« first day (3291 days earlier)      last day (1625 days later) » 

5:24 AM
@UlrikeFischer I have already updated the database and packages several times, to no avail. Indeed, the log did not match the log, sorry. I just put the code that matches the log in github.
 
 
1 hour later…
6:27 AM
@AndréC link?
 
7:07 AM
@AndréC what do you mean by the database? I'm not speaking of the FNDB but of the database you can update when you view the package list.
 
7:46 AM
@DavidCarlisle ooh
@DavidCarlisle so DPC = JW, JW = PC, then... PC = DPC?!
Let's try....
@JosephWright dinner!
 
8:22 AM
@barbarabeeton, @yo' speaking of Domino's. :)
 
 
1 hour later…
9:43 AM
@JosephWright I found the reason why chat.stackexchange.com/transcript/41?m=52419876#52419876 fails, with xkeyval \CurrentOption is passed twice.
 
9:53 AM
@UlrikeFischer Ah, right: is this our fault?
 
@JosephWright I just added an issue, a bit long for the chat.
 
@JosephWright ooh
 
@JosephWright actually one can simplify the problem. This here fails too:
\documentclass[]{article}
\PassOptionsToPackage{driver=dvisvgm,driver=dvisvgm}{expl3}
\usepackage{expl3}
\begin{document}
abc
\end{document}
 
@UlrikeFischer Give me five minutes
@UlrikeFischer Should be fixed: minded to release (as we are now really down to bugs-only in l3kernel changes)
@UlrikeFischer, @DavidCarlisle, @egreg, @PauloCereda, @PhelypeOleinik siunitx v3 now influencing kernel decisions ... :)
 
10:13 AM
@JosephWright ooh :)
 
@PauloCereda See mail
 
10:50 AM
@JosephWright Better than later having to fix packages or the kernel!
 
@egreg Well yes
 
11:04 AM
@JosephWright this here fails too:
\documentclass[]{article}
\PassOptionsToPackage{ undo-recent-deprecations,undo-recent-deprecations}{expl3}
\usepackage{expl3}
\begin{document}
abc
\end{document}
 
11:53 AM
@UlrikeFischer I'll look at that over lunch
 
12:11 PM
@JosephWright no need to hurry, I only thought I should check the other keys too.
 
12:25 PM
Whoever wrote the [H] float specifier has a LOT to answer for. Every single student report I read uses it, with messed up spacing as a result
 
@WillRobertson Er, well, it's kind of handy ...
@WillRobertson I mean, I favour ht, but ...
 
Also every second student report I read either uses blank line+\bigskip or \\\\ to add spaces between paragraphs manually :( :(
@JosephWright Yes, I can’t really blame anyone really. It just hurts me on the inside. Time to actually start LaTeX I guess rather than assume students pick it up by osmosis
 
@WillRobertson did you got the mail regarding the team meeting time?
 
@WillRobertson Hi!
@WillRobertson <- one of the most awesome persons I had the pleasure to meet! :)
 
@UlrikeFischer not yet... a bit crazy over here this week :-(
@PauloCereda no you are!
 
12:36 PM
@WillRobertson aww
 
@UlrikeFischer I mean I click on the button that says refresh the file name (or something like that). But since I'm obviously the only one with this problem, I uninstalled MikTeX and am reinstalling it.
 
@AndréC I meant this here ^^^. Do you really deinstall and reinstall the whole miktex because of one package? Why didn't you deinstall the iftex package and reinstalled it?
 
12:52 PM
@UlrikeFischer I did that too. But since the only functional solution is for me to add the iftex folder in my personal texmf, I conclude that there is a bug in my installation.
 
1:33 PM
@UlrikeFischer Looks like the middle thursday slot works for everyone. I might be a bit tired and/or in bed already by 10:30pm but should be fine most of the time...
@egreg close to your heart perhaps — digg.com/video/italian-chefs-carbonara
 
1:50 PM
If I were to talk/teach TikZ to a group of undergrads, who know bare minimum LaTex document preparation, what should my talk consist of? (I am neither a teacher nor speaker, this opportunity just opened up)
 
@PauloCereda Could you please verify that getCommandWithWorkingDirectory actually works?
 
