« first day (3121 days earlier)      last day (1798 days later) » 

9:07 AM
@JosephWright hmmm interesting
Friends, never install TeX Live at the same time you are updating your system's glibc... :)
 
9:24 AM
@UlrikeFischer That's back with needing to think about moving/reviewing Bruno's string conversion module
@UlrikeFischer Expandability is not vital
 
9:54 AM
@JosephWright With beamer, for example, \AtEndDocument is too late, it happens after shipping out the last page.
 
10:38 AM
@JosephWright yes, that should be reviewed (including the questions about other converting/escaping commands like \pdfstringdef stuff).
@JosephWright no, but it would be nice at least for such names if there were a sensible \pdfescapename-version (I think in some of Heiko's code there is one).
@AlexG I'm not worrying about being to late (I just tried my example in beamer and it seems to work fine) but being to early: I have to step over all pages and this is a problem if some pages are created later @JosephWright.
 
 
1 hour later…
11:43 AM
Just came across this github repository: github.com/pgf-tikz Looking at the list of authors, it is clearly a spoof, right?
 
11:56 AM
@LoopSpace no that's the real stuff.
 
@UlrikeFischer For \pdfcatalog it would be too late, at least.
 
@AlexG Could be, but probably there is no need to write it so late. But I made note so that I check it when I look at this (I wanted to do it next, when I have finished the resource stuff).
 
12:12 PM
@UlrikeFischer But ... but ... but ... @DavidCarlisle ??? It'll all be reimplemented in pictures!
 
@LoopSpace Henri invited me and Joseph too - which means everything gets reimplemented in expl3 ;-). And @DavidCarlisle has a gold tikz badge.
 
@UlrikeFischer It's amazing how much reputation you can get from simply writing "Try using picture" over and over again.
4
 
@LoopSpace but less than you can get by pointing out spurious spaces ;-)
6
 
@UlrikeFischer Banter aside, I'm pleased to see it's on github as (purely selfishly) it's easier to follow.
 
12:29 PM
@LoopSpace yes that's quite good, Henri made a good job here.
 
 
1 hour later…
1:39 PM
Hey there
A possibly weird question: bar charts and semilog axes don't really go well together, do they?
 
@Vogel612 Of course they do.
 
hmm ...
 
@LoopSpace wouldn't that be for the best?
 
^^ then I must be doing something really wrong...
I can't find a difference in the plot options between these two aside from the ymode=log
 
@LoopSpace that is just people who list their group membership in public, there are unlisted people, some of whom you may feel more associated with tikz:-)
 
1:48 PM
@Vogel612 hi! Nice to see you around!
 
Hey @PauloCereda 😀 Always when I'm struggling with TeX
@marmot I struggled a bit with the data labels, but got that hacked around with "explicit symbolic", but I can't get the offset from the bottom to vanish ...
 
2:06 PM
Okay I'm pretty sure I found a bug...
I think enlargelimits interferes with the layouting of bar plots.
specifically with the display layouting for values close to 0
\begin{tikzpicture}
 \begin{axis}[ybar
    ,width=6cm
    ,xtick=data
    ,enlargelimits=0.3
    ]
    \addplot+[point meta=rawy] coordinates{
      (1, 1) (2, 10)
    };
  \end{axis}
\end{tikzpicture}
\begin{tikzpicture}
  \begin{axis}[ybar
    ,width=6cm
    ,xtick=data
    ]
    \addplot+[point meta=rawy] coordinates{
      (1, 1) (2, 10)
    };
  \end{axis}
\end{tikzpicture}
^^ the first one's bar charts are "lifted off" from the display port, the second one's are not
apparently it's not the only contributing factor, though.
 
2:48 PM
@Vogel612 Well, in the first example you have ,enlargelimits=0.3.
 
3:03 PM
yea, about that... if you replace the first datapoint with a (1,4) or any larger y value, it behaves "as expected"
(was out to get sustenance, soz for unresponsiveness)
additionally @marmot it's not like the second one specifies enlargelimits=false, so it should behave as defaulting to enlargelimits=0.1
 
3:41 PM
@Vogel612 It might be that you are looking for log origin but this depends on the version you are using as the default value has changed. However, since you only post fragments one can only guess, and unfortunately my crystal ball got stolen...
 
if you want a standalone preamble in additon to that, it's \usepackage{pgfplots}\pgfplotsset{compat=1.16}...
I've worked quite a bit on getting it to a state that would be appropriate for asking on StackOverflow to rule out weird issues.
I fixed my immediate issue by setting enlargelimits=false and manually specifying the viewport
that does not change the fact that using (1,10) (2,10) as data with enlargelimits=0.3 behaves as expected of a bar plot and using (1,1) (2,10) as data does not...
and manually specifying a viewport does not fix the issue unless enlargelimits is set to false
 
4:29 PM
@Vogel612 Yes, I think it would make a lot of sense to ask a self-contained and clear question on the main site in which you provide a complete MWE and a clear description of the issue. You started out with a question bar charts and logarithmic axes, and it seems that now the focus is something entirely different.
 
