@DavidCarlisle On top of deprecating RevTeX 4.1 and breaking all documents, reference formatting is broken in RevTeX 4.2...

I'm pretty sure there should be a space between the first author and “and”.

I already emailed APS about it. Here is a patch.

--- apsrev4-2.bst	2019-01-12 11:30:04.000000000 +1300
+++ apsrev4-2.bst	2019-01-17 14:26:15.364510366 +1300
@@ -1571,7 +1571,7 @@
                     emphasize
                     *
                   } {
-                    bbl.and
+                    word.space * bbl.and
                     word.space * *
                     t *
                   }

Hi, is there some way of locally adjusting the space between the two lines in, say:

\node [font=\bfseries, align=center] at (0.2,-6.8){foo\\ bar};

That's TikZ syntax, I guess.

And also, why doesn't this split across lines?

\node[fill=white, shape=rectangle, minimum width=4.45cm, minimum height=0.5cm] at (8, -15.6){foo\\ bar};

I guess in both cases this could be handled by positioning the two pieces of text separately, but that's not a great approach.

@FaheemMitha You forgot to add align=something, that's why it does not split.

@TeXnician Oh. Why does that matter, though?

@FaheemMitha See TikZ manual: "You use \\ inside your node to mark the end of lines and then request TikZ to arrange these lines in some manner. This will only be done, however, if the align option has been given." (p. 228 of the current version).

@FaheemMitha You could use the optional argument \\[1em]…

@HenriMenke ah Mark Doyle mentioned he had to push out a bst fix (we were in discussion over teh 4-1 breakage) Karl has put 4-1 back into texlive

@TeXnician Ok, I'll try that, thank you.

@TeXnician I see - some TikZ specific thing. I thought it was some basic TeX thing I was missing.

@FaheemMitha Well, TeX cannot break lines either if it does not know where to align (how should it determine where to start placing the next line and where to end it?). The thing is that you won't notice this as there will be some predefined alignment in place when writing "ordinary text".

@TeXnician So inside a node there is no predefined alignment?

@FaheemMitha No, it's the same as if you'd do \hbox{a\\b} in normal text…

@TeXnician No, as in - no, there is no predefined alignment?

@FaheemMitha Yes, exactly ;)

@TeXnician I see. Thank you for the clarification.

And the explanations.

No, actually, I think there is something else I'm missing. But I can come back to it later.

@FaheemMitha actually the alignment is really a secondary thing, the main thing to keep in mind is that tikz nodes are either an \mbox or a minipage depending on whether a width is specified, so in the minipage case you can do multiline things like \\ or \raggedright or \[display math\] but in the \mbox case you can't.

@DavidCarlisle Specifying a width is not sufficient (that would be a \hbox to width which would not break lines) but you have to specify the alignment as well.

@TeXnician so to get a default justified minipage?

 A man in a movie theater notices what looks like a duck sitting next to him. "Are you a duck?" asked the man, surprised. "Yes." "What are you doing at the movies?" The duck replied, "Well, I liked the book."

@DavidCarlisle To be able to use line breaks. Without specifying an alignment (I guess default alignment is left but I don't remember exactly) even specifying a text width will not enable line breaking.

@TeXnician vvvv

\documentclass{article}

\usepackage{tikz}

\begin{document}

\begin{tikzpicture}

\node[fill=yellow]{one two three four};

\end{tikzpicture}



\begin{tikzpicture}

\node[fill=yellow,text width=1cm]{one two three four};

\end{tikzpicture}
\end{document}

@DavidCarlisle Oh yes, I always thought about manual line breaks using \\ …

@TeXnician this is a 1cm minipage with the default (unnamed) tex justification of both margins

@TeXnician \\ would work in the second case as well

@DavidCarlisle hmmm so you are really a TikZ expert.

The more you know...

@PauloCereda I Have a gold badge

@DavidCarlisle ooh

@DavidCarlisle Does the Italian bloke have it?

