TeX, LaTeX and Friends

Joseph Wright

9:07 AM

@DavidCarlisle Please remember the Doodle poll ;)

David Carlisle

@JosephWright how could I forget...

Joseph Wright

@DavidCarlisle Meanwhile, I'm getting the hang of branches, tags and switching in SVN :)

@DavidCarlisle Is there an 'easy' way to know what branches/tags are available in SVN other than checking out the entire repo? I presume there must be otherwise svn switch would be rather pointless.

@DavidCarlisle Thinking I'll wait for Frank's comments on $Id, tagging, etc. before doing anything

@DavidCarlisle Great: guess I'll prod Will next

David Carlisle

@JosephWright not really, unlike cvs or git svn doesn't have any primitive notion of tag or branch they are just the result of svn copy. Of course if you keep the convention of branches being in /branches rather than /trunk you can use svn ls ^/branches to list them all (the ^ notation is short for a relative svn: url so acts on the repository not requiring a checkout)

Joseph Wright

@DavidCarlisle Ah, that's close enough (I just had to remember to double ^ to get it to work here (Windows))

@DavidCarlisle I'm thinking if we do tag all of the L3 snapshots then I want some way to see them without checking everything out [though I'm hoping SubGit will make them all into Git tags :)]

Ulrike Fischer

10:04 AM

@DavidCarlisle I have minimalize the pdftexcmds error

\documentclass{article}

\directlua{
function sort_tdo (jobname)
file=io.open(jobname.."sorted", "w")
io.output(file)
io.close(file)
end
}

\directlua{
function shellescape()
if os.execute then
tex.write(os.execute())
end
end}

\directlua{sort_tdo("\jobname.tdo")

\makeatletter
%\def\pdf@shellescape{\directlua{tex.sprint(status.shell_escape .. " ")}} %works
\def\pdf@shellescape{\directlua{shellescape()}}% breaks
\ifnum\pdf@shellescape>0 %
\show\x
\fi
\begin{document}
\end{document}

David Carlisle

@UlrikeFischer thanks, so much of the lua side of pdfcmds needs updating.. I'll see what I can do can't think of lua at moment,: writing java to call xslt to read xml to generate C, thinking about lua at same time too much:-)

Ulrike Fischer

@DavidCarlisle Well the sort_tdo code is faulty imho too. It should really (as in your answer) store the old io-file handler. Should I add the example to some issue tracker so that it doesn't get forgotten?

Ulrike Fischer

10:25 AM

I added a tracker to ho-tex.

David Carlisle

10:38 AM

@UlrikeFischer thanks

CarLaTeX

12:08 PM

Hi guys, btw how does the TUG membership I won work? Do I have to wait that someone contacts me or do I have to do something? I don't want to miss the voting for TUG 2017 election! :):):):):)

Paulo Cereda

@CarLaTeX You will get an email soon. :)

David Carlisle

@PauloCereda demanding $100 to be paid for membership dues no doubt!

Paulo Cereda

@DavidCarlisle ooh

CarLaTeX

@PauloCereda OK, thank you!

@DavidCarlisle In that case I don't reply hahaha!

@PauloCereda However, I hope I notice the e-mail, because I'm signed in on TeX.SE both with the FB account and the "fake" gmail account that I use for avoid spamming, whereas I have my actual mail for the TeX.SE weekly newletter... a bit complicated :):):):):)

Joseph Wright

@CarLaTeX I guess I need to talk to the Powers: will do it now

CarLaTeX

12:24 PM

@JosephWright Thank you, perhaps my e-mail situation it's a bit complicated (see before)...

Paulo Cereda

@CarLaTeX Fake e-mail? Gasp...

Joseph Wright

@CarLaTeX I just pass the info to the Powers: they then contact you to get 'real' data I believe

Paulo Cereda

@JosephWright ooh powers

CarLaTeX

@JosephWright Thank you, my "real" e-mail is the one where I receive the newsletter...

@JosephWright but also the "fake" gmail is OK, I check it, too!

Sebastiano

1:12 PM

@ChristianHupfer Hi, Christian. I would need your help very important. I hope!!!!!

Hiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii to users LaTeX. Good morning everybody

David Carlisle

@Sebastiano aren't you in Europe? :-)

yo'

1:50 PM

Please, do you who is running LaTeXStudio GitHub?

@barbarabeeton I hope I got this right: github.com/tohecz/eqnalign/blob/master/eqnalign.sty#L47-48

barbara beeton

@yo' -- that looks plausible. however, \align@qed and friends aren't defined in amsmath, only in amsthm (and thus in ams document classes). so if one of those isn't being used, the \let shouldn't work. i don't see you testing for that, so maybe you need a \providecommand to avoid problems. (i'd need to test this all to be sure.) anyhow, take a look at amsthm.sty to see the definitions there.

yo'

2:05 PM

@barbarabeeton ah, thanks for this! Maybe I should \def rather then \let

I'll do some tests.

barbara beeton

@yo' -- should work, but i think you should also note in the documentation that using \qedhere won't do anything useful unless a "compatible" theorem package is being used.

yo'

@barbarabeeton which other "common" packages define \qedhere --- ntheorem probably?

David Carlisle

@yo' not \NewDocumentCommand ? :-)

barbara beeton

@yo' -- i'd have to check. i know there are some packages that have an amsthm option or a check for its presence. mdframed, for instance, checks whether amsthm is loaded; in ntheorem it's an option.

yo'

2:23 PM

@DavidCarlisle well, I'm not sure I need (or want) to handle expansion in any way.