@PauloCereda Canadian ducks in the news. cbc.ca/player/play/1524038723539
3
 
@marmot hmm ... I hope to get around to it on tuesday, then. Got to finish some other work before that ... I do have a workaround, so I'm honestly not too invested anymore :)
the logarithmic axes were the only thing that I could discern as different, because I didn't consider the possibility of the magnitude of data being relevant for the layouting of the bar chart.
but the actual root cause was the data magnitude.
 
 
1 hour later…
@CarLaTeX It's that time of the year. :)
 
@AlanMunn :)
 
Hello everyone!
Does someone know if there is a minimum reputation threshold to ask on meta.stackexchange.com?
The OP in this question says they don't have enough...
 
6:05 PM
@PhelypeOleinik I think you need to have 5 on the main site to post on the meta for that site.
 
@AlanMunn @CarLaTeX Oh, that explains it. Thank you both :)
 
5
participate in meta
Discuss the site itself: bugs, feedback, and governance
 
@CarLaTeX Makes sense. You first need to use the site, and then complain about it ;)
 
@egreg Thanks for your reply to my questions. However, below the comment block where your response was displayed, someone (or something) added the line: Please avoid extended discussions in comments. Would you like to automatically move this discussion to chat? I was reluctant to put my thanks there! What gives with such a remark? Can't an OP ask for clarification on some point made in an answer?
 
6:17 PM
@OneMug that is an automatic message from the system, if you click on it it opens up a new chat room specific to your question
 
@OneMug This is added automatically by the system when a comment thread gets too long.
 
@OneMug so the idea is that you can have a chat room and get clarifications, but the main site should be easily readable by others just reading question and answer posts without having to read through old comment threads
 
@DavidCarlisle Does "move discussion to chat" remove the question from the system? Or does it remain as it was prior to the "move"? That is, does the question remain in tact on the system so I can complete it (by selecting an accepted answer for instance) and the chat thread simply happens "in parallel"?
 
@OneMug no everything (including existing comments) stays just where it is, but a new room with the same interface as this chat is opened
 
@DavidCarlisle Ah, that makes a little more sense.
 
6:26 PM
@DavidCarlisle OK. Good to know. Thanks.
 
@LoopSpace some bloke called Till for example
@LoopSpace I am very hurt that you don't trust me with \put \draw etc.
 
@DavidCarlisle Is he still involved? I must admit I don't know all the story here but I'd vaguely picked up on Till having bowed out.
@DavidCarlisle I trust you with \draw as much as I trust you with \catcode. i.e., not at all.
 
 
1 hour later…
7:47 PM
what characters are allowed in \bibitem[label]{citekey} (for inline bibliography)?
and, where would I find out such a thing?
I found a hint somewhere that says citekey can be alphabetic? alphanumeric? “with punctuation except for the comma”, and it’s all 7-bit ASCII, that much I know, but…
(writing some code that translates nroff-like syntax to bibitems)
(I already wrote code that makes \bibitem{foo} the same as \bibitem[foo]{foo} (as I use the [foo] style of citation references, not [1]), counts the width of the items and stores it in the aux file so the next run will initialise the thebibliography environment with the correct size)
(in case someone is interested in that)
 
4
Q: What characters are allowed to use as delimiters for BibTeX keys?

andseliskSurprisingly, I wasn't able to find the list of symbols which can be used for separating fields in BibTeX keys. Many programming languages have their own lists of exceptions, but it seems there is none, neither for BibLaTeX, nor for Biber. Consider the following example: @article{Doe1970, aut...

 
@AlanMunn ah thanks… does this also apply if one does not use bibtex?
ah and it only shows a negative list, I kinda need a positive list because punctuation is very ambiguous
 
@LoopSpace I know nothing:-)
 
@mirabilos How so? Only comma is disallowed for the regular punctuation marks.
@mirabilos I would say yes.
 