@PauloCereda are you following the luatex list (reference our recent chat on google chat:-)

@PauloCereda s/have/stole/

@DavidCarlisle Yes, quite amusing. :)

@DavidCarlisle Oops, I should test with some real code next time…

@DavidCarlisle People really do take standards and recommendations into account!

@PauloCereda well given that the code under discussion is segfaulting, it's not a good time to ignore compiler warnings:-)

@DavidCarlisle uninitialized variables? Nah, what could possibly go wrong? :)

@DavidCarlisle on a serious note, definitely. :) It's the kind of issue that can be caught at compile time.

@HenriMenke tlmgr this morning:-) [19/25, 02:37/03:14] auto-install: revtex4-1 (49734) [3392k] ... done

@DavidCarlisle can we blame you: tex.stackexchange.com/q/470502/3929

@daleif it's a documented feature, so blame the OP for user error, not me:-)

@daleif ooh let's do it

@PauloCereda if I can answer enough questions like that perhaps I'll get a linguistics gold badge as well, and annoy @AlanMunn

@DavidCarlisle Thank god. This would have been a disaster if TeX Live 2019 shipped without RevTeX 4.1. The site would have been flooded with physicists thinking it's the apocalypse.

@DavidCarlisle ooh

@DavidCarlisle for all your picture mode answers?

"I find this a nice feature but it is not according to the documentation.
Or is it a BUG?"
"Let's call it an accidental feature. :-)"
             -- Larry Wall

@DavidCarlisle ^^

@DavidCarlisle There is documentation ... gasp

@daleif OH MY

@daleif @DavidCarlisle writes documentation for a living; which may explain why he does not advocate reading it.

@samcarter Thank you. I found \Big there. Seems I have to write it manually and write a fix for the package, where \Big is defined. AMS...?

@PauloCereda ^^^^^

@egreg ooh

@PauloCereda I'm much more expert in TikZ.

We use \ifnum or ifthenelse, when we have two conditions. Means (just for an e.g.) like $a =mod(2)$. When a=1, its odd or when its even for a=0. Then we can say \ifthenelse {a=1}{\node[black,below] at (\x-0.3, \y+0.3) {$odd$};}{\node[black,below] at (\x-0.3, \y+0.3) {$even$};}. What if we have more than two results, I'm struggling. like $a =mod(4)$?

Is there a way to say that \if {b=0} {...} and elseif{b=1} {..} ...

@DavidCarlisle I recall you once saying you answered TikZ questions with picture mode.

@FaheemMitha That was @DavidCarlisle.

@egreg Yes, I was responding to the wrong comment, sorry.

@Krishna You could try a real Linux in a Virtual machine like virtual box. Should be several orders faster.

@TeXnician @DavidCarlisle Since there seems to be some confusion about the line splitting thing, maybe it's worth asking a question about? Or even two?

@FaheemMitha No, there is no confusion. I have mixed some things up and the manual is a bit imprecise…

@TeXnician Sounds like a reason for a question.

@Shamina The face in your avatar. Is that Alizee?

@FaheemMitha haha yes!

@Shamina Ok. Just checking.

@FaheemMitha Do you know btw how to handle multi condition case?

Today I learned the correct name of the # symbol.

@Skillmon Octothorpe?

@egreg :) Yes.

@Skillmon THE SLACK LOGO

Not really :)

@PauloCereda oh slack... But octothorpe is correct. (which is quite funny as the English dictionary in my browser doesn't know it...)

@Skillmon I love the word too. :)

@Shamina as far as I know there is no elseif, you have to nest

@daleif Thanks! Is there a way to do that?

@Skillmon define "correct", Unicode calls it NUMBER SIGN for example

@Shamina If you mean in TeX, no.

