« first day (3681 days earlier)      last day (1250 days later) » 

Paulo Cereda
Paulo Cereda
8:59 AM
> A recent study found that bee populations in Michigan have increased by 50% since 2018.
@AlanMunn ^^ OOH
 
 
2 hours later…
Joseph Wright
Joseph Wright
10:44 AM
@DavidCarlisle I see stats are up
 
David Carlisle
David Carlisle
11:06 AM
@JosephWright oh I'll look
 
Ulrike Fischer
Ulrike Fischer
11:24 AM
@DavidCarlisle how are they?
 
David Carlisle
David Carlisle
@UlrikeFischer I guess you are the resident hyperxmp expert?
the difference is the new acmart.cls has \RequirePackage{hyerxmp} without that the ovefull box goes, but haven't traced exactly whatis happening yet — David Carlisle 6 mins ago
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle I can look, but while I saw that it is quite active currently, I didn't really follow what he is doing (but I will have to in case it clashes ...).
 
David Carlisle
David Carlisle
11:38 AM
@UlrikeFischer you get this box logged if you add \showoutput but ... 
\hbox(9.78558+0.0)x0.0
.\glue(\leftskip) 0.0 plus 1.0fil
.\kern-0.2024 (left margin)
.\hbox(0.0+0.0)x0.0
.\penalty 10000
.\pdfstartlink(*+*)x* attr{/Border[0 0 0]/H/I/C[1 0 0]} action goto name{Hfootn
ote.1}
.\hbox(9.78558+0.0)x4.592
..\mathon
..\hbox(5.4296+0.02199)x4.592, shifted -4.35599
...\T1/LinuxLibertineT-TLF/m/n/8.8 1
..\mathoff
.\pdfendlink
.\penalty 10000
.\kern-0.2024 (right margin)
.\glue(\parfillskip) 0.0
.\glue(\rightskip) 0.0 plus 1.0fil
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle yes found this too. And it seems to have the right size for the overful box.
 
David Carlisle
David Carlisle
@UlrikeFischer there is no difference in the shipped out box so it must be some trial typesetting somewhere?
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle yes must be. There is no link in the pdf.
 
David Carlisle
David Carlisle
11:58 AM
@UlrikeFischer \UseOneTimeHook {enddocument}
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle yes, hyperxmp adds lots of stuff there.
@DavidCarlisle \hyxmp@auto@assign@data seems to be the culprit.
 
David Carlisle
David Carlisle
@UlrikeFischer yes presumably in \hyxmp@parse@acmart
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle my guess is \setbox0=\hbox{\addresses}% in this command
 
David Carlisle
David Carlisle
@UlrikeFischer yes that's the only setbox around, better look whats in addresses I guess
> \addresses=macro:
->\@author {A}\@authornotemark \affiliation {}{\institution {T}\country {R}}\an
d \@author {B}\affiliation {}{\institution {T}\country {R}}.
@UlrikeFischer how bad woul dit be to simply do \setbox0=\hbox{\hfuzz\maxdimen\addresses}%
@UlrikeFischer it never uses the box at all there is an \endgroup immediately after the assignment
 
Ulrike Fischer
Ulrike Fischer
12:14 PM
@DavidCarlisle would probably work. But the whole command looks quite dubious. Why does hyperxmp check specific settings of a class? This looks like a bad interface. The class should do this.
@DavidCarlisle yes I think it only tries to test if some commands have been set.
 
David Carlisle
David Carlisle
@UlrikeFischer well quite especially as the class is actively maintained with updates this year
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle I bet they asked for better metadata support but didn't properly thought it through who should do which settings and checks, and which hooks are needed.
 