@AlanMunn what are “the regular punctuation marks”?
 
8:01 PM
@mirabilos at least .?!;:
 
oh well, I’m also exporting to XHTML, so the IDs need to comply to both HTML4 and XML id specs as well, and [A-Za-z][A-Za-z0-9_.-]* looks like it’s valid in Tₑχ
thanks
oh, not - then?
 
@mirabilos Yes, that's allowed too. (I don't know if I'd call that punctuation, though.)
 
@AlanMunn hm ok, then I guess I’m good; thanks!
 
@mirabilos Yes, you are fine with the spec you gave for sure.
 
and the label probably has less restrictions?
 
8:08 PM
@DavidCarlisle I have long suspected as much ...
 
@mirabilos No, it has more restrictions, since it must be valid textmode TeX, so _ will be disallowed, (since that's a mathmode character). But of course it can include macros and things too, so in that sense it is less restricted.
@mirabilos So e.g. Foo_Bar is an invalid label, but a valid key, whereas Foo\_Bar is a valid label but an invalid key.
 
@AlanMunn yeah, but I run the string through the totex escape function anyway, so that will be fine
 
@LoopSpace standard reply in this chat:
@LoopSpace you are mean
@mirabilos not the key, presumably
 
@DavidCarlisle like `echo "\\bibitem[$(totex "$label")]{$(totex "$key")}%" modulo syntax specifics
basically, escaping everything ever output ;)
 
@mirabilos no that would be wrong
 
8:19 PM
hmmm but that would break with the underscore, right?
 
@mirabilos all the quoting in the key is wrong, _, and & can be used as is and \& and \_ would not work (for example)
 
@DavidCarlisle mmh, and that even breaks the one-argument form of bibitem I wrote… thanks for the heads-up!
oooooh and, yak shaving, \_ in the label expands to \protect \T1\textunderscore in the aux file, which later breaks… niiiice
\unexpanded to the rescue
 
8:36 PM
Is there a way to make tlmgr reinstall a package that I might have messed up?
 
@PhelypeOleinik tlmgr install --reinstall <name>
 
@JosephWright Thanks :D
 
@AlexG Sure, but on a first pass \AtEndDvi doesn't work, so one has to put this somewhere; in many documents, it will be OK]
 
@MarcelKrüger did you see the luaotfload-features problem?
 
@JosephWright not following that closely, but why is atenddvi different on first pass?
oh it checks for last page?
 
8:47 PM
@DavidCarlisle Yup
@DavidCarlisle It's Heiko's code to add to the last shipout; necessary for things like closing objects. I think it LuaTeX one can handle this rather more easily
 
@JosephWright yes (I think I knew that once:-) it seems structurally wrong, you ought to be able to hook into \@@end but I guess that would require engine changes as the last page may have been shopped by then
 
Hello!
I have a question
 
@DavidCarlisle See Heiko's packages: one can create a hook there (as does etoolbox), and that's fine for some things, but it's not good for shipout. Like I say, in LuaTeX, one can hook the end-of-PDF, which what you really want
 
I finally cracked and bought a new phone today (as banking app refused to work as mine was too old:-)
@JosephWright yep
 
We know that some users predicate about the use of \[\] instead of $$$$ when a user ask in the site. That is fine, excellent
 
8:50 PM
@DavidCarlisle I dont think it's possible to know which is the last shipout other than in a two-pass sense in pdfTeX
 
@DavidCarlisle as Alex just reminded me: there is a problem with atenddvi, it adds an additional box: github.com/ho-tex/oberdiek/issues/28.
 
However, when a user post a question/answer using $, the same users do not say "Don't use $, use \("
So my question is: we know that in LaTeX one should use \( instead of $, but why the users don't suggest that?
 
@manooooh that's because $$ does not work in latex but $ is fully supported syntax.
 
@DavidCarlisle it's that why you wrote an answer without zzz? Couldn't you find the key? ;-)
 
@UlrikeFischer oh I was just reading the dtx which says there is no box
@UlrikeFischer atenddvi.dtx says % Unlike \cs{AtBeginDvi} the \meta{code} is not put in a box and
 
8:53 PM
@DavidCarlisle hi! Why do you say "$$ does not work in latex" if I can write it without any problem?
However
 
@manooooh until a couple of years ago $ was recommended over \( in many situations as it is robust but \( was fragile (it is robust in recent releases)
 
@DavidCarlisle well the example in the issue clearly shows that somewhere there is an additional box level.
 
