12:50 AM
@DavidCarlisle but a normal person would find it in the CTAN page -> amsmath package...

1:38 AM
Quick question: Is it possible to use \enumerate list items but with two columns instead of one? Or would you need to use something different?

2:10 AM
@Quintec I don't think so. See the tasks package and use it as \begin{tasks}(2) \task Text \task Text...\end{tasks}
The "(2)" means "two columns"

Thanks :)

@DavidCarlisle that commit is for me? lol

4 hours later…
5:54 AM
good morning guys, how to configure texstudio every time I click on a file to run as an admin? this issue is causing texstudio to crash.

6:08 AM
So either my last question was very obviously stupid, or my suspicion that nobody understands how bibtex's internals work is correct.

6:19 AM
@Maeher Huh?

@DavidCarlisle nope. Mine works without tokenizing, so you can write a macro \foo with \verb in its argument.

@Maeher Oh, the one about MACRO and .bst: those are substituted into the .bib material before the .bst ever sees them

@JosephWright Which I suppose means that there is no way to actually access those values in the .bst (and in extension that there is no way of passing global (as in not belonging to an entry) information from the .bib to the .bst)

@Quintec with multicol you can do \begin{multicols}{2}\begin{enumerate}...\end{enumerate}\end{multicols}.

@Maeher You can pass info to a ,bst by using a special citation which has dedicated fields: see for example IEEEtran, achemso or biblatex

6:26 AM
@CroCo running arbitrary binaries with admin rights is not a good idea. For everyone (ok, perhaps not everyone) else it's working without admin rights. You should focus on the issue causing this rather than solving it by granting privileges.

@JosephWright Right, I didn't think of that. Thanks.

@Maeher I think IEEEtran is the originator of this idea: I got it from there for achemso

@Skillmon but why if I run texStudio as an admin I have no issue but if I click on a file.tex, texStudio crashes. It seems texStudio needs admin privileges to communicate with Miktex. This issue occurs with the new update of texStudio.

@Maeher The trick is to write a special .bib files with one entry that you automatically \nocite. The special citation has fields with names that are dedicated, like ctrl-some-aspect, which you can then convert to a variable in the .bst.

@CroCo I don't know. I don't have access to your PC, personally am neither using Windows nor TeXstudio, so can't help you any further except saying you that I know it does work out for others without admin rights.

6:32 AM
@JosephWright Thanks, I'll have a look and think about whether that's applicable in my case.

6:49 AM
@Skillmon so lrbox environment then (that was one of main reasons for adding an environment form of \sbox)
@manooooh no. Why do texlive and miktex go to the trouble of installing documentation on millions of end users machines, if "normal" users ignore that and then trawl the web to find the same files on a remote server?
@KhaledHosny ./build.sh at the top level failed to find or make a source/configure for me, I tried to copy in some bits from my checkout of the github texlive-source repository but it was gone midnight and I gave up:-) I'll look again at the weekend but did you have something missing (I've compiled both full texlive-source and the luatex-specific luatex experimental branch on this machine before so I think I should have all the necessary tools installed (on cygwin))


@Skillmon no, but verb not working in macro arguments is a design feature, see the repeated refusals to make verb work in \footnote even though it works in plain-tex \footnote or to make it work in \mbox which could have been defined as \def\mbox{\leavevmode\hbox} but is defined as \def\mbox#1{\leavevmode\hbox{#1}} and more or less the only effect of that #1 is to freeze the catcodes.

@DavidCarlisle \boxgrab currently has the syntax \boxgrab<register>[<inject pre contents>]<hbox/vbox>[<inject post contents>]<next>.

@Skillmon if we keep chatting long enough @JosephWright or @egreg will pop up and mention xparse v arguments:-)

@DavidCarlisle Of course

7:03 AM
@DavidCarlisle It might be a design feature, but it bothers me sometimes. And grabbing something as a box is handy if you don't have to parse its contents but only need the size of the box anyway.
@DavidCarlisle which can't be nested, too.

@Skillmon sure, but no verb macro can be nested (without re-implementing things with \scantokens or whatever

@DavidCarlisle but box macros can :)

@Skillmon no, that's what I mean, if the reason for not using lrbox syntax is so that you can accept verbatim arguments, then that works but means that your macro becomes verb-like, and doesn't work in the argument of another command (or at least doesn't accept verbatim). \mbox is defined as it is so that it works the same way in all contexts, the simpler definition would mean that \mbox{..\verb|\foo|} worked or didn't work, depending where it is used

@DavidCarlisle yes, that's true, but in that case it would have been used in an ordinary argument, which is not my fault. ducksay does this with its box type argument parsing, too. And I like the fact that I don't impose further restrictions upon users where it's not necessary.

