« first day (4728 days earlier)      last day (210 days later) » 

cfr
cfr
3:04 AM
@PauloCereda Glad I waited. expl3 update breaks my code :(.
 
 
5 hours later…
mickep
mickep
8:23 AM
@JosephWright Sounds like a good change! I liked the "A little while ago now" part. Wasn't it a rather long time ago that \expanded was added?
 
Joseph Wright
Joseph Wright
@mickep Not in LaTeX terms, no
@mickep Remember we need to think really at least on a 5 year timescale
 
mickep
mickep
@JosephWright I would have guessed at least 10 years, but maybe that is not correct then.
 
Joseph Wright
Joseph Wright
@mickep In LuaTeX, from the word go, but in the others on since 2019 - ftp.math.utah.edu/pub/tex/historic/systems/texlive/2019/web/…
So pretty recent
 
mickep
mickep
@JosephWright Ah, so recent for the others, OK.
Then it indeed makes more sense.
 
Joseph Wright
Joseph Wright
@mickep For LaTeX, requiring a primitive that's only been available for 4 years in pdfTeX is 'really recent' :)
 
Skillmon
Skillmon
8:36 AM
@JosephWright there is no 2e-interface to .value_required:n, is there?
 
David Carlisle
David Carlisle
Oct 4, 2016 at 22:18, by David Carlisle
@Canageek 1st rule of tex support: never believe users when they describe what happened.
 
Paulo Cereda
Paulo Cereda
8:57 AM
@cfr oh no
 
mickep
mickep
9:09 AM
@PauloCereda And not possible to blame arara...
 
Paulo Cereda
Paulo Cereda
@mickep success!
ooh for a minute I thought this was in Germany
2
 
mickep
mickep
@PauloCereda Haha, wonderful.
 
Joseph Wright
Joseph Wright
@Skillmon At present no, but these are just 'friendly' names anyway - you can use the expl3 one without issue
@DavidCarlisle ??
 
Skillmon
Skillmon
@JosephWright yes, the second part I knew, the first I wasn't sure about :)
 
David Carlisle
David Carlisle
@JosephWright expl3 ~ change
 
Joseph Wright
Joseph Wright
9:12 AM
@DavidCarlisle Ah, right
@DavidCarlisle Well yes, we haven't changed anything
 
daleif
daleif
@UlrikeFischer hyperref with the pagebackref option. Should that only work if it is specified as a package optionn and not \hypersetup? We have a doc where if set in hyperetup it complains that the setting is already set and does not work. Adding it as a package option it works and no complains.
 
daleif
daleif
9:26 AM
@UlrikeFischer forget it, our bad. It was not super clear from the manual that those options listed should only be used as package options.
But there might be something wrong with the \backrefalt example in section 7.27
 
Ulrike Fischer
Ulrike Fischer
@daleif what?
 
samcarter
samcarter
9:45 AM
What colour does the new Jag have? :)
 
Skillmon
Skillmon
@samcarter dark green
 
Joseph Wright
Joseph Wright
 
samcarter
samcarter
@Skillmon oh, that sounds fast - I like it :)
@JosephWright :)
 
Paulo Cereda
Paulo Cereda
@JosephWright ooh Lord Percy Percy
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle well the good news is, that filecontents actually writes with unix line endings ;-). I got a different length from a manually created support file ...
 
Joseph Wright
Joseph Wright
10:08 AM
@PauloCereda :)
 
daleif
daleif
@UlrikeFischer We tried using \backrefalt in a thesis and it kept just having the value 0 in #1 and #3 (where we expected the number of back refs). I can see if I can make an example later. (we dumped the four args to log and #1 and #3 kept just beign 0)
Hmm, seems to be interaction from something else. Will dig
 
David Carlisle
David Carlisle
10:26 AM
@UlrikeFischer ah so blame @UlrikeFischer
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle ;-). Now I'm down to the dvips tests failing on github. And I don't see how they should ever work, so how can I disable them fully on github? @JosephWright?
 
David Carlisle
David Carlisle
@UlrikeFischer disable just on github, or everywhere?
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle only on github (or on all non-windows-OP). I want to run the tests locally.
 
daleif
daleif
@DavidCarlisle It seems to be an interaction between chapterbib, natbib and hyperref
Here is a MWE:
\documentclass[a4paper]{book}

% remove one of these and \backrefalt works
\usepackage[sectionbib]{chapterbib}
\usepackage{natbib}

\usepackage[pagebackref]{hyperref}
\hypersetup{
colorlinks,
linktoc = all,
final,
}

\renewcommand*{\backrefalt}[4]{%
% alternative interface
% #1: number of distinct back references
% #2: backref list with distinct entries
% #3: number of back references including duplicates
% #4: backref list including duplicates
\par
#3 citation(s) on #1 page(s): #2,\par
\ifnum#1=1 %
 
David Carlisle
David Carlisle
why is the dvips output windows specific?
@daleif thanks will look but possibly later
 
Ulrike Fischer
Ulrike Fischer
10:38 AM
@DavidCarlisle that are pdf-tests, and the pdf-version is different, font descriptors are different, and other small stuff, that I basically can't control. You can see the diffs here github.com/latex3/pdfresources/suites/17109067091/artifacts/…
 
David Carlisle
David Carlisle
@UlrikeFischer seems wrong to test for github (although we could test for windows) couldn't we diff the postscript rather than the pdf?
 
Ulrike Fischer
Ulrike Fischer
11:07 AM
@DavidCarlisle well perhaps (but it wouldn't be the same, as one also wants to see what ghostscript is doing with the pdfmarks ...), but on the whole I'm not sure if it worth the trouble, the dvips support is squetchy anyway. Running only locally the tests now and then should be ok, I only don't want failures on github.
 