@DavidCarlisle which recent releases?
 
@manooooh it is not supported, and it does not work if for example you use fleqn documentclass option
@manooooh from 2015 i think
@UlrikeFischer you are having a bad influence, I read the doc not the code
 
@UlrikeFischer No, which problem?
 
8:55 PM
@DavidCarlisle oh, ok, thanks. However, you are not being consistent, because if you suggest \[ "For math mode" and $ "For inline mode" it is not the same in the sense that the way of how you write a symbol like $ or \[
 
@manooooh I do not understand that comment
 
\[ for math mode, but $ for inline mode??? Wow, that is inconsistent
 
@manooooh no latex documentation mentions $$ at all, the latex book documents $ as an alternative to \(
 
@MarcelKrüger luaotfload-tool doesn't like the changes in luaotfload-features: github.com/u-fischer/luaotfload/pull/52#issuecomment-493697582
 
It should be \[ and \(, or $$ and $
@DavidCarlisle because inline or display, the math is the same. You write math in both environments. So if you do that, then you need to be consistent with the symbols you are using
 
8:57 PM
@UlrikeFischer We can solve that I think
 
@manooooh no. There are many environments that start display math, \[, equation, math, align` etc, why should they have corresponding inline math versions?
 
@DavidCarlisle yes, I know it is an alternative. I'm going further. I'm talking about whether it's consistent or not (which clearly is not)
 
@UlrikeFischer Probably we need the same split as the begin-shipout case: a token hook, and a box one
 
@DavidCarlisle I was referring to "math", not some specific environments
 
@manooooh just because \( has $ as an alternative syntax why do you assume just one of the many display math environments should have an alternative syntax?
 
9:00 PM
@DavidCarlisle if you use \[, then the logic would say that the symbol to write in inline would be \(, even though $ is an alternative of \(
 
@manooooh I do not see that at all
 
@manooooh Well not exactly: the current team position is I think to favour $ for inline math, simply as over time it gets very tedious to use \(. I think on balance this was a less-good plan by Lamport.
 
@JosephWright so you prefer not to write tediously over consistency. Nice
@DavidCarlisle imagine you write a document using \[ for math mode. In one paragraph, you use $$ instead of \[. That is an inconsistency. The same here: if you use \[ but not \( for inline mode then it is inconsistent
 
@manooooh Like @DavidCarlisle says, I'm not seeing the consistency argument here. If we go that way (and one can for display math), you'd use \begin{equation*} for display math, and have some new environment \begin{math} ...
 
@manooooh your consistency argument is spurious. There is no way you could assume just because one environment is called by the unusual \[ syntax that the math environment should use the unusual \( syntax
@JosephWright actually that is not a new environment:-)
 
9:04 PM
@DavidCarlisle Ah, well .... i was mainly thinking of math mode in ConTeXt, to be honest
 
@JosephWright I don't see an inconstency there, why do you say that? If you always use \[ then it is fine, but for inline mode, to being consistent with the syntax of \[, then you write \(
 
@manooooh no. Where do you get this [ --> ( correspondence from?
 
@DavidCarlisle I am saying that; if you don't respect the \[ syntax then you are being inconsistent in the code
 
@manooooh Well for a start it's not 'always' \[. That's a shortcut for \begin{equation*}, but does not apply to the unstarred equation environment, any of the AMS ones, etc.
 
@manooooh I give up.
 
9:06 PM
@DavidCarlisle well, they have in common a backslash. \( and $ don't have in common anything
 
@manooooh It depends how far you want to go. XML is totally consistent in this sense, but most people don't choose to hand-edit XML. ConTeXt for MkIV has moved from \item, to \startitem ... \stopitem in this vein, but it's not gone pure-XML
 
@manooooh sorry your argument is impossible to follow and in any case pointless as none of this syntax has changed since 1983 and isn't going to change now.
 
@JosephWright so we need to discuss if a shorcut brings problems, which is not true, I guess
 
@DavidCarlisle It may ;)
2
 
Now the important issue of seeing if Manchester City win Eurovision.
2
 
9:10 PM
@JosephWright I just want to be consistent. It is normal that the syntax can change over time. If \item is now called \startitem ... \stopitem it is okay, but may not respect others way of writing an bulleted/numbered list, which in that case would be inconsistent
 
What is the reference for texmf.cnf settings such as half_error_line and the like?
 