7:25 AM
@DavidCarlisle Which don't work in arguments to other commands. ;-)

22 mins ago, by Skillmon
@DavidCarlisle which can't be nested, too.
@egreg :)

7:43 AM
@DavidCarlisle See the build instructions in the README,md file

7:54 AM
@KhaledHosny reading instructions? It will never catch on:-) it was rather late to be reading so ./build.sh was all I could manage, I'll try again when I'm awake thanks:-)

@DavidCarlisle You basically need to run ./reautoconf from source, since I dropped the generated build files from the repo (they make reading the diffs extremely hard when they are modified).

8:29 AM
@KhaledHosny yes thanks it ran and main build is now building something, I probably can't look at it until this evening but it's doing something in another window at the moment:-)

1 hour later…
9:30 AM
Is it possible to keep braces in an O type argument? Consider \NewDocumentCommand \foo { O{} }{...} with usage \foo[{arg1}] will have #1=arg1 without the braces, while \foo[{arg1}{arg2}] was #1={arg1}{arg2}. How could I get #1={arg1} in the first and #1={arg1}{arg2} in the second case?

10:05 AM
@DavidCarlisle wow I am sorry,I did not know that they were different. So is amsmath package different installed on texlive than miktex?

10:16 AM
@manooooh the files are the same, that's what I meant. I was commenting on your assertion that normal users would use the files from ctan rather than the files they have installed on their machine. I don't see why they would do that necessarily.
@Skillmon I suspect you can't: that brace stripping happens at the lowest level in tex's argument scanner,

@DavidCarlisle \futureletting through it could work, but I don't think I'll do that...

@Skillmon When TeX looks for an argument, it automatically strips off a surrounding pair of braces, provided this doesn't leave unbalanced braces. So {arg1} becomes arg1 and {arg1}x remains the same as well as {arg1}{arg2}.

@Skillmon at a lower level (and giving up some error checking) you can define \def\foo#1]{...} then #1 is [{arg} so you just need to \@gobble the first token and you have {arg} so there are things you can do, but they all avoid passing {arg} to tex's low level argument scanner

@egreg yeah, I know, but I wondered whether I could somehow prevent that.

@Skillmon {{arg1}}

10:26 AM
@egreg he means a definition that doesn't require the user to do that.

@DavidCarlisle while this would certainly work (great idea btw.) it doesn't check for matching brackets like an O type would.

@Skillmon I know (you can't have everything:-)

@DavidCarlisle I think I'll stick with the user has to protect it if it's only one braced arg.

11:07 AM
Hello @Skillmon ;)

@Skillmon That's deliberately not supported: classical LaTeX2e requires for example [{\foo[aaa]}], but we don't in xparse so want the outcome to be the same: [\foo[aaa]]

Does anybody remember what is the first TeX Live version to run latex with the e-tex engine by default? And when was \unexpanded introduced to e-tex, if it hasn't been there since the beginning? (I ask because I am trying to come up with a good help text in case \unexpanded is not defined.)

@HaraldHanche-Olsen \unexpanded was in v1 of e-TeX, so mid-1990s

@JosephWright Thanks, good to know. Bonus question: Is there a source repository or other place where I could easily discover such things for myself? I get a ridiculous number of irrelevant hits when I try googling.

@HaraldHanche-Olsen tug.org/svn/texlive/trunk
@HaraldHanche-Olsen SVN goes back to 2005 ....

11:21 AM

@JosephWright Ah. So I can safely say their TeX distribution is seriously out of date if there is no \unexpanded. Thanks!

@HaraldHanche-Olsen The only case I can see where that might not be the case is for things like PCTeX or server-based set-ups (journals and the like), but even then it's extremely unlikely today

@JosephWright so, generally the stuff provided in brackets should be given to the last argument of \boxgrab as is.

@JosephWright Oh my, how time flies …

@cis hello :)

11:37 AM
@JosephWright Um, I was thinking about the e-tex sources.

@JosephWright I wrote some recursive macro that works: 1. Checking for an opening brace (but not removing it), 2. reading up until the next closing brace, 3. removing the first opening brace, 4. checking for another opening brace, 5. if none is found we're done, 6. else put the closing brace back into the read information and read till the next closing brace (recursion to 4).