David Carlisle
David Carlisle
@UlrikeFischer some of us may want to test as well:-) but I think simplest is put them all in a separate config and then set checkconfigs conditionally to have or not have that config depending if os.typeis windows
 
Ulrike Fischer
Ulrike Fischer
11:57 AM
@DavidCarlisle do the dvips tests break also on windows-in-denial?
 
David Carlisle
David Carlisle
@UlrikeFischer yes as it's basically a linux tex
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle could you resale them ĺin another folder or another branch)? Then one could test if they run 🏃‍♂️ ( <- suggestion from my mobile) on github)
 
David Carlisle
David Carlisle
12:44 PM
@UlrikeFischer yes but if os.type=="windows" add config-dvips to testconfig seems simpler
 
Ulrike Fischer
Ulrike Fischer
1:01 PM
@DavidCarlisle yes, to suppress the tests, but you said some of you may want to test as well, so I wanted to provide a version for you ;-)
@daleif try this (I'm not fully sure that it will work always, all this extra-info commands are bit difficult to track, with biblatex it would be easier ...): 
\makeatletter
\def\BR@backref#1{%
  \newblock
  \begingroup
    \csname @safe@activestrue\endcsname
    \expandafter\providecommand\csname brc@#1\@extra@b@citeb\endcsname{0}%
    \expandafter\providecommand\csname brcd@#1\@extra@b@citeb\endcsname{0}%
    \csname @safe@activesfalse\expandafter\endcsname
    \ifBR@BackrefAlt
      \ifx\backrefentrycount\BR@BackrefEntryCountUnused
      \else
        \BR@PopulateEntryCount{#1}%
      \fi
      \expandafter\backrefalt\csname brc@#1\@extra@b@citeb\expandafter\endcsname
 
David Carlisle
David Carlisle
@UlrikeFischer if needed we could look at doing more agressive normalisation (eg removing random identifiers from font subset names) but not today
 
cfr
cfr
@DavidCarlisle In this case, especially. Nonetheless, the update broke my code. I just shouldn't have tried to remove the Welsh from my example.
 
David Carlisle
David Carlisle
@cfr Dylech chi feio @UlrikeFischer bob amser
 
cfr
cfr
@PauloCereda And, being an idiot, I screwed up my question.
@DavidCarlisle Am y Gymraeg? Neu yn gyffredinnol?
 
David Carlisle
David Carlisle
1:17 PM
@cfr Bob amser, am unrhyw beth.
 
David Carlisle
David Carlisle
1:43 PM
@cfr has that ever worked? don't you need \string~ to stop ~ outside of expl3 meaning a non breaking space?
 
daleif
daleif
@UlrikeFischer thanks, passed on to the student with the issue. Your code also explains where those macros were used, which was not obvious from the manual
 
cfr
cfr
2:03 PM
@DavidCarlisle Yes. It worked until the day before yesterday. Or, rather, my real version worked until then. I hadn't constructed a MNWE at that point. However, my MNWE works fine with TL 2022. Also, \input{~/...} still works regardless of which version of TL I use. I've edited my question in response to @UlrikeFischer 's comments, so hopefully it's a bit clearer now?
 
David Carlisle
David Carlisle
actually you are right I get (with 2023) `! LaTeX Error: File `~/Dogfennau/Fy_Ysgrifau/dysgu/ymresymiadau/ysgrifennu-LEAR
N-compact/hello.tex' not found.` which means we stringified the `~` and passed it to kpathsea
in texlive 2020 you would have had `! LaTeX Error: File `protectunhboxvoidb@xprotectpenalty@M {}/Dogfennau/Fy_Ysgri
fau/dysgu/ymresymiadau/ysgrifennu-LEARN-compact/hello.tex' not found.`
@cfr but the change to make ~ work there was in 2021 (so probably a side effect of making utf-8 filenames work) I don't see a change this week
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle it can't be this week. I get the error in miktex where I have 2023-08-29.
 
cfr
cfr
@UlrikeFischer I can't tell you exactly when TL 2023 changed because I hadn't updated it since installing. It changed for me because I updated TL 2023. So, the version I had last week worked, but that version was presumably older than the version then current.
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle I guess it is this commit github.com/latex3/latex3/commit/…
 
cfr
cfr
@DavidCarlisle TL 2023 available on 17th June worked. That is probably the version I had installed last week. I had no reason to update really, so I didn't until an OS update broke arara. Then I updated yesterday and it didn't work.
@UlrikeFischer I take it a 15th June commit wouldn't have been in TeX Live by 17th, so that seems entirely plausible timewise.
 
David Carlisle
David Carlisle
2:21 PM
@UlrikeFischer @JosephWright ^^
@cfr hmm seems like we are both right, it doesn't work now, and it wouldn't have worked for the 30 years prior to 2021, but it did work last year...
 
cfr
cfr
@DavidCarlisle I don't think that's right. But let me check ...
 
David Carlisle
David Carlisle
@cfr the simple \input{~/foo} case not with the expl3 wrapping
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle the expl3-version works for me also in older texlive.
 
cfr
cfr
@DavidCarlisle My example compiles fine with TL 2021, TL 2020 and TL 2022. It fails with TL 2019 because I'm not loading expl3 and \ExplSyntaxOn is undefined.
@DavidCarlisle No, both.
 