 
1 hour later…
Pablo González L
Pablo González L
1:23 PM
@DavidCarlisle Just to leave it in the list of future updates (texfaq.org/FAQ-whatengine)
 
David Carlisle
David Carlisle
1:42 PM
@PabloGonzálezL hmm I suppose you want it all replacing with iftex .... :-)
 
Pablo González L
Pablo González L
1:54 PM
@DavidCarlisle Hehe... Maybe take a look at (texfaq.org/FAQ-osf)
 
David Carlisle
David Carlisle
2:05 PM
@PabloGonzálezL try now
 
Pablo González L
Pablo González L
@DavidCarlisle Great, I know it's almost impossible to update everything related to LaTeX that's on the net, but, the "almost official" reference sites is quite positive, thank you very much.
 
Paulo Cereda
Paulo Cereda
@PabloGonzálezL ooh los textos acerca de TeX
 
David Carlisle
David Carlisle
@PauloCereda ooh los textos sobre cocinar patos
 
Pablo González L
Pablo González L
@PauloCereda Hehehe...I was checking an old 8gb hard drive (a monster in its time) and I came across several LaTeX tutorials and things that are still going around on the net and are totally obsolete :(
 
Paulo Cereda
Paulo Cereda
@PabloGonzálezL ooh let's blame someone
@DavidCarlisle OH NO
 
billyandriam
billyandriam
2:45 PM
Hello all! This is my first time joining the chat. I have one small question, so small I thought it would be better to ask it here rather than to open one on the site.
I've had the same problem as the OP on this question: https://tex.stackexchange.com/questions/31150/bodegraph-on-ubuntu-10-04
It has been solved by @MartinScharrer. I'm on Win10 running TexLive but I was just confused why gnuplot wasn't found as a command so I installed it manually. Is it the correct way or should gnuplot be pre-installed and bundled altogether with TexLive when first installing TexLive? That is, did I mess u…
 
David Carlisle
David Carlisle
@billyandriam gnuplot is not distributed with texlive.
 
billyandriam
billyandriam
Thank you very very much! So it must be manually installed.
 
David Carlisle
David Carlisle
@billyandriam yes for some definition of "manual" Many linux distributions will have it installed anyway or you can instal it via the system package update system etc.
 
Paulo Cereda
Paulo Cereda
@billyandriam welcome to the chat room, by the way!
 
billyandriam
billyandriam
@DavidCarlisle Oh yeah! I'm on Windows anyway so I am nearly a GNU/Linux illiterate.
 
Paulo Cereda
Paulo Cereda
2:50 PM
@DavidCarlisle Does WSL have it?
 
billyandriam
billyandriam
@PauloCereda Thank you :) Please let me know if there is any rule that one should know for first-time users of the chat. Just like the usual welcome to tex.sx rules.
 
David Carlisle
David Carlisle
@PauloCereda it's unbuntu so it would only be an apt-get away if not there, but downloading the windows gnuplot likely to be 1000 times easier than making the wsl one work with windows texlive
 
Paulo Cereda
Paulo Cereda
@billyandriam Don't worry too much, it's cool. :)
@DavidCarlisle full ack. :)
 
David Carlisle
David Carlisle
@billyandriam main rules, are: if @PauloCereda shows up, say you like eating duck. If @CarLaTeX (or any other italian) shows up, say you like putting pineapples on pizza.
8
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle oi
 
Plergux
Plergux
2:54 PM
@DavidCarlisle LOL
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle III. if @DavidCarlisle is around, say longtable is awesome. IV. When an error is reported, blame @UlrikeFischer. :)
2
@Plergux ooh hi sheep person!
 
Plergux
Plergux
@PauloCereda Hi duck person!
 
billyandriam
billyandriam
Ok all! I have to leave for now. Thanks again! Goodbye or "Veloma" in malagasy!
 
Paulo Cereda
Paulo Cereda
Ta-ta
ooh that was very British of me
 
Henri Menke
Henri Menke
3:22 PM
@UlrikeFischer I'm not speaking as the PGF maintainer but as one of the millions of document authors and I find this attitude pretty arrogant. Just shitting over people's documents by silently breaking them and then telling them to just adapt because you have more pull here. So far I have rolled back every regression in PGF because I actually value my users.
 