@JosephWright Oh, wait, I just found out: “The de­vel­op­ment source for e-TEX is the TEX Live source repos­i­tory.”

@JosephWright "currently working on tagged PDF": we should start to make some concrete plans ...

11:55 AM
Exploring the TeX Live repository through the web interfarce (← inadvertent typo, but it's too good to correct) is just incredibly clunky. Is it available for cloning from somewhere, or is that reserved to TL developers?

@UlrikeFischer Yup
@HaraldHanche-Olsen You can do an anonymous SVN checkout
@HaraldHanche-Olsen I use a checkout for doing engine fix stuff ...
@UlrikeFischer I have a few other things to do, then back to the driver support stuff. If you work on the 'visible' part, we can close in on each others positions ...

@JosephWright fine. I have next week a talk about it and I'm currently making a list of "open points" while preparing it. Perhaps we could then speak about it in november.

12:13 PM
@UlrikeFischer Sounds good

3
@JosephWright ^^^I even have a shirt!

@JosephWright But from what URL? tug.org/svn/texlive/trunk only provides a web view, as far as I can tell.

/duck

@cis I think this only works when @Psmith is in the room.

12:30 PM
Where is @Psmith? @PauloCereda give us someone to talk to.

12:41 PM
@JosephWright highlighting that @egreg has more rep than he deserves:-)
@HaraldHanche-Olsen svn where Joseph just showed or git for the main sources at github.com/TeX-Live

@DavidCarlisle Ah, maybe I can make do with the git repo. Thanks. … Oh wait, git svn clone svn://tug.org/texlive/trunk texlive works too.

1:10 PM
@JosephWright More people to blame!

@HaraldHanche-Olsen that's @DavidCarlisle's text ;-)

@UlrikeFischer I am learning from him, of course.

@DavidCarlisle so if the files are the same, then both files should include the same data (maybe not structure, and that's probably the point), but it's not like that. In CTAN there is a section 4.1 but in the other there is not section 4.1

1:38 PM
@manooooh no the files are exactly the same. I ran latex on them on my machine and copied the pdf to ctan and texlive copies the pdfs off ctan and distributes it to users. amsldoc.pdf (not amsmath.pdf) has a section 4.1 headed matrices, and that's what you get from texdoc amsmath

@manooooh the files are not the same. CTAN links to amsmath.pdf which contains the code documentation and not to the user documentation amsldoc.pdf, imho that is simply an error on the ctan page (@DavidCarlisle).

@UlrikeFischer oh but @manooooh was describing it the other way round, that ctan had section 4.1, but both files are on ctan and in texlive anway

@DavidCarlisle I never believe in descriptions ;-). So I checked ...

Jan 14 '15 at 21:56, by David Carlisle
@PauloCereda never believe what they tell you

Hey, how can I get Unicode characters, like \write{U+2654}? Is this a trivial question or a complicated problem?

1:44 PM
@cis depends on the engine and the unicode char.

@cis in pdftex? (harder) or luatex or xetex? (easy)

Aha. I know LuaTeX... jaja... I asked myself, is there a possibilty to get the unicode symbold by there "number ponits"?

@cis in luatex ^^^^2564 (or \Uchar "2654)

Mmmhhh....

2:15 PM
@UlrikeFischer thank you Ulrike! That's probably my problem, I cannot distinguish which file I am reading. What's the difference between a code documentation and user documentation? When you suggest read a doc to a user in this site, which one you pick?

@manooooh users should read user documentation. What else? And if you want to see the difference find amsldoc.pdf and compare it with amsmath.pdf. And write to ctan an tell them that the link on the catalogue site is wrong.

@UlrikeFischer if I remember I will do it. I thought they were the same but in different version, because I remember (I am on mobile) that the doc that contains section 4.1 were launched on... 2002/2003 maybe? And the doc in CTAN (not contains 4.1) were launched on 2014
I will look it better, don't trust me at all

2:36 PM
Wow, my git svn clone … command that I started almost two hours ago is still running. Guess I shouldn't be too surprised, TeX Live is a big hunk of stuff after all. Oh, it's at revision 7344 out of 48923! This will not finish today.

2:57 PM
@UlrikeFischer I think the Bär should also have a shirt:
4

@samcarter Oh yes. He was already quite jealous ...
@samcarter you got mail ...

@HaraldHanche-Olsen it would have been a lot faster from github, their servers are rather more designed for that than tug's...