David Carlisle
David Carlisle
@cfr I get LaTeX Error: File `protectunhboxvoidb@xprotectpenalty@M {}/D on your example in texlive 202, 2021
@cfr because ~ is \unhbox\voidbox\penalty\@M
 
cfr
cfr
2:29 PM
@DavidCarlisle I don't get that error with TL 2021.
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle why is ~ not so defined that it is safe in a csname?
 
cfr
cfr
@DavidCarlisle If I add \usepackage{expl3}, \myinput works with TL 2019, but \input fails there.
 
David Carlisle
David Carlisle
@cfr sorry 2021 is the first that works, 2020 gives what I showed
@UlrikeFischer because it was defined in 1982
 
cfr
cfr
@DavidCarlisle I can't get that error in any version of TL I have. Neither 2020 nor 2019.
 
David Carlisle
David Carlisle
$ /usr/local/texlive/2020/bin/x86_64-cygwin/pdflatex ee212
This is pdfTeX, Version 3.14159265-2.6-1.40.21 (TeX Live 2020) (preloaded format=pdflatex)
 restricted \write18 enabled.
entering extended mode
(./ee212.tex
LaTeX2e <2020-10-01> patch level 4
L3 programming layer <2021-02-18>
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/article.cls
Document Class: article 2020/04/10 v1.4m Standard LaTeX document class
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/size10.clo))
(/usr/local/texlive/2020/texmf-dist/tex/latex/l3backend/l3backend-pdftex.def)
@cfr ^
 
cfr
cfr
2:32 PM
This is pdfTeX, Version 3.14159265-2.6-1.40.21 (TeX Live 2020) (preloaded format=pdflatex 2021.1.13)  11 OCT 2023 15:31
entering extended mode
 restricted \write18 enabled.
 %&-line parsing enabled.
**prawf
(./prawf.tex
LaTeX2e <2020-10-01> patch level 4
L3 programming layer <2021-01-09> xparse <2020-03-03>
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/article.cls
Document Class: article 2020/04/10 v1.4m Standard LaTeX document class
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/size10.clo
 
David Carlisle
David Carlisle
I used 
\documentclass{article}
\ExplSyntaxOn
\cs_new_protected_nopar:Nn \my_input_add:n
{
  \seq_put_right:Nn \l_file_search_path_seq { #1 }
}
\cs_new_protected_nopar:Nn \my_file_input: { \file_input:n {ysgrifennu-LEARN-compact/hello} }
\cs_new_eq:NN \myinput \my_file_input:
\cs_new_eq:NN \myinputadd \my_input_add:n
\ExplSyntaxOff
\myinputadd{~/Dogfennau/Fy_Ysgrifau/dysgu/ymresymiadau}

\begin{document}

\input{~/Dogfennau/Fy_Ysgrifau/dysgu/ymresymiadau/ysgrifennu-LEARN-compact/hello}
\myinput
\end{document}
 
cfr
cfr
I used 
\documentclass{article}
% \usepackage{expl3}
\ExplSyntaxOn
\cs_new_protected_nopar:Nn \my_input_add:n
{
  \seq_put_right:Nn \l_file_search_path_seq { #1 }
}
\cs_new_protected_nopar:Nn \my_file_input: { \file_input:n {ysgrifennu-LEARN-compact/hello} }
\cs_new_eq:NN \myinput \my_file_input:
\cs_new_eq:NN \myinputadd \my_input_add:n
\ExplSyntaxOff
\myinputadd{~/Dogfennau/Fy_Ysgrifau/dysgu/ymresymiadau}

\begin{document}
% \input{~/Dogfennau/Fy_Ysgrifau/dysgu/ymresymiadau/ysgrifennu-LEARN-compact/hello}
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle so \protected\def~{\ifincsname\string~\else\nobreakspace\fi} ?
 
cfr
cfr
:64558350 This is pdfTeX, Version 3.14159265-2.6-1.40.21 (TeX Live 2020) (preloaded format=pdflatex 2021.1.13)  11 OCT 2023 15:33
entering extended mode
 restricted \write18 enabled.
 %&-line parsing enabled.
**prawf
(./prawf.tex
LaTeX2e <2020-10-01> patch level 4
L3 programming layer <2021-01-09> xparse <2020-03-03>
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/article.cls
Document Class: article 2020/04/10 v1.4m Standard LaTeX document class
(/usr/local/texlive/2020/texmf-dist/tex/latex/base/size10.clo
@DavidCarlisle ^^ and copy-pasted code from your message. (Because I'd altered my code for 2019.)
 
David Carlisle
David Carlisle
@UlrikeFischer something like that, yes
@cfr I'm confused but out of time, will look later
 
cfr
cfr
2:36 PM
@DavidCarlisle Thanks. My TL 2020 was updated, so it is not the initial release. But I can't reproduce your error even with TL 2019.
 
David Carlisle
David Carlisle
@cfr linux?
 
cfr
cfr
@DavidCarlisle Linux, yes.
 
The_AH
The_AH
How do I start learning LaTeX?
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle I think it is differences between \input and \file_input:n.
 
Paulo Cereda
Paulo Cereda
@The_AH hey! this seems like a good start: learnlatex.org
@UlrikeFischer ooh
Germans are fast
 
The_AH
The_AH
2:40 PM
Thanks.
 
cfr
cfr
@The_AH You can use one of the online sites with a free account to try it out without installing locally. I used overleaf in classes for beginners (because the computer lab machines don't have TeX Live).
 
The_AH
The_AH
Hmm.
I am fine with locally installing though.
 
cfr
cfr
@UlrikeFischer If I go back to TL 2019, \input fails but \file_input:n works.
@UlrikeFischer But I still don't get the error @DavidCarlisle showed.
 