@GermanShepherd I think you need at least to tell them how to make a 'bare bones' document of the 'Hello world' variety
 
!config
identifier: all
name: All TeX files
commands:
- name: All
  command: >
    @{
        foreach(ref : listFilesByExtensions(".", ["tex"], true)) {
          var parts = ref.toString().split("/");
          var path = ".";
          for (int i = 1; i < parts.length - 1; ++i) path += "/" + parts[i];
          System.out.println(
            getCommandWithWorkingDirectory(toFile(path), [command, parts[parts.length-1]]));
          var res = unsafelyExecuteSystemCommand(
            getCommandWithWorkingDirectory(toFile(path), [command, parts[parts.length-1]]));
@PauloCereda That is, whether my answer here actually answers the question…
 
@GermanShepherd Something like 'LaTeX is a bit like a program to describe how to make a document: it uses \foo as 'commands' to tell the computer what to do, and \begin{foo} ... \end{foo} for 'environments' containing related instrutions
@GermanShepherd 'A minimal file might look like'
@GermanShepherd ... then say 'TikZ is an add-on that can do diagrams: we load it like ... and then use it like ...'
 
@TeXnician let me check
 
2:16 PM
@WillRobertson, I found out a couple of my friends have you as their final year project supervisor. I'm sad to find out you are not having success in convincing them to use LaTeX... :)
 
ooh mr. koala is here
 
@WillRobertson Even Italians disagree on how to make “real” carbonara; one point is sure: no cream!
 
@TeXnician: The unsafely thingy does not offer support for working directory: github.com/cereda/arara/blob/master/application/src/main/java/…
Would this code be acceptable?
!config
identifier: all
name: All TeX files
commands:
- name: All
  command: >
    @{
        a = [];
        foreach(ref : listFilesByExtensions(".", ["tex"], true)) {
          var parts = ref.toString().split("/");
          var path = ".";
          for (int i = 1; i < parts.length - 1; ++i) path += "/" + parts[i];
          a.add(getCommandWithWorkingDirectory(toFile(path), [command, parts[parts.length-1]]));
        }
        return a;
    }
arguments:
- identifier: command
  flag: >
    @{
 
2:31 PM
@PauloCereda How trivial. Well, that should be documented somewhen…
@PauloCereda Let me try.
 
@TeXnician my bad. :)
 
@PauloCereda It works great. Thanks :)
 
@TeXnician My pleasure. :)
 
2:59 PM
@PauloCereda -- I wonder, if they are all lined up single file, close behind one another, if someone pushes the one in the back of the line, will they all topple over, one after another?
@WillRobertson -- Unfortunately, I've seen places where someone teaches the use of \\ as a preferred way to break lines. There is even a video on youtube that dies this (although I don't remember at the moment how to find it). I've just taken a look at the LaTeX WikiBook to see what it says. On the page en.wikibooks.org/wiki/LaTeX/Paragraph_Formatting instructions are ambiguous.
I think a great improvement there would be to carefully distinguish between "paragraph" (a block of text) and the sectioning command \paragraph. Then the examples given might be more clear.
 
3:17 PM
@barbarabeeton ooh
 
@PauloCereda -- Actually, that was intended for @WillRobertson, but if someone else reads it who can do something about the WikiBook, that would be great. I've tried a few times to enter edits, but they've always been refused, presumably because whoever is in charge doesn't believe I know what I'm talking about.
 
3:54 PM
@UlrikeFischer I just reinstalled MikTeX and now everything works normally : -)
 
@JosephWright do you have texshop?
 
@UlrikeFischer No, no Mac any more: ask @egreg or Frank
@UlrikeFischer This is why I need a Macbook Pro ;)
 
@JosephWright ;-).
 
4:24 PM
@UlrikeFischer Only largely joking ... there are reasons for wanting access to a Mac
 
4:40 PM
@JosephWright Mac games
 
@PauloCereda Er ...
 
@UlrikeFischer Yesterday you wanted to discuss the "merge / tag order" for luaotfload. What did you mean there?
 
5:16 PM
@MarcelKrüger I don't want to make a mess out of the branches and the tag. If I want to release 3.1, I have to - in my master-XX branch change build.lua to the correct version number, tag the files, test and build everything - sent to ctan - set a tag v3.1 (which will create a release) - merge the code into master - merge into dev - delete the master-XX. Does this sounds correct and should some step should done earlier or later?
 