@Shamina your question isn't at all clear as \if {b=0} isn't TeX syntax even without the question about ifelse but you can use \ifcase#1 a\or b\or c\else zzz\fi for example which will print a if #1 is 0, b if it is 1, c if it is 2 else print zzzz

@FaheemMitha Oh no! We can't do it in TikZ then. I have a multiple condition plot. Which will need these conditions. Means when a=1, then this and when a=2, then this and when a=3 then this

@Shamina I meant I don't know how to do it. But I'm barely a TeX beginner.

I'm sure one can write conditionals in TeX.

@DavidCarlisle Sorry. I was just giving an example, to show my problem

@Shamina Is the delegation of such calculations to an external source a possible solution?

Good afternoon everybody of the users into chat.

I can make it clear, just to repeat. Means when a=1, then this and when a=2, then this and when a=3 then this

@PauloCereda Hola' :-) my best and big huge.

@Sebastiano Aloha!

@Shamina as David already wrote, you can do it with \ifcase and similar commands.

@PauloCereda Didn't get it completely ;(

@UlrikeFischer Thanks. Let me check

@DavidCarlisle I guess, instead of $a$ we can put like \node[type\type] at (\x,\y) {\color{white}$a$}?

@Shamina probably ifcase as I showed above will do what you want then

@JosephWright Thank you very much for your recent contribute and for the refined response you gave to the user on Meta. Thanks for deleting the bad stuff I wrote almost two years ago if I remember correctly on Excel2LaTeX. I kindly wanted to ask you for a kindness. Why don't people get a downvote when they edit their answers and I systematically get a downvote after improving and integrating my previous answer?

@DavidCarlisle where \type can give #1, #2 #3 or #4

@PauloCereda Here the climate change in Sicily is evident. In my opinion, penguins and polar bears will move to Sicily :-)

@UlrikeFischer Thank you for your comment on the issue of not installing MikTeX. Luckily I didn't have to format the pc. :-)

@StefanKottwitz Hello and welcome moderator. Even if I forwarded a message to your colleague Joseph Wright could you give me a suggestion, or an explanation why when I improve even a little a question I get a downvote? In my opinion he's the same user.

@DavidCarlisle Last night, when there was a vote of confidence/faith in the House of Commons about Teresa May, I thought of you. I wonder if you were also on television watching live.

@Sebastiano I also think that polar bears move there. The ice is melting and houses seem to very affordable. ;-)

edition.cnn.com/travel/article/…

@marmot Ahahahahah, dear marmot, but where do you find these links? ahahahh. In my opinion at such a low price will be the ruins. It is true that the cost of living in Sicily is below average but you live very badly. I can assure you. Almost non-existent services, hospitals with poorly qualified staff, etc. ... Maybe I heard a bear knocking just now? What do you think about I'll let him in or eat me? ahahahh

@Sebastiano If downvotes or comments are related to the content and are somehow technically reasonable, we hardly can start tracking the user and do something about the account, for what reason.

@Sebastiano I was just kidding. (I actually liked it when I was there, but of course I didn't have to use any of the public services.)

ooh hi Dr. @StrongBad! :)

(")>/

@Sebastiano If there's objectively bad stuff happening, we can delete it. If it happens often, we can suspend a user. If it's still piling up, it can be escalated to the company employees that run the site.

@StefanKottwitz In the meantime, with politeness and respect, I thank you for your response. This afternoon I changed a link that didn't work and I edited my answer integrating it with another GUI that creates XCircuit electrical circuits. Could you check if it's okay? tex.stackexchange.com/questions/26972/…

$ sudo apt remove united-kingdom
[sudo] password for eu:
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following packages will be REMOVED:
  united-kingdom england wales scotland northern-ireland
0 upgraded, 0 newly installed, 5 to remove and 0 not upgraded.
After this operation, 1 GB disk space will be freed.
Do you want to continue? [Y/n]

@DavidCarlisle @JosephWright ^^

@StefanKottwitz We discussed this very kindly but the core of my thinking is: but why am I always subject to criticism and strong contrasts from some user? I confirm with my heart what I have always written: I do not care about the score but I do not like the evil and especially "hurt others" unjustly. I hope the translation with Deppl translator is good.