Paulo Cereda
Paulo Cereda
user image
2
It's a pity that Gert definitely does not see this chatroom ^^ @UlrikeFischer (ooh a plan)
 
cfr
cfr
@The_AH Fair enough. I just mention it because it's useful, especially, for people who aren't sure if it is for them. Also Overleaf does some things for you, so some of the complexity is hidden. This can be good or bad. Good in that it makes starting easier. Bad if it means you expect the same when you install locally!
 
samcarter
samcarter
2:46 PM
@PauloCereda Let's sell a vampire chess to avoid this "problem" :)
 
Ulrike Fischer
Ulrike Fischer
@cfr I wouldn't spent time on tracking down the changes. For now use \string~, and we can think about how to correct this.
 
Paulo Cereda
Paulo Cereda
@samcarter OOH
 
cfr
cfr
@UlrikeFischer Thanks. I can't go back further than TL 2019 without a lot of work anyway. I was just curious why I got different results from @DavidCarlisle.
 
gusbrs
gusbrs
@DavidCarlisle There's a chance you might like this: tex.stackexchange.com/questions/697954/…
@daleif too, if I recall. ^^^
 
Paulo Cereda
Paulo Cereda
@samcarter see Matrix
 
David Carlisle
David Carlisle
3:03 PM
@gusbrs thanks
 
Joseph Wright
Joseph Wright
@UlrikeFischer, @DavidCarlisle, @cfr Ah, I think the real issue is the definition of ~: it doesn't account for csname contexts ...
And it's not \protected
 
cfr
cfr
@JosephWright Thanks. That sounds like @UlrikeFischer's comment above.
 
daleif
daleif
@gusbrs to be honnest, I'd rather have a doctex mode where the coloring does not bog down emacs on long dtx files. It has been better on newer emacs versions, but still bad.
 
gusbrs
gusbrs
@daleif That's fair, of course. Lucky me I don't need to deal with something as large as memoir.dtx on a regular basis. :-)
 
Joseph Wright
Joseph Wright
3:18 PM
@cfr Indeed: I guess I'll see if the team are happy with a change
@cfr It's well known I think that I'd make a lot of stuff \protected
 
cfr
cfr
@JosephWright I'm unblissful ignorant. (Until now.)
unblissful -> unblissfully
 
David Carlisle
David Carlisle
@daleif secret preferred mode: M-x fundamental-mode
 
cfr
cfr
@JosephWright (& @UlrikeFischer @egreg et al.) And sorry about my first question. I thought it would be clearer if the paths weren't in Welsh, but almost all the actual paths in my home directory are Welsh. (Dogfennau is what Documents becomes if you use Welsh on Linux.) Clearly, I should have left it alone so it was properly tested. (But I could still screw it up.)
 
The_AH
The_AH
Can someone recommend a TeX Distribution for MacOS?
 
cfr
cfr
@The_AH MacTeX
 
Joseph Wright
Joseph Wright
3:27 PM
@cfr It's OK
 
The_AH
The_AH
@cfr Sounds like made up in 2 seconds but okay.
 
cfr
cfr
@The_AH MacTeX is TeX Live for Mac OS. I think MikTeX is now available for Mac, but it's still mostly used on Windows. MacTeX is basically TeX Live plus some Mac GUI goodies to make life easier and a bundled editor, if I remember correctly. When I used it, it was really very nice. Much easier than what had been available before then. If it worked on Linux, I'd still use it!
@The_AH To prove I didn't make it up: tug.org/mactex.
 
The_AH
The_AH
I know. I KNOW. I KNOOOOOOOOOW.
 
cfr
cfr
@JosephWright Not really. I don't like wasting people's time. Or looking like an idiot.
(Being an idiot is bad enough.)
@The_AH ???
 
The_AH
The_AH
@cfr I knew. I KNEW. I KNEEEEEEEEEEEEEW.
 
samcarter
samcarter
3:34 PM
@The_AH I second what @cfr said. For a beginner, mactex is probably a good choice. Once you are more familiar with things, you can decide if you actually use all the extra stuff and switch to normal texlive if you don't.
 
cfr
cfr
@The_AH Then why did you ask? ASK? AAAAASSSKKK?
 
The_AH
The_AH
@samcarter I now understand. I NOW UNDERSTAND . I NOOOOOOW UNDERSTANNNNND.
@cfr I meant I knew MacTeX wasn't made up! I MEANT I KNEW MACTEX WASN'T MADE UP! I MEANT I KNEEEEEEW MACTEEEEEEX WASN'''''T MAAAADE UUUP!
 
cfr
cfr
@samcarter The utilities for managing TeX Live are useful. Or is it no longer as easy to switch from one version to another? On Linux, I basically emulate the way MacTeX did (does?) that. The one thing I don't like is that MacTeX installed TeX Live as root, which is something I much prefer to avoid.
But Apple have messed that stuff up, as I understand it, so I've no idea what MacTeX does now.
 
Joseph Wright
Joseph Wright
@cfr It still works, although the author has been ... less than pleased with Apple
 
samcarter
samcarter
@cfr I haven't use mactex for some time, so I'm not up-to-date on the current state. Personally I didn't see a need for the extra stuff, so I switched to vanilla texlive. I have a zsh function to switch between versions.
 
Joseph Wright
Joseph Wright
3:45 PM
@samcarter Does that work for GUIs?
 
samcarter
samcarter
@JosephWright no, it simply sets the path for the current session.
 