@PhelypeOleinik texdoc web2c
@manooooh you can use \begin{displaymath} for displaymath and \begin{math} for inline math, that uses the standard environment syntax and consistent naming. Those environments have been in latex since the start but they are far from the most popular environments that are defined.
 
@PhelypeOleinik texdoc tex, page 7
 
@DavidCarlisle I found memory settings in section 2.5 Runtime options, but nothing about half_error_line :/
 
@PhelypeOleinik oh I didn't try reading the doic:-)
 
9:16 PM
@DavidCarlisle wow, nice to know. I like those tags as I like \[ and \(, as I also like $$ and $. I don't care if they are not popular, they have the correct syntax (and works perfectly in LaTeX, which is very important)
 
@DavidCarlisle I thought so ;)
 
@manooooh $$ is simply wrong
 
@DavidCarlisle I know.... just an example of what a consistency is
 
@egreg Thanks, that's it. Now I just have to understand that :)
 
@manooooh like all environments, they are fragile though
 
9:17 PM
@DavidCarlisle hm, well, a system like LaTeX can not be perfect
 
@DavidCarlisle You can probably guess what I'm going to say ...
 
@PhelypeOleinik Look at modules 311 and 315–317
 
@egreg well that's the internal tex version, not that it can be set in texmf.cnf
@JosephWright use context not latex environment syntax?
 
@DavidCarlisle No ....
 
@JosephWright I know you are a context mole really:-)
 
9:19 PM
@DavidCarlisle \RequirePackage{etoolbox}\robustify\begin\robustify\end, or more simply \protected
 
@DavidCarlisle There's nothing about them in texdoc web2c, because those variables are explicitly said to be possibly different in INITEX and in virtex
 
@JosephWright what is it for?
 
@DavidCarlisle Well yes
@manooooh Deals with environments being fragile
 
@egreg yes but isn't the texmf.cnf documented somewhere
 
@JosephWright I don't understand. Are array, multline,... fragile?
 
9:21 PM
@DavidCarlisle Not completely, it appears.
 
@manooooh yes, or at least \begin{array} is
 
@manooooh Well, \begin is ...
 
@egreg the comments in the standard texmf.cnf :-)
 
@egreg Thanks, the description of module 311 says what they are. At least now I can know what I'm breaking :D
 
@JosephWright I'd have to think whether an etex protected \end was always safe if \end{array} comes at the start of a cell...
 
9:22 PM
@DavidCarlisle hm, well, I have stopped using \begin{array} long time ago. Why is array and others fragile?
@JosephWright \begin is...? :O
 
@manooooh what do you mean stopped using?
 
@DavidCarlisle I don't use array anymore. Now I use pmatrix, vmatrix...
 
@manooooh oh OK, well same applies to them of course
 
:C
 
@manooooh \show\begin
 
9:24 PM
% It's probably inadvisable to change these. At any rate, we must have:
% 45 < error_line      < 255;
% 30 < half_error_line < error_line - 15;
% 60 <= max_print_line;
% These apply to TeX, Metafont, and MetaPost.
error_line = 79
half_error_line = 50
max_print_line = 79
@DavidCarlisle Only this ^^
@DavidCarlisle At least it says it's inadvisable to change :)
 
@JosephWright what does that have to do with the fragility or not of the environment?
 
@PhelypeOleinik yes I was going to point at those comments but thought I would do the right thing and reference the web2c doc instead;-)
@manooooh it is \begin itself that is fragile, so \begin{anything} is fragile
 
@DavidCarlisle why?
 
@DavidCarlisle Good practices :-)
 
So what environments do you guarantee that are not fragile for LaTeX?
 
9:25 PM
@manooooh because that's what being fragile means
 
@manooooh Fragile = explodes inside \edef
2
@manooooh None, because \begin itself is fragile
 
@manooooh none, as I said above, all environments are fragile
 
@DavidCarlisle well, if we live with that, then we should never use LaTeX to write a math document......
 
@UlrikeFischer You got a PR. Basically we can just load the tables, it kind of makes sense to have them in luaotfload-tool too: The tables are mapping to human-readable names, so luaotfload-tool might have even more use for them than the remaining loaotfload.
 
@manooooh Not really, you just have to stick to the rules of LaTeX syntax
 
9:27 PM
@MarcelKrüger yes, thanks just saw it. I'm just pulling.
 
