hello

@barbarabeeton oh no

In LaTeX3, we have \skip_horizontal:N, which is \hskip joining the costume party. Is \kern also to be found somewhere in that party?

@Gaussler no you can essentially always use a skip

But isn’t there the issue of line break or not?

@Gaussler possibly, but then there's always penalty you can add (not with an l3 name though:-) or use a kern

@DavidCarlisle ... which gets us right back to \kern :-)

Maybe I should be an evil bitch and use \tex_kern:D, just for fun, if that exists

I like the :D commands since they resemble a very happy smiley

@Gaussler better to define your owm \mypackage_kern_horizontal:n that uses the primitive internally so it isolates such things in one place

@Gaussler At present, we've not got a 'typesetting' module but there are experimental ideas, e.g. xbox in l3trial

% \begin{macro}{\kern_horizontal:n, \kern_vertical:n}
%   \TeX{} does not treat kerns differently in horizontal and vertical mode.
%   As we want skips and kerns to have similar behaviour, we try to get similar
%   results.
%    \begin{macrocode}
\cs_new_protected:Npn \kern_horizontal:n #1
  {
    \mode_leave_vertical:
    \tex_kern:D \dim_eval:n {#1} \exp_stop_f:
  }
\cs_new_protected:Npn \kern_vertical:n #1
  {
    \mode_if_vertical:F
      { \par }
    \tex_kern:D \dim_eval:n {#1}  \exp_stop_f:

@JosephWright xbox? Are you saying that I can soon play Halo in TeX? :-D

@Gaussler It's the use x... pattern :) No Fallen Order for TeX

@JosephWright Damnit ;-). Speaking of boxes, I often run into the command \@begin@tempboxa. It has a not very enlightening definition. What is the benefit of using this over ordinary, primitive (or l3) box handling commands?

@Gaussler It's doing things to set up \height, etc.; again, take a look at xbox and it's \@@_via_internal_box:nn function (does the same thing, is perhaps clearer in why)

I di not see any \@@_via_internal_box:nn

*do

@Gaussler Ah, sorry: I'm doing a bit of tidying up, the version on GitHub calls it \@@_via_tmp_box:nn

@JosephWright But it’s not up on CTAN, right? I could not find l3trial nor xbox there, so I assumed the github version was the only one avaiable?

@Gaussler Yup:it's something experimental, really a proof-of-principle (I'm just working on a few things on it)

@JosephWright Great! Will it also have stuff like a re-implementation of \mathmakebox from mathtools? I feel slightly uncomfortable about using it too much in LaTeX3 writing since its implementation is very non-LaTeX3

On the other hand, if I try to re-implement it, I run into \@begin@tempboxa

@Gaussler No, at the moment it is about seeing if I can cover all of the kernel box structures

@Gaussler the height thing as Joseph said but perhaps more importantly it is color-safe if you use a primitve \setbox you are quite likely to trash the color stack:

\long\def\@begin@tempboxa#1#2{%
   \begingroup
     \setbox\@tempboxa#1{\color@begingroup#2\color@endgroup}%
     \def\width{\wd\@tempboxa}%
     \def\height{\ht\@tempboxa}%
     \def\depth{\dp\@tempboxa}%
     \let\totalheight\@ovri
     \totalheight\height
     \advance\totalheight\depth}

Thanks to all :-). With what you said, I provided an answer to my own question here: tex.stackexchange.com/a/559720/19809 However, I’m quite unsure about whether I am doing it the right way. Any comments are appreciated :-)

@DavidCarlisle mathb.in

@DavidCarlisle ^^ an important app

I just wonder what a cactus is doing there

@DavidCarlisle is learnlatex down? (I mean latex online)

@UlrikeFischer oh. let me check...

@UlrikeFischer try now

@DavidCarlisle compiles again ;-)

@UlrikeFischer applied standard high tech solution to computer issues.

@DavidCarlisle ashts= reboat?

@UlrikeFischer sudo service apache2 restart

@PauloCereda presumably it's an edible variety

@DavidCarlisle oh no

@PauloCereda I think you are missing an ó in your machine name

@DavidCarlisle Just in terminal (and hostname). :) The machine name has the correct ó in other places. :)

@PauloCereda we linguists worry about getting the right squiggle thing on letters

@DavidCarlisle ooh we also have one under c

@UlrikeFischer you broke hopatch at miktex?

@DavidCarlisle looks so. I also broke zref ;-(

I don't understand this release system. Do I have to undefine a new command for older releases?