Joseph Wright
Joseph Wright
@samcarter Hmm, so MacTeX (or rather the TeX Live Utility) still needed for that - I go with MacTeX then remove most of the extras :)
Then again, I also use my Windows system to switch TL versions as running older ones on the Mac is .. interesting
 
samcarter
samcarter
@JosephWright for the pre 2019 ones?
 
Joseph Wright
Joseph Wright
@samcarter Yup
 
samcarter
samcarter
@JosephWright Ah, the sacrifice of older programs to the M* gods :)
 
Alan Munn
Alan Munn
3:53 PM
@samcarter @The_AH MacTeX is TeXLive for the Mac. There is no sense not to use it.
 
samcarter
samcarter
@AlanMunn Storage space
 
Alan Munn
Alan Munn
@samcarter Well if you're concerned about storage space, then install BasicTeX and go from there.
 
mickep
mickep
1h to monthly ConTeXt meeting. 😌
 
cfr
cfr
@JosephWright You don't really need MacTeX for that. At least, you didn't. I do the same on Linux. Works for GUIs etc. and isn't session-dependant.
Partial listing of /usr/local/texlive: 
drwxr-xr-x   9 texlive texlive 4.0K Hyd   4  2019 2019/
drwxr-xr-x   9 texlive texlive 4.0K Ion  13  2021 2020/
drwxr-xr-x   9 texlive texlive 4.0K Maw   9  2022 2021/
drwxr-xr-x   9 texlive texlive 4.0K Tach 12  2022 2022/
drwxr-xr-x   9 texlive texlive 4.0K Hyd  10 20:39 2023/
lrwxrwxrwx   1 texlive texlive    8 Meh  17 17:45 bin -> bin.2023/
lrwxrwxrwx   1 texlive texlive   21 Mai  16  2019 bin.2019 -> 2019/bin/x86_64-linux/
lrwxrwxrwx   1 texlive texlive   21 Tach  7  2020 bin.2020 -> 2020/bin/x86_64-linux/
^^ is essentially what MacTeX does (or used to do). When you change versions, it just adapts the symbolic links, but there's only two to do. Then you don't need to include YYYY in PATH. 
if [ $UID != 0 ]
then
        [ -d "/usr/local/texlive/bin" ] && export PATH="/usr/local/texlive/bin:${PATH}"
# need to add to INFOPATH but this seems not to be set and I'm not sure how to handle that...
        [ -d "/usr/local/texlive/current/texmf-dist/doc/info" ] && export INFOPATH="${INFOPATH}:/usr/local/texlive/current/texmf-dist/doc/info"
fi
TeX Live's installer still tells me to set MANPATH, even though this has been a bad idea for years.
I know this is probably wrong for current Mac OS because Apple, but it probably still uses something similar, even if the locations have changed.
(Also, setting PATH is probably more complicated on Mac. On my system, I just drop the snippet into /etc/profile.d and make it executable. Then all users get the correct PATH. (root is excluded by design.)
 
Joseph Wright
Joseph Wright
@cfr macOS GUIs don't read the PATH, etc., so whilst you can adapt symblinks, you can't do (to me) the natural thing and simply change the PATH (which is what I do on Windows)
@cfr Anyway, the main reason for installing MacTeX for me is so I can match end-user setups - that's important when asked odd questions (e.g. I have TeXshop installed but don't use it myself)
 
Paulo Cereda
Paulo Cereda
4:09 PM
@JosephWright you could definitely get rid of your MacBook: chat.stackexchange.com/transcript/message/64552661#64552661 :)
 
Joseph Wright
Joseph Wright
@samcarter Indeed - on Windows I can still run TL'09 when I have to
(And yes, I know I could do the same on Linux, just haven't got all of the setups downloaded)
 
cfr
cfr
@JosephWright That's part of what makes it more complicated. Even before Apple's changes.
@JosephWright People don't seem to do it this way, though, because this isn't the default. I don't know it would have occurred to me unless I'd been coming from MacTeX. But Apple put me off Macs for both hardware and software reasons.
@PauloCereda Or just install Linux on it ;).
 
Paulo Cereda
Paulo Cereda
 
cfr
cfr
@JosephWright I liked TeXshop.
 
Paulo Cereda
Paulo Cereda
@cfr ^^ I am actually running Fedora on an old iMac :)
@cfr I am trying to convince Joseph to donate his lappy to the island. :)
 
cfr
cfr
4:18 PM
@PauloCereda I ran Debian on an old PowerBook for a while, although it wasn't possible to get everything working.
 
Paulo Cereda
Paulo Cereda
@cfr ooh that sounds great
 
cfr
cfr
@PauloCereda It was one of the few distributions which supported PPC.
 
Paulo Cereda
Paulo Cereda
@cfr how nice!
 
cfr
cfr
@PauloCereda But it couldn't sleep due to poor support for the graphics card. And I had my mother running Debian on an eMac. Again, PPC was a constraint.
@PauloCereda Since then, I've looked for intel wifi and graphics in laptops. And a US keyboard ;).
 
Paulo Cereda
Paulo Cereda
@cfr Ouch. I had Fedora installed on an old MacBook (first Intel-powered). It ran okay-ish, and then I went for Arch. Now it's retired. :)
 
cfr
cfr
4:21 PM
And small. And non-glossy screen.
 
Paulo Cereda
Paulo Cereda
@cfr ooh I got me a Lenovo with an AMD chipset for the first time. :)
 
cfr
cfr
@PauloCereda My PowerBook was one of the last PPC. They'd started switching to Intel, but I didn't want to be one of the first as Apple have a bit of a reputation for initial lemons.
 