Skillmon
Skillmon
3:52 PM
What would be a good place to test whether characters are made active, e.g., by babel? \AtBeginDocument could be before babel, depending on the load order.
 
Ulrike Fischer
Ulrike Fischer
@HenriMenke Well we have a rollback system. If you don't like the changes, use it. But we don't have only users who don't want changes, we have also users who are in bad need of better support of various pdf standards and accessibility. And we value these users too. And if may say: we are providing development versions so that everyone can check what is coming and give feedback. You could have tested your documents for months against the new latex.
@Skillmon you can use \AtBeginDocument and setup a rule that your tests are after babel?
 
Skillmon
Skillmon
@UlrikeFischer ohh, forgot about lthooks new ordering rules :)
 
Paulo Cereda
Paulo Cereda
Paulo: Mind if I tell a joke about sodium?
Joseph: Na.
@Skillmon hi mr. rabbit!
 
Skillmon
Skillmon
@PauloCereda Hi, Mr. Duck!
 
Alan Munn
Alan Munn
@PauloCereda This is very good news. Bees are super important to agriculture, especially in this state.
 
Paulo Cereda
Paulo Cereda
4:03 PM
@Skillmon <3
 
Skillmon
Skillmon
@PauloCereda oh, a carrot!
 
Alan Munn
Alan Munn
@PauloCereda Ba
 
Paulo Cereda
Paulo Cereda
@AlanMunn Yes, indeed! My grandfather was a beekeeper, so I wholeheartedly cheer up with such good news!
@AlanMunn LOL
@Skillmon ooh
 
Skillmon
Skillmon
@PauloCereda uh, uh, I know a stupid joke as well: I once told a joke to Helium. It didn't react.
 
Paulo Cereda
Paulo Cereda
@Skillmon LOL
Damn those snob noble gases
 
 
3 hours later…
CarLaTeX
CarLaTeX
7:15 PM
@billyandriam and don't listen to @DavidCarlisle if he's talking about food
 
David Carlisle
David Carlisle
 
Plergux
Plergux
@DavidCarlisle heh, Iceland. :p
 
Skillmon
Skillmon
I just had a bit of panic today... My initramfs image was corrupted, luckily the fallback was working fine, so fixing it was easy enough, but that could've ended with loads of work.
 
Frank Mittelbach
Frank Mittelbach
@HenriMenke Henri can we do a step back a try to discuss this a little less hostile? To accuse use of "shitting over our users" is more than unfair given that we have managed to maintain LaTeX for over 30 year successfully and to the benefit of millions of users.
4
@HenriMenke We are concerned about not breaking documents but any change for the better will break a few documents and even for those cases we work hard on providing solutions, but the solution can't be not to make improvements for the majority if a handful of cases require adjustment.
@HenriMenke I'm going to address your topic but please let us first get back to civilized discussions and not unfounded accusations.
 
Skillmon
Skillmon
7:44 PM
@LaTeXereXeTaL I saw others have already answered, but I think a bit more on the limitations to the package options would be a good idea here. So: Passing anything in that needs spaces, braces, commas, or macros might break. Spaces will be zapped, braces break LaTeX code to remove used global options from that list (and might break elsewhere), and additionally are likely removed during parsing at some point, commas don't work as braces don't work reliably, and macros might get expanded by mistake.
 
Skillmon
Skillmon
8:00 PM
@LaTeXereXeTaL and the "does every package" part of that question: No, but for the more widespread packages (and my own, which aren't wide spread) there are auxiliary packages adding package-option support, e.g., pgfkeys->pgfopts, l3keys->l3keys2e, keyval->xkeyval or kvoptions; and my own stuff: expkv->expkv-opt. And other key=val packages might directly include the package options support, e.g., options (which is pgfkeys inspired), or xkeyval (if one sees it not as [...]
@LaTeXereXeTaL [...] an extension to keyval but as a standalone package inspired by it)
 
LaTeXereXeTaL
LaTeXereXeTaL
8:21 PM
@Skillmon Thanks for the clarifications. I think one detail I wasn't aware of until yesterday was that "options" generally refers to packages, not commands. At least that's what I got from my discussion with @DavidCarlisle. And "keys" generally refers to commands, not packages, correct? So pgfkeys and pgfopts are a similar pair, as are l3keys and l3keys2e, and as are keyval and either xkeyval or kvoptions? I already have some small demos with kvoptions working for package options.
Honestly, I think a lot of my conceptual difficulties are just getting the terminology correct.
 