6:05 PM
@WillRobertson shocking, evil terrible idea. Who was that?
 
@JosephWright I was thinking I could talk about nodes, drawing simple lines, curves, and perhaps get them to draw something.
We are talking to an audience familiar with LaTex, atleast the basics.
 
6:34 PM
@GermanShepherd Ah
@GermanShepherd I'd start with lines and curves then nodes
 
@UlrikeFischer some smaller remarks: I would create the tag first, then send to CTAN. Then the Travis build is a last check that the tests really pass and it makes sure that a released version is always a tagged release. Also for the merge to master we should ensure that it is always a fast-forward merge, otherwise there is something seriously wrong. So maybe this should be done first.
I would do the merge to dev last because it is the least important step and the only step where we it is quite likely to get merge conflicts. So my order would be: Do a fast-forward merge to master, then set the tag. After the Tag tests ran through send the code to CTAN and delete the master-xx branch. Finally merge master back into dev.
 
@MarcelKrüger ok, so I will build, then merge with master and if this gives a conflict ask for help ;-)
 
@UlrikeFischer you could try git 2.24 :-)
 
@DavidCarlisle what will it break?
@DavidCarlisle did you see the iftex issue?
 
@UlrikeFischer oh everything probably:-)
@UlrikeFischer no?
 
6:42 PM
@DavidCarlisle ooh
 
@DavidCarlisle we overlooked that scrbase has a \ifpdftex command (and a \ifVTEX)
 
@UlrikeFischer I must have missed the mail notification as well, odd.
 
@DavidCarlisle I didn't get one either. I looked at komascript.de as I got a warning and saw there the reference.
 
@UlrikeFischer Not one here either
 
@UlrikeFischer making a \ifpdftex that isn't a tex \iftrue or \iffalse is just weird.
 
6:49 PM
@DavidCarlisle Not a great plan, no
 
@UlrikeFischer @JosephWright I think the answer has to be too bad here, it's not as if we added new names so scrbase was already incompatible with ifpdftex and ifvtex packages
 
@DavidCarlisle True
 
@DavidCarlisle didn't you add new casing variants?
 
Package scrbase Warning: \ifpdftex was redefined
(scrbase)                at the document preamble.
(scrbase)                If you did not redefine it by yourself, have a
(scrbase)                a look at the description of this command at
(scrbase)                the manual of every package you are using.


Package scrbase Warning: \ifVTeX was redefined
(scrbase)                at the document preamble.
(scrbase)                If you did not redefine it by yourself, have a
(scrbase)                a look at the description of this command at
 
@UlrikeFischer just checking.... iftex's was \ifPDFTeX
@JosephWright anyway I think scrbase can load iftex then redefine anything it wants to define so documents using that package shoudl be affected
@UlrikeFischer @JosephWright you two are not watching iftex (as it was transferred from bidi) si you wouldn't have been notified github.com/latex3/iftex/watchers
 
6:55 PM
@DavidCarlisle Ah
 
@DavidCarlisle the issue at koma is here komascript.de/node/2262. There is also a comment from Markus.
 
@UlrikeFischer I guess google translate got the general gist that he's annoyed.
 
@DavidCarlisle etoolbox has tons of \if-commands which aren't \iftrue or \iffalse. It is not quite clear were the border is here.
@DavidCarlisle yes, sounds so. He also deactivated for some time the forum.
 
@UlrikeFischer yes I know. But here that's really just a side issue, the definitions are just different so you can't really resolve that, it seems trivial to make scrbase definitions win if that package is loaded.
 
7:20 PM
@DavidCarlisle yes, but imho it would be even better if koma would remove all this definitions (there are lots more) and refer to etoolbox and iftex and expl3 ;-).
 
@UlrikeFischer I think Marcus has had this for a long time, and wants to stay in control
 
@UlrikeFischer I need an account to comment at komascript.de/node/2262 I think?
 
@DavidCarlisle yes, but currently no new accounts can be created: komascript.de/node/2263 (and I don't have an account either, I always sent mails).
 
@UlrikeFischer OK I'll leave it then, Markus could respond on the iftex issue.
 

« first day (3291 days earlier)      last day (1625 days later) »