The_AH
The_AH
WHAT THE N-NOT-N IS A POWERBOOK?
 
cfr
cfr
@PauloCereda How's that? Any problems? My last laptop was my first Lenovo and now I've another. Both intel, though. And Lenovo let me have a US keyboard.
 
Paulo Cereda
Paulo Cereda
@cfr Wise move. :) It was shipped with Leopard, then Snow Leopard, then... Lion. :)
 
cfr
cfr
4:24 PM
@The_AH Sort of an early version of the abacus.
 
The_AH
The_AH
what the n-not-n is the abacus?
 
Paulo Cereda
Paulo Cereda
@cfr So far, so good. I am really enjoying it. And I could even upgrade the firmware!
 
cfr
cfr
@PauloCereda Yes, I didn't regret it, even though an intel chip might have been easier later when they dropped support for the hardware.
 
Paulo Cereda
Paulo Cereda
@cfr I bought mine because the price was inviting back then (not like it is today, where a decent MacBook costs the price of a new car).
 
cfr
cfr
@PauloCereda Oh, but you can do that on the intels, too, now. It took me a long time to realise, but it's great. Mine isn't even certified for Linux, but I can still get the firmware updates. Soooo much easier than finding somebody to burn me a disk and lend me a CD drive!
 
Alan Munn
Alan Munn
4:27 PM
@PauloCereda Clearly you haven't bought a car... :)
 
Paulo Cereda
Paulo Cereda
@AlanMunn :D
 
cfr
cfr
@PauloCereda I know. I had an ibook, then a powerbook. At the time, you could get better hardware for the price than buying a PC. Then that all changed, so I switched to ThinkPad.
 
Paulo Cereda
Paulo Cereda
@cfr I so wanted a ThinkPad! They are expensive though. So I went for another Pad — the IdeaPad one. :)
 
cfr
cfr
@AlanMunn Me neither. I can't drive. I could buy quite a few bikes for the cost of an Apple, though.
 
The_AH
The_AH
ThinkPad, IdeaPad, I want the BrainPad now.
 
Paulo Cereda
Paulo Cereda
4:29 PM
@cfr I used to have Dell machines, but Dell became the new Apple around here. I bought a Lenovo lappy with better specs than a Dell counterpart by half of the price.
 
cfr
cfr
@PauloCereda My first ThinkPad was not expensive, but this one was. They stopped doing the lower end one, I think. But IdeaPad's are too big.
 
Alan Munn
Alan Munn
@cfr That's probably true. It's impressive that you managed to survive in the US without driving (I know it's not impossible, but in most places it's not simple.) I didn't learn to drive until I moved here for my PhD
 
Paulo Cereda
Paulo Cereda
@cfr I usually favour 15" screens when I have the option, but I think 14" are okay in some cases. :)
 
cfr
cfr
@AlanMunn No, it's not. It was easy in Berkeley, but less so elsewhere.
@PauloCereda 13" is too big, but I'd have one if I had to. Anything larger is too large. This one's 12.5". My last one was 11.?", I think.
 
Paulo Cereda
Paulo Cereda
@cfr Really? Oh my, that's too small for me. :)
 
cfr
cfr
4:31 PM
I used a 15" for a year or so at Berkeley and was determined not to buy one.
 
Alan Munn
Alan Munn
@cfr I agree, small(ish) screens for laptops are best, and add an external monitor for stationary use.
 
cfr
cfr
@PauloCereda Really. But they are increasingly hard to find. Everyone wants to lug a desktop around!
@AlanMunn I don't have that, but it would be ideal.
 
Alan Munn
Alan Munn
@cfr External monitors are pretty cheap these days, and for me having one makes a world of difference.
 
cfr
cfr
@AlanMunn If you don't want a netbook, though, it pushes you into things like ThinkPads. Especially if you want a non-glare screen.
@AlanMunn It's more the space, to be honest.
 
Paulo Cereda
Paulo Cereda
@cfr Indeed. I actually wanted a smaller screen for mobility purposes (I often have to ping pong between São Paulo and Rio de Janeiro), but I could not find anything relevant. I even considered a Microsoft Surface, but Microsoft doesn't sell them around here (and I think the OS is locked, but I am not sure).
 
Alan Munn
Alan Munn
4:35 PM
@cfr True. Everything depends on your specific environment for sure.
 
cfr
cfr
@PauloCereda I know. It isn't good. An option for US keyboard is also tricky, although I'd sacrifice that if necessary.
 
Paulo Cereda
Paulo Cereda
@cfr In the past, you could even find US keyboards here. Nowadays, only Brazilian keyboards.
 
cfr
cfr
@AlanMunn Websites which take half my screen are increasingly common, despite the prevalence of mobile devices.
 
Paulo Cereda
Paulo Cereda
I learned how to type with an US keyboard, I was quite good at it. :)
 
cfr
cfr
@PauloCereda Interesting.
 
barbara beeton
barbara beeton
4:38 PM
@PauloCereda -- That's Lake Webster to you.
 
cfr
cfr
@PauloCereda I originally learnt on a UK keyboard, but then I got very used to a US one and certain keys are in more convenient places. The only downside is typing sterling signs. Not a problem in LaTeX, but elsewhere it is.
 
Alan Munn
Alan Munn
@cfr On a Mac (at least with the Canadian keyboard layout, it's just option-3, which is pretty easy.)
 
Paulo Cereda
Paulo Cereda
@barbarabeeton ooh it's like Gloucester, it's spelled that way but pronounced Glouster. :) (@DavidCarlisle)
@cfr Am I a nerd for admitting I sometimes type with Unicode codepoints? :)
 
cfr
cfr
@PauloCereda I don't know. How do you do that?
 