Skillmon
Skillmon
@LaTeXereXeTaL I'd not generalise that statement about the two terms. All that pgfopts does is call pgfkeys on the individual entries of the package or class option lists (or both). l3keys2e and expkv-opts do exactly that as well.
 
LaTeXereXeTaL
LaTeXereXeTaL
@Skillmon I'm also reading past TUGboat articles by @JosephWright.
@Skillmon I'm going to experiment more today to see if I can get the hang of it.
 
Skillmon
Skillmon
@LaTeXereXeTaL don't remind me of those... We wanted to write a follow up article of that in spring... :)
 
LaTeXereXeTaL
LaTeXereXeTaL
@Skillmon Are they outdated or something like that?
 
Skillmon
Skillmon
@LaTeXereXeTaL "options" is often used for package or class options, but "key" is just a named entry in some list. A "key" isn't necessarily for a command, but also for package options (if you use packages like kvoptions, pgfopts, or similar).
 
LaTeXereXeTaL
LaTeXereXeTaL
8:28 PM
@Skillmon Ah okay then.
 
Skillmon
Skillmon
@LaTeXereXeTaL no, not necessarily. But there are just new things to consider, and if the team manages to give better support for package and class options (without this \zap@space code), we have a totally new situation. So maybe it's best to not write a new TUGboat article until things are sorted out in this regard.
 
LaTeXereXeTaL
LaTeXereXeTaL
@Skillmon I've seen that \zap@space thing mentioned recently. While I don't know the details, I gather that it breaks some things.
 
Skillmon
Skillmon
@LaTeXereXeTaL also I did some (read: much) benchmarking on the different key=value interfaces and checked some of them for bugs or misfeatures. The essence of this benchmarking is contained in the documentation of expkv.
 
David Carlisle
David Carlisle
@Skillmon stop mentioning \zap@space :(
 
LaTeXereXeTaL
LaTeXereXeTaL
Oh dear. I'm just a bystander. Didn't mean to start something. LOL
 
Skillmon
Skillmon
8:32 PM
@LaTeXereXeTaL well, it works in the cases it was designed for, is simple in nature, and doesn't have too bad performance (actually quite good performance). It just breaks more "modern" use-cases one might like to have.
@LaTeXereXeTaL don't worry. Everyone here knows that @DavidCarlisle's code doesn't have bugs, so there can't be any discussion you started :)
 
David Carlisle
David Carlisle
@LaTeXereXeTaL when we designed \usepackage and its option system for latex2e the main idea was to fix this which did not work in latex2.09@ 
\documentclass[
fleqn,
twocolumn
]
{article}
 
Skillmon
Skillmon
@DavidCarlisle of course it didn't, there is no articlle.cls :)
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle Because the options are broken over multiple lines?
 
David Carlisle
David Carlisle
so we applied a space removing pre-pass first but as the only options we had considered were simple names like twocolumn or 12pt the code to remove spaces is somewhat aggressive and pre-expands (destroys) any tex code added there.
 
Skillmon
Skillmon
@LaTeXereXeTaL because each of those line endings translates into a space.
 
David Carlisle
David Carlisle
8:37 PM
@LaTeXereXeTaL yes, or even on one line if you did [ fleqn , twocolumn ] it would fail.
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle extraneous spaces
 
David Carlisle
David Carlisle
At the time we had to make it fit in a 640K PC and there were no packages existing that had any options so it was hard to know how it would be used.
 
LaTeXereXeTaL
LaTeXereXeTaL
Aha so that means keys made of multiple words won't work because the space is zapped?
 