@UlrikeFischer yes probably then update the annoying rollback tests to keep travis happy (simpler just to do it directly than via a pr:-) You can do it in the existing latexrelease block though I think

@UlrikeFischer was an issue reported for currentcounter or is it just something you spotted?

@DavidCarlisle I was checking some older documents against lthooks and ran into an error and then looked. basically this failed:

\documentclass{article} %
\RequirePackage{zref-lastpage}
\RequirePackage{zref-counter}

\begin{document}blub
\end{document}

@UlrikeFischer can you run Frank's update-rollback-tests.sh in the git bash on windows? (or just run the equvalent in the cmd window since it's only one line

@DavidCarlisle I can try. What does it do?

@UlrikeFischer just calls l3build to update two tests that always fail if you touch latexrelease (it's only a one-liner you can look at it in an editor)

@UlrikeFischer I left a comment on your checkin....

@DavidCarlisle grr ;-(

@UlrikeFischer now I am confused though, how come travis just passed?

@DavidCarlisle where do you see that it passed? I see a failure.

@UlrikeFischer so do i now, for some reason I was being shown travis-ci.org/github/latex3/latex2e/builds/720374828 (which is green) later ones fail

@DavidCarlisle I did replace the path manually, I hope it works and the file ending is not wrong now ...

@UlrikeFischer now it's happy

@egreg I tried taking your excellent answer to a question of mine (tex.stackexchange.com/questions/557655/…) and made a modified version, generalizing it to arbitrary symbols. I was wondering if you have any comments? Again, thank you very much for your excellent answer :-)

@samcarter_is_at_topanswers.xyz I know you are no longer active here. But would you consider editing your answer tex.stackexchange.com/a/445641/35864 along the lines suggested in my comment? If not, would you mind me editing the answer? Or would you rather have a separate answer? (The problem with a separate answer that comes in later is that it gets fewer upvotes and can look 'worse', but given the total number of votes that probably doesn't matter.)

@moewe god, it is difficult to keep track with biblatex ;-). Which reminds me: it would be perhaps good if the extended name input (with giveni=... would be mentioned in the biblatex docu too, and not only in the biber docu).

@UlrikeFischer Well, sometimes I wonder whether the stuff I pressure PLK into putting into biblatex is useful... Especially with stuff like this, it's documented somewhere, but it's hard for people to find out about.

@UlrikeFischer Agreed about the documentation for the extended name format. Tracked at github.com/plk/biblatex/issues/1038. I'm bad with documentation, though, so this may go to the back of the queue quickly

@moewe ;-). I only realized that it is missing because of question - I was so sure that this format exists that I tracked it down in the biber docu.

@UlrikeFischer Oh, I can never remember if it's given-i, giveni or given_i. And I'm never sure about where to put commas and braces, either. Hopefully PLK can find the time to write something up.

@MarcelKrüger did you see the harf mode question?

@UlrikeFischer Yes, I just pushed the fix.

@MarcelKrüger sounds as if we should really start to work on an update. Do you have list of the important changes for the docu?

@JosephWright the question tex.stackexchange.com/questions/559757/… used the tag "harftex". Can/should one rename this to harfmode?

@UlrikeFischer Yes: I guess I could mod-rename

@UlrikeFischer I added a list of the imprtant thing I could think of to the docu. There was also some Windows business, You probably know better what that was about.

@MarcelKrüger why does travis fail on the aaa-loading test? locally it worked for me.

@UlrikeFischer I don't know why it works for you, but here all unicode-math tests fail because unicode-math isn't yet compatible with the current latex-dev.

@MarcelKrüger ah right, the filehook problem, and the amsmath problem. Well locally I get errors in the log but the check pass nevertheless.

@moewe I just made the answer community wiki. Please feel free to edit whatever you like. (Thanks a lot for asking!)

@moewe and as you are already here I'd like to say thanks a lot for v2.15 - I love the new hooks!

I want to shift many \draws to the left, right, upper or down. Do you know an easy way of doing that?

@nbro \begin{scope}[xshift=1cm] ... \end{scope}

@samcarter_is_at_topanswers.xyz the values specified in (x, y) of the \draw command are in cm?

or what?

@nbro By default yes, but this can be changed

hm, I am not sure cuz I have a \draw that starts at (1.5, 0) -- (2.5,0)

so the bottom points are 1cm apart

oh, never mind

smaller than \scriptstyle?