Paulo Cereda
Paulo Cereda
@cfr Ctrl+Shift+u (you will get some sort of "underline u", which indicates you are in this mode) followed by the Unicode codepoint. Eg, U+DF will typeset ß :)
 
Alan Munn
Alan Munn
4:41 PM
@PauloCereda Place of my birth. But would say pronounced "Gloster"
 
Paulo Cereda
Paulo Cereda
Ah, and hit space after the code. :)
@AlanMunn ooh <3
 
cfr
cfr
 
Paulo Cereda
Paulo Cereda
 
David Carlisle
David Carlisle
@AlanMunn of course, no one pronounces the ce in the middle of town names.
 
Paulo Cereda
Paulo Cereda
@cfr I love this so much <3
 
cfr
cfr
4:44 PM
@PauloCereda There's also a youtube lesson on how to say it ;).
 
Paulo Cereda
Paulo Cereda
@cfr ooh I need to learn that so I can have fun when I visit the UK
 
cfr
cfr
 
David Carlisle
David Carlisle
@PauloCereda the standard pronunciation is "the station with the long welsh name"
 
Paulo Cereda
Paulo Cereda
@DavidCarlisle ooh teh Welsh
 
cfr
cfr
@DavidCarlisle 'Llanfair PG' is probably too demanding for people over the border.
 
barbara beeton
barbara beeton
4:49 PM
@PauloCereda and @cfr -- According to Wikipedia (and the translation that I learned: The humorous translation is: "You fish on your side, I'll fish on my side, and no one shall fish in the middle". Its humorous translation was perhaps invented by Laurence J. Daly, editor of The Webster Times.
 
Alan Munn
Alan Munn
@DavidCarlisle Yes, but I was pointing to @PauloCereda's 'ou' (he'd already dumped the 'ce')
 
Paulo Cereda
Paulo Cereda
@barbarabeeton lol
 
Joseph Wright
Joseph Wright
5:07 PM
@cfr I guess because I'm from a Windows background changing the path feels pretty normal
 
Alan Munn
Alan Munn
5:56 PM
A TeXShop user, Jeroen Scheerder, recently created an engine file that allows TeXShop to typeset Typst sources. Version 5.20 of TeXShop contains that engine in ~/Library/TeXShop/Engines/Inactive/Typst. This folder also contains simple typst source files, called templates, and instructions explaining how to download and install typst. Thus users can easily experiment with Typst using TeXShop.
2
From the most recent TeXShop update.
 
cfr
cfr
@JosephWright On Linux it's not usually necessary, but it isn't abnormal. It's more that the TL installer is somewhat dictatorial. It tells you do (a), (b) and (c). It also probably does those things for a lot of people because most people install by root. I don't, so it moans and tells me I have to do a, b and c. In that context, being confident you can do d instead of a and b and forget about c is a significant ask.
@AlanMunn What's Typst?
 
Alan Munn
Alan Munn
@cfr typst.app It's intended to be a modern replacement of TeX. The scripting language is python-esque which in principle will make package creation easier. It remains to be seen whether it will take off, but it has solid backing.
@cfr For those of us invested in particular packages (like linguistics example packages and forest in my case) it will be hard to convince me to shift. For people with less dependence on such packages I can definitely see the appeal. And because it's likely to be easier to write new functionality, maybe replacements will be forthcoming. But TikZ, for example, is a huge undertaking by any stretch.
 
Joseph Wright
Joseph Wright
6:18 PM
@AlanMunn Ooh
 
Ulrike Fischer
Ulrike Fischer
@AlanMunn I'm completly distressed about this engine: it has no accessibility support whatever. There is no way to create tagged pdf as they don't have pdf primitives and as the engine looses the structure information before they create the pdf, and they can't produce html either. And I doubt that that will change in the near future, as there are only some vage promises ("we know it is important"). That a "modern" engine so totally ignores a modern requirement makes me quite sad.
 
Joseph Wright
Joseph Wright
@AlanMunn Unfortunately, I think they've made some pretty series errors in inputs, etc.
 
Alan Munn
Alan Munn
@UlrikeFischer Wow, that does seem like a really bad design choice. I had no idea.
@JosephWright Yes. I'm not promoting it :), but I thought the addition to TeXShop was certainly worth noting.
 
cfr
cfr
@AlanMunn Does it have a licence?
 
David Carlisle
David Carlisle
 
Joseph Wright
Joseph Wright
6:28 PM
@AlanMunn Sure - just that like @UlrikeFischer, I fear its going to be a missed opportunity
 
Alan Munn
Alan Munn
@JosephWright Yes, it seems like it had a lot of promise.
 
David Carlisle
David Carlisle
@AlanMunn or made a lot of promises
 
cfr
cfr
@DavidCarlisle Hmm. So not really open and not accessible ....
 
Alan Munn
Alan Munn
@DavidCarlisle True. There's always enough frustrated first time LaTeX users to attract with the slogan "LaTeX is simply a hassle to use"
 
cfr
cfr
@DavidCarlisle Hmm. So not really open and not accessible ....
 
The_AH
The_AH
6:38 PM
La-Te-X. LTX. Linus Tech Expo. Leslie is secretly Linus! He was always hinting at this!
 
cfr
cfr
@AlanMunn It has a package called prooftrees, but not for constructing tableaux. Apparently based on bussproofs.
 