Skillmon
Skillmon
@LaTeXereXeTaL yes, also the space zapper might accidentally strip braces due to TeX's parsing rules for delimited parameters.
 
David Carlisle
David Carlisle
@LaTeXereXeTaL yes and worse values with tex macros in get destroyed, imagine something like font=\bfseries `\bfseries will expand and give weird internal errors.
 
LaTeXereXeTaL
LaTeXereXeTaL
8:42 PM
@DavidCarlisle I remember that example from yesterday, yes.
 
Skillmon
Skillmon
(but braces give bad results in package or class options anyways...)
 
David Carlisle
David Carlisle
@LaTeXereXeTaL it could be easily changed except for the fact that there are 1100002346 local thesis classed out there that patch \documentclass in unknown weird ways and whenever we have tried ti change anything something breaks and we back it out... We do have a plan for a more conservative safer change but it's not easy
@LaTeXereXeTaL doesn't have to be a font command, more or less anything, eg name=david likely to fail if your name isn't david and has non ascii letters. Really only single word ascii options were in the design.
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle I would think it's up to the authors to accommodate the changes, but I understand weighing that against backwards compatibility.
 
Skillmon
Skillmon
@DavidCarlisle that's the reason why LaTeX3 would've been a good idea. Getting rid of all those historic baggage.
 
Phelype Oleinik
Phelype Oleinik
@DavidCarlisle Blame the user for not being named David?
 
Skillmon
Skillmon
8:45 PM
@PhelypeOleinik Phelype seems fine, too.
 
Phelype Oleinik
Phelype Oleinik
@Skillmon Oh no, too complicated
 
David Carlisle
David Carlisle
@LaTeXereXeTaL scroll up a bit and see discussion about an (unrelated) change. If you are writing your thesis, changing it isn't so bad. If you are maintaining arxiv and have several million tex documents that you need to be able to process, changing them after a system update is less fun.
@Skillmon I guess you didn't try the latex3 format of the time. it was a little bit slow.
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle I can imagine.
 
Skillmon
Skillmon
@PhelypeOleinik As long as you're not named Ṕĥéḷỳ˞pē
 
LaTeXereXeTaL
LaTeXereXeTaL
Well, I plan to use single word ascii options so I'm hoping to avoid all this. LOL
 
Skillmon
Skillmon
8:50 PM
@DavidCarlisle I was <3 years old when Frank typeset the LaTeX3 manual... So no, I didn't.
 
David Carlisle
David Carlisle
@Skillmon your parents parenting skills were clearly lacking
 
LaTeXereXeTaL
LaTeXereXeTaL
A key=value interface seems to be the preferred way around providing so many arguments to commands, which is why I want to learn to implement it.
Trying to move with the times.
 
Skillmon
Skillmon
@LaTeXereXeTaL then you're fine. Also, there are use cases where you have to use a package option.
 
LaTeXereXeTaL
LaTeXereXeTaL
@Skillmon For packages I'm using a \hypersetup interface.
 
Skillmon
Skillmon
@LaTeXereXeTaL That's a good approach. But some packages I wrote do use a dual approach. Things which have to be known during package load time (because macros are defined differently for some option combinations, for example) are simple non-value taking options, everything else is managed through a key=value interface with a \...setup macro.
 
LaTeXereXeTaL
LaTeXereXeTaL
8:54 PM
Since July I've had a huge awakening: unicode, LuaLaTeX, key=value, font encoding (still a bit shaky on that), expansion,...the list goes on.
 
David Carlisle
David Carlisle
@LaTeXereXeTaL package options work best when they affect the code at load time eg \documentclass[12pt]{article} causes different code to be loaded than \documentclass[10pt]{article} so that works better being an option on the load rather than a later setup. but if it's just setting some choices, and in particular if you may want different choices in different parts of teh document, having a separate settings commadn works better/
 
Skillmon
Skillmon
@DavidCarlisle I'll forward that to my mum, can I also give her your email-address, she might like to have a word with you afterwards :P
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle I'm thinking booleans that act as a switch.
 