@PauloCereda but I have a secret plan, pull out the plug: it stops the system clock and causes a reboot (or well sometimes it permanently wrecks the machine, but a risk worth taking if everything else failed)

@Sebastiano when you edit an answer, it jumps to the top of the front page. It means people are more likely to see it and if they think it is bad, down vote it (or if it is good up vote it).

@DavidCarlisle l3build-upload.lua:189: The field author is longer than 128 ;-(

@DavidCarlisle ooh we could force with systemctl poweroff -i

@UlrikeFischer how many people did you list? (don't blame me they are ctan's limits)

@PauloCereda No I think someone may have installed a rootkit, I don't trust system commands, the plug is a safer bet.

@StrongBad Hi, I'm doing my best.

@DavidCarlisle en.wikipedia.org/wiki/Pedro_I_of_Brazil first emperor of Brazil, check his full name :)

@DavidCarlisle ooh

@DavidCarlisle I added the whole list from ctan.org/pkg/luaotfload (including the "inactive" text).

@DavidCarlisle you might want to go with a big magnet (or hammer)

Dear users, I feel cold, I'm going to go make myself an infusion. I greet all of you. A big hug.

@PauloCereda That explains why he never uploaded anything to CTAN

@DavidCarlisle Exactly!

@Sebastiano Good night

@UlrikeFischer for the oberdek bundle, if I use the web form, it auto-fills in the description and then rejects the submission as the description is too long so I always have to open out the "all fields" then delete it and write ... as before...

@DavidCarlisle do you have a suggestion what I should add in the author field? "unchanged" "see catalogue", a few names and then ";et al"?

@DavidCarlisle I think I write "... as before..." to and add a note.

@Sebastiano I think you misunderstood climate change. Just because it is unusually cold at this time of year in Sicily this is not a sign of climate change. And it is no proof, that there is no climate change. Weather is not climate.

@DavidCarlisle from an example in a german group: \def\linewidth{0.3\textwidth}.

@DavidCarlisle Is there an easy way to calculate the width of a D column? E.g. if I have a D{.}{.}{6.2} how can I determine its width?

@DavidCarlisle Do you know what is wrong here: ["WARNING","License not found","GNU General Public License, version 2"]?

@UlrikeFischer Comma, perhaps?

@UlrikeFischer I think it needs to be gpl2 as that is what the upload form uses:

<option value="gpl2"
            >
            GNU General Public License, version 2
           </option>

@PauloCereda no, it looks as if I need the internal shortcut (found in the source code of the ctan page (gpl2).

@UlrikeFischer ah :)

@DavidCarlisle yes just found it.

@AlanMunn it is the width of $123456.12$

@UlrikeFischer @JosephWright should have documented that in l3build doc:-)

@DavidCarlisle Ok. That's what I thought. Just wanted to check I wasn't missing something. Thanks.

@UlrikeFischer interesting

@DavidCarlisle after getting told that it is not a good idea the code was changed too \newlength\bildbreite \def\bildbreite{0.3\textwidth} ;-)

@UlrikeFischer we could validate that list in the lua before submission, as we do for lengths, but if it changed you would have to change l3build to match

@DavidCarlisle well it is not so difficult to find the shortcuts - if one realize that one need them ;-). So some documentation would be good.

@DavidCarlisle but both uploads were successfull, let's see if ctan complains ...

@ChristianHupfer Hi, :-) I have written in sicilian language. Here we mean climate as a synonym for time or weather.

Hi, can a ask a quick question. I just have the final version of my PhD thesis pending approval. Upon approval, I'd like to archive the entire code that reproduced the final PDF, including style-files, engine version, font files

How can I do that?

The snapshot package does not seem to be updated since 2002? I am using luatex. I am not sure if the snapshot package will work with my document

@Krishna that's impossible, basically. You can list all used files using \listfiles, but you will still miss engine binaries, fonts etc.