Ulrike Fischer
Ulrike Fischer
the one thesis says "Apart from purely visual content, PDF files can also encode a document’s structure. At a language level, Typst has this rich structural information. However, the current implementation does not yet make the best use of it.
Since the frames currently only contain locators and not a full reverse
mapping to the document’s structure, the exporter cannot write this structure
information. As a result, PDFs produced by the Typst compiler are not yet
accessible for visually impaired people." Which bluntly said means that they don't really have a plan here.
3
 
cfr
cfr
7:08 PM
@UlrikeFischer Judging by github, they are more interested in HTML for accessibility ....
 
Ulrike Fischer
Ulrike Fischer
7:20 PM
@cfr perhaps, but they have for this imho no concrete plan either how to access and convert the structure.
 
Werner
Werner
7:38 PM
@DavidCarlisle: Regarding tex.stackexchange.com/questions/698274/…; we've been trying to keep on top of this. Feel free to flag a user that is suspect.
 
David Carlisle
David Carlisle
@Werner I never read the small print, is it actually against the rules to have an account a day, or just annoying.
 
Werner
Werner
@DavidCarlisle Well, the current situation points to someone circumventing a suspension, which qualifies as sock puppetry. See meta.stackexchange.com/q/57682/168244.
Personally, it's a bit annoying. In the bigger scheme, I don't think it's harmful, but the current questions seem to have limited quality.
 
David Carlisle
David Carlisle
@Werner very basic, almost duplicate, questions laced with constantly rude side comments. But chaning the account is annoying as it means there is always a slight possibility it's someone different so you can't refer to an answer you gave a few hours earlier in quite the same way
 
Joseph Wright
Joseph Wright
@DavidCarlisle Flag and let the mod team look at the data we can see
As with anything that needs mod intervention
 
David Carlisle
David Carlisle
@JosephWright I flagged after Werner suggested it
 
cfr
cfr
8:05 PM
@UlrikeFischer It's not really an answer anyway, as people have pointed out.
@DavidCarlisle Sometimes there's a new account on the same day in the middle of the flow of questions. At least, it seems that way.
 
David Carlisle
David Carlisle
@UlrikeFischer @JosephWright ooh secret email looks interesting
 
Joseph Wright
Joseph Wright
@DavidCarlisle Indices?
 
David Carlisle
David Carlisle
@JosephWright lwg
 
Joseph Wright
Joseph Wright
@DavidCarlisle Ah - just re-read - yes agreed
 
samcarter
samcarter
8:23 PM
@cfr Looks like a new account every 4 questions - is this when rate limiting for new users kicks in?
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle math lwg?
 
Joseph Wright
Joseph Wright
@UlrikeFischer Yup
 
mickep
mickep
@DavidCarlisle Ohh, secrets.
 
Joseph Wright
Joseph Wright
@samcarter Er, no ...
 
samcarter
samcarter
@JosephWright ok, was just a guess :)
 
Werner
Werner
8:34 PM
@samcarter That's when it starts becoming obvious and one can make some informed decisions.
 
samcarter
samcarter
@Werner ah, ok :)
 
cfr
cfr
8:54 PM
@Werner What's the advantage of doing this? I can understand somebody creating a second account to vote on their own posts or whatever. I could also understand creating one because you 'messed up' and wanted a fresh start. And I can understand posting anonymously. But I don't understand the benefits in this kind of case. It just seems annoying.
 
Werner
Werner
@cfr The advantage for the person in this case is that they escape a suspension. For example, they may have a lot of question, then they land here, ask a bunch of poor questions, get suspended, but still have a lot of questions... so they create a new account, and continue.
This in itself is not all that bad if there's a benefit to the community as well (content being discussed and contributed to the site). But if the content persists to be bad then it just is annoying. Users who post a bunch of questions that eventually have negative scores (for whatever reason) distracts from the community's efforts.
But, ultimately, this violates the SE community agreement and falls under the banner of sock puppetry - creating new users to do things one had been previously told not to do (even if just for a cool-off period).
 
David Carlisle
David Carlisle
9:48 PM
another hour yet another account tex.stackexchange.com/q/698288/1090
 
Werner
Werner
Off the bat it seems like a new sock: tex.stackexchange.com/users/305804/vaite
 
Ulrike Fischer
Ulrike Fischer
10:19 PM
@DavidCarlisle and which one is the american english? some german are confused ;-)
 
David Carlisle
David Carlisle
@UlrikeFischer drop the u
@UlrikeFischer Americans don't like u so: color, behavior, ...
@UlrikeFischer someone broke includegraphics, who could I blame?
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle I bet there are some exception somewhere
@DavidCarlisle I'm just wondering about that ;-(
 
David Carlisle
David Carlisle
@UlrikeFischer your
 
Ulrike Fischer
Ulrike Fischer
10:40 PM
@DavidCarlisle oh well, \fp_set:Nn \l__tag_graphic_scale_fp { \Gin@scalex } isn't the best idea, if \Gin@scalex can be !. What are the values they can take? Can both be non-numbers or is one always a scale factor?
 
David Carlisle
David Carlisle
@UlrikeFischer one should always be a number, ! means use the same scale as the other one, but if you put scalex=blub, scaley=blub then you get what you get
 
Ulrike Fischer
Ulrike Fischer
@DavidCarlisle so testing if scalex=! and then use scaley should normally work. (Probably there will be edge cases where the bbox is not correct, but I decided not to care ...).
 
David Carlisle
David Carlisle
@UlrikeFischer yes
 
cfr
cfr
11:15 PM
@Werner Thanks. I didn't realise you could get suspended just for asking poor questions.
 
Werner
Werner
@cfr Repeated poor questions, I guess. But I think poor questions usually result from people who fight back in comments about their needs/wants through rude statements.
 

« first day (4728 days earlier)      last day (210 days later) »