David Carlisle
David Carlisle
@LaTeXereXeTaL sure it's a matter though of when you read them eg amsmath has a fleqn option to set things flush left instead of centered. the option itself just sets a boolean flag so the code knows which form to use but (mostly) that flag is read at package load time so if you use fleqn the default centred definitions are simply not there. that's a pain as people sometimes ask to do just one equation in the other style and it i not easy.
 
Phelype Oleinik
Phelype Oleinik
@Skillmon No need to zalgo it. My name is... weird... for Brazilian standards, so people usually get it wrong. Really, I've seen that y be put in all possible positions :P
 
David Carlisle
David Carlisle
8:59 PM
You could now define it so both definitions loaded always and the boolean tested as equation is used, but there was not space to do that in machines of the time.
 
Phelype Oleinik
Phelype Oleinik
@Skillmon But I guess in English (or for people familliar with English) it's okay
 
Skillmon
Skillmon
@PhelypeOleinik Well, I never would've guessed you're Brazilian when I first read your name, but otherwise, there are much weirder names :)
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle Understood. I need to experiment more.
 
Phelype Oleinik
Phelype Oleinik
@Skillmon I shouldn't have let people know ;-)
 
Skillmon
Skillmon
@PhelypeOleinik I did input that directly with my keyboard, no generator involved.
 
Phelype Oleinik
Phelype Oleinik
9:01 PM
@Skillmon Oh, wow. I don't know my keyboard that well
 
LaTeXereXeTaL
LaTeXereXeTaL
Now, I need to do some yardwork before the sun sets!
I shall return later. Thank you all!
 
David Carlisle
David Carlisle
@LaTeXereXeTaL it's set already
 
Skillmon
Skillmon
@PhelypeOleinik neo-layout.org
@DavidCarlisle since hours!
 
LaTeXereXeTaL
LaTeXereXeTaL
@DavidCarlisle You're five hours ahead of me.
And about twenty degrees higher in latitude.
 
Skillmon
Skillmon
@LaTeXereXeTaL uh, Timelords!
 
David Carlisle
David Carlisle
9:03 PM
@LaTeXereXeTaL GMT is obviously the correct timezone, i like to think of it as you being out by 5 hours.
 
LaTeXereXeTaL
LaTeXereXeTaL
@Skillmon Astronomers!
@DavidCarlisle As an astronomer, I completely agree.
 
Skillmon
Skillmon
@DavidCarlisle well, that's debatable.
 
Phelype Oleinik
Phelype Oleinik
@Skillmon Oh, cool! But how long did you take (assuming you did) to adapt from qwerty to that?
 
David Carlisle
David Carlisle
@PhelypeOleinik He's German so had qwertz not qwerty so was probably desperate to find something else.
 
Skillmon
Skillmon
@PhelypeOleinik depends, on normal text I was confidently typing after two weeks. Programming symbols took a bit longer. Not sure how long it took to reach my former speed (but now I'm faster, especially while programming).
@DavidCarlisle whyile qwertz (thanks to the symbols) is a bit worse than qwerty, both are horrible for typing.
 
Phelype Oleinik
Phelype Oleinik
9:12 PM
@Skillmon Oh, that's surprisingly low! I'd take weeks to adapt...
 
Skillmon
Skillmon
@PhelypeOleinik two weeks is "weeks" :)
@PhelypeOleinik and I typed more than usual during those two weeks to train.
 
Phelype Oleinik
Phelype Oleinik
@Skillmon But I said weeks (the emphasis is important :)
 
Skillmon
Skillmon
@PhelypeOleinik ^^^
 
Phelype Oleinik
Phelype Oleinik
@Skillmon Keyboard invaders :)
@Skillmon I died right after a blew up a marriage
 
Skillmon
Skillmon
9:29 PM
@PhelypeOleinik you can shoot a wave by hitting enter which kills everything near you, if you're cornered.
 

« first day (3681 days earlier)      last day (1250 days later) »