@barbarabeeton you know, I could also wrap in \AtBeginDocument{\@ifpackageloaded{amsthm{...}{}}}

David Carlisle

@yo' oh I was just reprimanding you for using tex primitive syntax in a package. I'm sure I've never done that....

Artemis

4:08 PM

@UlrikeFischer Maybe it is easier to sort out some more short questions here?

Joseph Wright

4:47 PM

@CarLaTeX I've heard back from the Powers (@pops): he's on it

CarLaTeX

@JosephWright Thank you very much!

Paulo Cereda

@JosephWright ooh Pops

Joseph Wright

5:44 PM

@WillRobertson Don't forget the team Doodle poll :)

Paulo Cereda

@JosephWright ooh poll

Joseph Wright

@PauloCereda Team meeting

Paulo Cereda

@JosephWright Saturday, David's place, Coca-Cola and BBQ. :)

Sebastiano

6:18 PM

Hola', bienvenido:-) a todos

@PauloCereda Hi Paulo, Hi Joseph

@DavidCarlisle ahahahahahahahah

@DavidCarlisle hola' david. They have European dear david. But I wanted to make them understand that we had morning.

@TorbjørnT. hiiiiiiiiiiiiiiiiiiiiiii

David Carlisle

6:43 PM

@PauloCereda i doubt there's ever been any coca cola in this house:-)

Paulo Cereda

@DavidCarlisle oh no

@DavidCarlisle: I had horchata in Spain, it was delicious. Can I find some near your house? :)

David Carlisle

@PauloCereda compared to Brazil, Spain is quite close.

Paulo Cereda

@DavidCarlisle ooh

Christian Hupfer

8:36 PM

@Sebastiano Well.... go on, but I am not sure I can help at all...

Joseph Wright

9:42 PM

github.com/latex3/latex3/commit/…

David Carlisle

@JosephWright and I just needed a random number generator for integers:(

3

A: OpenType 'rand' (Randomize) feature not correct implemented in XeTeX?

I don't think +rand is supported however you can access all the variants by name (a.1 to a.31 in the case of "a") and then randomise in TeX eg using the PGF random function. \documentclass{article} \usepackage{fontspec} \setmainfont[Letters=Random]{punknova-regular.otf} \usepackage{pgf} \be...

Joseph Wright

@DavidCarlisle Yes, I'd seen that: is it in the dev version of XeTeX?

David Carlisle

@JosephWright don't think so, as far as I can see it's not in harfbuzz at all

@JosephWright benefits of writing a font shaping engine in a scripting language:-)

Joseph Wright

@DavidCarlisle I meant the random number primitive for the FPU :)

@DavidCarlisle :)

David Carlisle

@JosephWright oh did you mean the random number generator rather than +rand

Joseph Wright

9:52 PM

@DavidCarlisle Yes

David Carlisle

ah yes, Not sure (and it's a while since I built xetex so I'm just stock texlive 2016 for that)

@JosephWright "we" didn't and rand to expl3 yet did we?

Joseph Wright

@DavidCarlisle Huh?

\documentclass{article}
\usepackage{fpu}
\begin{document}
\LaTeX{} can now compute: $ \fpeval{rand()} $.
\end{document}

David Carlisle

@JosephWright I thought it did wasn't there discussion about possible algorithms, or was that just in context of adding it to xetex. anyway I used pgf today (I don't have a gold badge for expl3)

Joseph Wright

@DavidCarlisle Algorithms was partly about issues in the pdfTeX (MetaFont) code and partly about the fact one can do it expandably (but slowly) without primitive support ...

David Carlisle

@JosephWright meanwhile did you see the xparse keyval question?

Joseph Wright

10:05 PM

@DavidCarlisle No?

David Carlisle

0

Q: Is it possible to define pgf keys that accept values with an xparse-like syntax?

The pgfkeys package has the /.code handler that allows one to define the code to be executed when the user sets a key, as if it executed a macro with #1 as the value. The /.code args handler allows to provide an actual argument list in \def-style so one can define fancy keys, like rectangle/.cod...

@JosephWright ^^

Joseph Wright

@DavidCarlisle Ah

Max

Where can I find documenation for outputpenalty and its magic values in the kernel?

I perused source2e but it didn't really enlighten me

egreg

@JosephWright I see \NewExpandableDocumentCommand :)

Max

I have someone whose label get swallowed because he ends up in the dospecialoutput part of the output

why would his outputpenalty be set to -10001 ?

David Carlisle

10:17 PM

@Max outputpenaly itself is a tex primitive so the texbook (or I would guess tex by topic for free alternative) for the specific values used in latex, source2e (ltoutput.dtx)

Max

@DavidCarlisle I know how it works in general and the latter didn't yet clear up the problem

David Carlisle

@Max it's \clearpage

\def\clearpage{%
  \ifvmode
    \ifnum \@dbltopnum =\m@ne
      \ifdim \pagetotal <\topskip
        \hbox{}%
      \fi
    \fi
  \fi
  \newpage
  \write\m@ne{}%
  \vbox{}%
  \penalty -\@Mi
}

@Max any penalty less than -10000 forces the output routine but the specific values tell the output routine why it was called, -10000 is \pagebreak, -10001 is \clearpage, -10002 is a float in hmode, etc

Max

but doesn't that set the clearpage penalty to -10000 not -10001?

or is there a -1 I don't see?

David Carlisle

@Max no \@Mi is 10001

Max

ah sorry

misread that for \@M

thank you very much

David Carlisle

10:52 PM

@egreg ooalign, you need to do a horse now to go with the wagon wheels