@Krishna copy the whole texlive tree, including texmf-var, copy all fonts mentioned in the log

@Krishna especially as you are using luatex (the least stable of all the tex engines) you can only guarantee reproducibility if you archive the entire tree including binaries. Although for most practical purposes you'd expect that if you just archived your actual thesis document it would work to produce something reasonable in any future tex system

@DavidCarlisle There are some things breaking between luatex 1.09 and luatex 1.07 . I can't afford that!

@UlrikeFischer and binaries will likely stop working at some moment anyway. The only option would be to preserve the whole computer, but even that's not a surety.

So, arlatex, bundledoc, snapshot - all useless for my case, huh?

@Krishna well, they all have ever been useless, IMHO :-/

I am asking about real-arxiving because, between the time my PhD thesis got submitted, viva'ed and corrected, 3 different packages had changed behaviour

I remember the diffcoeff package just broke with a major update

I'm not sure why you wanna do that. It makes sense to want to preserve the working environment until the final theses is defended, but that's just it.

@Krishna well, there is a reason why I was not updating my TeX distro for over 2 years of my PhD studies...

It was typesetting pdes without producing a rvert to indictae evaluation conditions.... My examiner went red in the face

Yeah.....my supervisor asked me nicely if I could give him the TeX code for the benefit ofthe group

my research is meant to be open.

So, I said yes. No problem

But when a new student joins in 2020 and wants to run my code, their TeXLive would contain updated packages which most likely breaks a few macros

@Krishna well, it shall be understood nowadays that a LaTeX code that uses many packages will not be 100% working after some time.

Yeah....I agree....

But I'd like to archive it.... Show to my nephew, niece and my kids when they grow up :)

@Krishna I believe you are overthinking this.

@Krishna Using too much packages is a guarantee of breaking sometime in future.

No....the diffcoeff update broke my PDF between the time I got viva-ed and the time corrections got submitted

Not if I populate the folder with the .sty files as of today right?

The local sty files will be prioritised by the engine over that installed in the texmf-dist, am I wrong?

@Krishna as I said, never update packages in critical times :-)

@Krishna it will.

So, one good idea to minimize changes is to populate today's sty files

Already, I have the fonts (libertinus) in the right folder

within the project

As Ulrike said, copying the whole texmf tree is overkill...... I will just copy all the sty files of the packages into the project's root

@Krishna You're forgetting that the LaTeX kernel changes from times to times as well, or font definitions etc, as well interdependencies between packages. Copying just packages A.sty, B.sty etc. is not enough, if A.sty loads Z.sty...

@Krishna the problem is that you are using luatex, and it loads lua files and some of them aren't even mentioned in the log.

@JosephWright For the first time I've seen three misses ending a frame in snooker. (Ding vs Brecel, Ding losing 4-5 after this mistake, played to 11 frames.)

@yo' Oooh

@Krishna things have broken every time the luatex version number has changed.

@Krishna sure but in practice your fully archived tree isn't going to be a lot of use. If in 10 years time someone wants to use your thesis code, it is likely to be much easier to just take the tex code of your thesis and make whatever macro patches are needed if necessary to make it work with packages of the time than it will be to set up a virtual machine with all the right system libraries to run your fully archived binary texliv etree.

@Krishna if you are arxiv with hundreds of thousands of documents that you really don't want to edit for package updates, it makes sense to preserve the entire virtual machine so you can re-run the entire corpus, but for one document even if it is a large thesis, it is unlikely to be a real benefit

@DavidCarlisle got it. Thanks a lot for explaining

I'd like to report a possible behavioural change with luatex 1.08 and 1.09 with respect to 1.09. If a tabular* is set to 0.7\textwidth with some footnotes in a threeparttable, then the table width that gets typeset is different enough to cause one of the table notes to move to the next row

@Krishna that sounds far more likely to be a change in a latex package than the change in luatex

Nope...it's just luatex

I build it on my local machine with TL2018

and then push it to travis, which builds it with TL2019/dev (luatex 1.09) and then places that version into dropbox

@Krishna oh OK, but to report to luatex list you'd have to cut that down to a simple plain tex example, and likely to get the answer that luatex development doesn't guarantee compatibility and it's not a bug. but...