3:14 PM
@DavidCarlisle Well, I on github I find only the source part. But it turns out I don't really need more than that anyhow, so I guess I'll interrupt that svn clone.

@UlrikeFischer Well, he has to understand that with his custom size it can take a bit longer until he gets his shirt :)

@samcarter we eat animals, not make them team members.

@DavidCarlisle now he is crying and hiding behind the sofa ...

Oct 10 at 14:47, by Paulo Cereda
You are mean

Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

3:31 PM
@DavidCarlisle Can @Psmith automatically insert such quotes? (@PauloCereda)

4:08 PM
@UlrikeFischer With texdoc chessboard I only get a readme of the package. I had to go to CTAN to find the documentation. Is it supposed to be that way?
@UlrikeFischer Actually, searching for chessboard in the TeXLive folder does not show any pdf...

@PhelypeOleinik on texlive is it like this -- until I find the time to clean up the sources and upload them. I should do it, I know.

@UlrikeFischer Ooh. Okay, I thought it was an unknown feature :)

@PhelypeOleinik no it is a well known feature to me -- everytime I want to look in the docu I have to remember to use miktex. But I don't dare to look in the sources know. Perhaps they no longer compile, or look so horrible that I spend hours to repair them... better not start it.

@UlrikeFischer hahahhahahah. I know the feeling :P

@JosephWright are there mnemonics for the test files endings? lvt= ?? tlg= test log? pvt= ?? tpf = ??.

4:23 PM
@UlrikeFischer The first two @DavidCarlisle is likely better placed to know: I think 'Latex Validation Test' and 'Test LoG', and so went with 'Pdf Validation Test' and 'Test PdF' as equivalents ...

4:45 PM
@JosephWright ok, lets hope that I know can remember this better ;-).
is kpsewhich --expand-path=$texmf the correct syntax to output the full pathes on linux? 5:20 PM @StefanKottwitz you probably know this ^^^ @UlrikeFischer do you mean something like this? kpsewhich --var-value=TEXMF gives on the TeXample server {/home/texample/.texmf-config,/home/texample/.texmf-var,/home/texample/texmf,/e‌​tc/texmf,!!/var/lib/texmf,!!/usr/share/texmf,!!/usr/share/texlive/texmf,!!/usr/lo‌​cal/share/texmf,!!/usr/share/texlive/texmf-dist} @StefanKottwitz a yes, that probably works too. Would --expand-path=$texmf work too for you?

@UlrikeFischer empty output for all path vars I tried, still checking

@StefanKottwitz well perhaps it is --expand-path '$TEXMF' on linux. But if --var-value works I will use this, then I don't have to fight with the dollar ... @UlrikeFischer yes, just found it: quotes are needed kpsewhich --expand-path='$TEXMF'
and it is case sensitive
I get as result: /home/texample/.texmf-var:/home/texample/texmf:/etc/texmf:/var/lib/texmf:/usr/share/texmf:/usr/share/texlive/texmf:/usr/local/share/texmf:/usr/share/texlive/texmf-dist


1 hour later…
7:02 PM
Do you think the name boxgrab is better or grabbox?

'ello, Psmith is now fully operational.

@DavidCarlisle you are mean

Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

@DavidCarlisle you are mean

7:04 PM
Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

/love

Skillmon loves me!

/help

'ello, I am Psmith, the TeX bot!

List of commands:
/8ball     Simulates a magic 8-ball.
/battle    Displays the battle of titans!
/cricket   Fetches cricket news from ESPN.
/duck      Displays a random duck picture.
/football  Fetches football news from ESPN.
/help      Displays this help message.
/learn     Teaches the bot a new command.
/roll      Rolls a variable-sided die or makes a choice.
/say       Makes the bot say something.
/shutdown  Terminates the bot (admins only).

/unlearn love

7:04 PM
@Skillmon Hurrah, command forgotten.

/learn love @Skillmon loves me!

@Skillmon Saved.

/love

@Skillmon loves me!

/8ball is grabbox a better name than boxgrab?

7:05 PM
@Skillmon Okey-dokey.

@Skillmon 'ello, I bid you good day! Type /help to see all my commands.

/borussia

/say

7:06 PM
He is content to order his movements in the tone of a glove-fight, a spine-shaking word-picture of a number.

/8ball is grabbox a better name than boxgrab?

@Skillmon Most likely.

Thanks, @Psmith, then I'll stick with that one.