@manooooh nonsense of course, every user has lived with that and many math documents have been written.
 
@JosephWright yes! That's what I mean by consistency! Thanks Joseph! And what is the LaTeX syntax, where can I find it?
 
@manooooh the latex book (or any latex documentation)
 
@DavidCarlisle by Lamport?
 
@manooooh As in 'what is described in Lamport's book'. He never puts an environment into a place where they are not safe
 
9:28 PM
@manooooh yes
 
@DavidCarlisle well, Overleaf documentation, ShareLaTeX (not available now)...?
 
@manooooh Lamport's book defines LaTeX syntax
 
@JosephWright ... but he do puts environments... and environments are, by definition, fragile... so?
 
@manooooh Yes, he uses environments, which is fine as he uses them where they are safe
 
@manooooh I haven't read them but same as answers on this site, they either describe the correct syntax or they are wrong. the latex book is different as that defines latex.
 
9:30 PM
@JosephWright when an environment is considered safe?
 
@manooooh when they are not in a moving argument
 
@DavidCarlisle again my bad influence ?
 
@UlrikeFischer yes
 
@manooooh When it is not used in a moving argument or similar. Basically, you can't put them inside things like captions
 
@JosephWright @DavidCarlisle what is a "moving argument or similar"?
like {ccc} in \begin{array}{ccc}?
 
9:32 PM
@manooooh a fragile command is one that sometimes needs \protect, and a moving argument is an argument in which some commands need \protect
 
@manooooh something that is copied to another place, e.g. a caption that also goes in the listoffigures, or a section title that gets also in the header,
 
@UlrikeFischer thanks Ulrike and David!
 
@JosephWright @UlrikeFischer Should I close the issue I opened about the messages?
 
@PhelypeOleinik If you are happy it's resolved, then close
 
@JosephWright Happy is not the word, but if there's no solution then It's good enough :)
 
9:43 PM
@PhelypeOleinik I don't think we can tell from the macro end that the value has altered. I guess we could add another variable (line the one for max_print_line), to allow the code to adjust
 
@JosephWright If it's possible, I think it would be nice.
 
hm, does texlive looks for scripts only in the main texmf?
 
@UlrikeFischer There's the business about setting TEXMFCNF
 
@manooooh I think before you continue this discussion you should read tex.stackexchange.com/q/4736/2693
 
@JosephWright vv
\documentclass{article}

\begin{document}

\begin{tabular}{|cc|}
  a&b\\
\hline
\end{tabular}


\bigskip

\let\oldend\end
\protected\def\end{\oldend}

\begin{tabular}{|cc|}
  c&d\\
\hline
\end{tabular}

\end{document}
 
9:50 PM
@DavidCarlisle Yeah, I have something about that in xparse
 
@JosephWright ? what has texmfcnf to do with it?
 
@UlrikeFischer Oh, sorry, misread your comment
@UlrikeFischer It does now: I raised that there was a security issue otherwise
 
@JosephWright etex v 1 expanded protected commands at start of table cells and we persuaded them to change:-)
 
@DavidCarlisle It's still fixable, though, just have to work a bit harder
 
@JosephWright everything is always fixable:-)
 
9:53 PM
@AlanMunn thanks Alan! I will read it
 
@JosephWright hm, I wonder since when, I never realized until I now try to test a new luaotfload-tool version.
 
@UlrikeFischer When we moved l3build to working as a script; behaviour was not consistent between Linux and Windows
 
@MarcelKrüger and luaotfload-tool --version gives /luaotfload/luaotfload-tool.lua:345: bad argument #2 to 'stringformat' (number has no integer representation)
@JosephWright this makes testing such scripts a bit complicated ;-) How do you do it with l3build?
 
@UlrikeFischer texlua l3build.lua: done that way, we have usual kpsewhich searching
 
10:15 PM
@MarcelKrüger hm, it looks as if luaotfload has a bit curious idea about the luatex version, it reports a table consisting of "1.1" and "10" an "0".
Oh a "update miktex in both modes" question on the latex list ;-)
2
 
@UlrikeFischer I leave that one for you, we blame you for all miktex issues:-)
 
@UlrikeFischer Oh yes, it's the same all the time
 
@DavidCarlisle !
 
@DavidCarlisle tomorrow ... I'm going to bed now.
 
@UlrikeFischer I added the fix to the old PR.
 
10:29 PM
@UlrikeFischer night
 

« first day (3121 days earlier)      last day (1798 days later) »