac39fd5845d1aac1d16ef8efdd
@HenriMenke I',m thinking of boxes, mainly: TeX trucates long lines and you have only ...: this comes up with LuaTeX vs the rest as it truncates differently (also pTeX as you have the direction stuff)
@JosephWright If anything TeX needs some sort of stack trace primitive so that I can call \trace or so at any point in the document and TeX tells me what macros were expanded to get there.
@JosephWright Ah, but that's \showoutput, not \tracingall.
@HenriMenke Hmm, perhaps I'm mistaken: I know there are issues with normalising line wrapping, but I can't find the example I had in mind. Perhaps we 'fixed' that in l3build by fiddling with line wrapping anyway
@UlrikeFischer The issue is if the line just befreo in the log has exactly 79/80 lines, as latexmk pre-processes the files to unbreak broken filenames, and I suppose there's no way for the tool to process the file twice.
(And yes, it completely reasonable to say that this lies on latexmk improving their log parser, and I'm completely fine with that answer. However, it doesn't solve the issue in the end, so it's not that much useful...)
@UlrikeFischer Yeah. That's the solution we suggest actually to the users: to add \typeout{} before \bibliography :) (We used to have a dirty latexmkrc hack for this until recently, because we didn't know where the issue is. Then I got mad at this and decided to track the bug down.)
@barbarabeeton ah of course :) I thought of "fozzle" as the combination of "fog" and "drizzle" :p but now I can't stop thinking about Fozzie the bear and that maybe a "fozzle" might be a very bad joke :p
@Plergux -- Yes, it could mean that poor Fozzie is all wet. (The "m" in "mizzle" = "miist"; the rest means the motion is steady and mostly downwards. Certainly pervasive.)
@yo' -- AMSLaTeX now by definition no longer exists as a unit. amsmath is now separated and maintained by the LaTeX team. AMS classes are still allegedly maintained by AMS, although mostly in the breach.
@DavidCarlisle If I am using a class that has changed the math fonts (to newtxmath), is there an easy (\usepackage) way to restore the default CM fonts in math mode?
This question, tex.stackexchange.com/questions/209060/…, sort of addresses it with \mathversion, but it does not restore things like integral signs (which just disappear) (using T1 encoding)
@UlrikeFischer Actually, preventing packages from loading is something I regularly do, especially in the use case of some class providing specific font setups. So it's good to know there's now a command for that.
I would be grateful if anyone could help me with this basic question
I found out that the main cause of the issue in my question https://tex.stackexchange.com/q/573233/2288 is due to the multiple calls of \AtBeginDocument (inside and outside of \ExplSyntax)
So, how can I efficiently call \AtBeginDocument only once and pass to it all the commands I need to execute?
@DavidCarlisle But removing "\string\AtBeginDocument{\string\printanswersfalse}" from inside \ExplSyntax (under the comment % Build LaTeX questions command) makes the header be printed properly adjacent to the page top border as expected.
So, I think one of them is the suspect in one way or another
@Diaa no one should be using shell-escape by default so if you want people to run it you should flag that very clearly (and to be honest post a much simpler example, you are asking people to trust you a lot to post that much code and ask they run t with shell escape)
@Diaa why such weird filenames? spaces in filenames are bad enough but filenames starting with [ !!!
@Diaa actually looking at the code, do you have to use latexmk?
@Diaa it makes doing anything with the files harder and you are asking people to use the files, even just passing the file to a pdf viewer takes more effort.
@Diaa no what I mean is does your example work without. I was trying to work out where \printanswersfalse is used but it isn't used anywhere I think except ``-usepretex="\string\AtBeginDocument{\string\printanswersfalse}` and (I assume??) that usepretex is a latexmk option?
$ grep -i undefined *dd512*log
[Exam] dd512.log:! Undefined control sequence.
[Exam] dd512.log:and I'll forget about whatever was undefined.
[Exam] dd512.log:! Undefined control sequence.
[Exam] dd512.log:and I'll forget about whatever was undefined.
[Exam] dd512.log:! Undefined control sequence.
[Exam] dd512.log:and I'll forget about whatever was undefined.
[Exam] dd512.log:! Undefined control sequence.
[Exam] dd512.log:and I'll forget about whatever was undefined.
The control sequence at the end of the top line of your error message was never \def'ed. If you have misspelled it (e.g., `\hobx'), type `I' and the correct spelling (e.g., `I\hbox'). Otherwise just continue, and I'll forget about whatever was undefined.
! Undefined control sequence. \tcb@startbox ...x }{\tcb@w@upper }\tcb@lua@color {tcbcolupper}\kvtcb@fontup...
Maybe this is because I compile two files: one with the editor and the another externally with shell-escape
inside the \ExplSyntax so that I won't have the error and get the header properly placed in the [Exam] file?
The problem is I get two files produced: one file has the name of the tex file and it doesn't have any issues the second file ([Exam]) is the one I need to pass the two lines I mentioned inside \ExplSyntax to avoid the error I get in its log
@PhelypeOleinik I was just looking at the beginning of expl3.ltx and am wondering if we could simplify the \meaning dance by temporarily changing the catcode of _ (like this) or if that would break in some weird edge-cases?
@DavidCarlisle you mean you didn't see all the nice easter eggs I planted in the code? Btw: I collected the hyperref color suggestions, and there is a \seq_const_from_clist:Nn \c__hyp_colorscheme_david_seq ;-)
@UlrikeFischer well henry authored that one, I just passed it on (and it would make more sense to more people reading it, it's quite a well known scheme under that name)
@DavidCarlisle ok. But it is currently only a draft, probably one should use properties after all, then colors can be added more easily to scheme, in case people would also like black citation or backref links ;-)