Please ignore the changes in the citation numbers for now. I can confirm that even with the same citation numbers in [], stuff broke

Yeah....but I have to make someone aware, right?

@Krishna .. while i was typing that you posted the image and the footnote numbers are differemt colours

I don't have the ability to do these kind of things

Please don't attribute the difference to all these

This is becauase, I am working on the document for online upload

@Krishna not really, you chose to build on luatex and the luatex developers explictly do not guarantee the kid of stability that you get from classic tex, so things changing when the version change is expected behaviour

Wheras the first one was produced for my defense (viva)

and I didn't want to pay for colour, and colored hyperlinks when printed in B&W could become distorted, I forced Black & White for hyperref colours......anyway, the point is

1. I used to build the document locally, see the table typeset according to what I want

@Krishna well certainly I wouldn't even consider comparing line breaks unless you make two version using identical tex macros and identical options. you have used color and using or not using colour is highly likely to change spacing

No.....Aaargh...It's a misunderstanding due to these screenshots

I was lazy to report the bug back then, and was hurrying against a deadline to finsh the thesis that I didn't bother

@Krishna yes I believe you can make an example using the same macros but the above isn't it and as I say you'd need to cut that down to a plain tex paragraph if you wanted to report it to the luatex list.

This was/is my workflow. 1.) Build the document, using TL2018 locally, get a satisfactory table 2) Push to github, travis starts building with TL2019/dev, and the table linebreak gets messed up .....for the same identical source code

@Krishna why do you assume it's a bug?

I never said, bug.,.....I said "changed behaviour"

for the same source code

between LuaTeX 1.08 and 1.09

@Krishna .........................................^^^^ :-)

Okay...sorry about that

"I'd like to report a possible behavioural change with luatex 1.08 and 1.09 with respect to 1.09." was my original statement

Okay, anyway back in August 2018, you were interested to know if something changed

I started noticing this in September 2018, but then my submission deadline was Oct 3, 2018

So, I didn't report and lived with it

@Krishna basically it's likely to be a lot of work to make an example, and the response is likely to be just that change is expected as they changed the linebreaking code,....

Okay...cool ...no issues

I don't have a problem with the behaviour

Just saying.....just in case, it piques your interest or someone else's

@Krishna it depends. every luatex release we need to run lots of tests to make sure lualatex isn't completely broken, but if it's making reasonable but different output then it moves down the priorities (you were asking the other day about balancing time and real life...:-)

yeah..

But should this be in even in the back of the list is the question

@Krishna even with interest. Without concrete code to test one can't investigate.

As in, should this be recorded?

I am happy to give the code

I can invite you to my private github repo

and point out the specific line where this happens

@Krishna you've been on site long enough to know a thesis repo isn't an ideal debugging test, can you not make a one-table example file?

:)

Of course, I can

@Krishna Can you turn it into a truly minimal example?

Well....the other day, I was pointing to you about motivation.......

I am not sure, if I have the will-power ....:)

I may do it...please don't judge me....I don't mean to be rude

but it's just that the thesis took me to breaking point

I shall try to do a minimal example....no worries

@Krishna I don't think anyone cares so much one way or the other (so you shouldn't worry about seeming rude.) It might be helpful for the luatex folk to know, if it is indeed a luatex problem, but you shouldn't feel obliged to report it either. But if you do want to report it, then it is really only helpful if you also construct a MWE.

@AlanMunn Ok, that's very clear. Got it. I cannot make a plaintex example for the lua folks mailing list, but I sure can provide a clean, latex-MWE here.