@Skillmon 'ello, I bid you good day! Type /help to see all my commands.

/8ball do you like me?

7:09 PM
@PauloCereda I am not sure I can answer that.

oh
/say

New York, the magnet, had been the order of the ways

@Skillmon boxgrab sounds a bit like a funeral.

@UlrikeFischer I don't get the association. Mind explaining?
@UlrikeFischer oh, in German....

@Skillmon think german ..

7:15 PM
@UlrikeFischer but so does grabbox (sounds like Sarg)

/8ball Pizza with ham or with salami?

@cis It is certain.

@cis isn't he great! /love

@Skillmon not to my ears. I'm interpreting the first as a verb and the second as noun so grab a box is okay, but box a grab sounds odd.

@Skillmon Of course.

7:18 PM
/love

@Skillmon loves me!

Asking the final question (@DavidCarlisle @CarLaTeX @egreg):
/8ball does pineapple taste good on pizza?

@Skillmon Chances are lower than skynet.

So it is solved, once and for all!

/say

7:23 PM
I wished to consult his confidential secretary and adviser on some aspect of these determined spirits? Huh? Did they carol snatches of song as they've started, what it will be compelled to place yourself under police protection

7:35 PM
/8ball @Psmith Who was the murderer of J.F.k.?

@cis 'ello, I bid you good day! Type /help to see all my commands.
@cis Outlook good.

/8ball are you reliable?

@PauloCereda David is mean.

ooh
@DavidCarlisle ^^
@DavidCarlisle you are mean

Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

7:37 PM
oh

7:58 PM
@PauloCereda first useful thing your bot has ever done

@DavidCarlisle you are mean

Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

/pancake

8:01 PM
oh no

@PauloCereda don't blame me. It's your bot what did it.

@Psmith Very good. ;)

@cis 'ello, I bid you good day! Type /help to see all my commands.

You are mean
Oh no, the bot already left :(

/8ball is @samcarter mean?

8:07 PM
@DavidCarlisle It is certain.

@samcarter aren't you glad he's still here ^^^

@samcarter you have to address @DavidCarlisle I guess, to say that he is mean.

@DavidCarlisle, @egreg, @UlrikeFischer Simplified error messages from L3 code off to CTAN ...

@DavidCarlisle you're mean
@DavidCarlisle you are mean

Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

8:10 PM
Next try: You are mean!

@DavidCarlisle You are mean!
@DavidCarlisle You are mean
you are mean

@DavidCarlisle you're mean

@samcarter see, the string must be exactly matched.
@DavidCarlisle you are mean

Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

@DavidCarlisle you are mean

8:12 PM
Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

/love for his simplicity!

@Skillmon loves me!

@PauloCereda The bot is awesome!

@JosephWright trying to squeeze a sensible message in between the nonsense?

@DavidCarlisle Indeed: see the diffs ;)

8:13 PM
@JosephWright and my alias names ;-)

@Skillmon Figured that out, I just did not know which is the exact string :)

@DavidCarlisle oy

@samcarter Saved.

/realpancakes

8:14 PM

@samcarter those look great!
@egreg is it ok if I mention (and thank) you in the documentation of grabbox?

/shutdown

Initiating shutdown sequence. Cheerio!

8:32 PM
@Skillmon I'm not sure how I contributed. :-)
/shutup
@PauloCereda ^^^^

@PauloCereda I keep reading that as "Irritating shutdown sequence."

@egreg back when I wrote version 2 of the ducksay package you helped me with some stuff related to the mechanism I now use (though I got rid of the \let\futer=` to remove the opening brace). I think I don't use anything you helped me with back then, but if you hadn't helped me I wouldn't have further considered it.

@egreg he needs someone to blame in case of bug reports

@DavidCarlisle great idea, but no :)
@DavidCarlisle there will of course be no bugs, only hidden features (hidden so well, I didn't even see them).

9:02 PM
@egreg so is it ok if I acknowledge your help?

9:14 PM
+

9:35 PM
@Skillmon If you so wish

2 hours later…
11:23 PM
@DavidCarlisle @UlrikeFischer Here is a fork of luatex-harfbuzz that should theoretically work with luahbtex github.com/khaledhosny/luatex-harfbuzz, but it doesn’t as I’m getting some weird error in luatex-fonts-merged.lua. Just in case you managed to compile luahbtex and want something to try with.
I tried to cross-compile for Windows with mingw32, but my distribution packages for it are utterly broken.