any one can guess if the following was done using Latex? I doubt it myself, but can't think what software could do this if not Latex? This is a page from textbook called "Calculus" by James Stewart, publisher is CENGAGE learning. Here is the page

The math looks like it is Latex? But the overall design is very nice, and I do not see how they did all of this in just Latex. inDesign may be, but indesign do not do Latex.

@PauloCereda -- There are more strange things about that score ... I've never before seen the clefs reversed, and there are also some symbols that are quite new to me. (Thanks for the pointer.) There also seems to be an anomaly in the notes: the first selection is definitely a rag, and the second isn't. I think somebody wasn't paying attention when posting it.

@Nasser -- If you can find a page with a \sum that might give a stronger hint. The computer modern \sum doesn't look like any other currently in use, as far as I am aware, and it's very easy to recognize.

@barbarabeeton thanks. I will now look for such page. May be add one or two pages also.

The following are 2 pages from the same book. It has sums and also integrals.

I'd love to be able to do the above using Latex only. But if the above was done in Latex, then some real Latex expert must did all the above!

Fyi, I found an amazon link to the above book. amazon.com/Single-Variable-Calculus-Early-Transcendentals/dp/… it has see inside also. So few more pages are there to see as well.

@Nasser Stewart was a Canadian mathematician. He had a 32 million dollar house built in Toronto called "Integral House". Clearly his textbooks made him some money. :) en.wikipedia.org/wiki/Integral_House

@AlanMunn Yes, he is very famous author of math books. He has a wikipedia page also en.wikipedia.org/wiki/James_Stewart_(mathematician) . For long time, I've tried to find what software is used to make his math books. They look really nice. I doubt that these were all done in Latex. But have no idea what other software could do such a nice job.

@Nasser Maybe this: tex.stackexchange.com/a/120496/2693

@PauloCereda Very interesting. I'd never heard of him before.

@Nasser -- The sums on these pages are not Computer Modern. They look like they are from a Times-like font. So there isn't a clear answer. (Thanks for looking further.)

@UlrikeFischer Will be sorted soon

@Nasser I'm not too impressed of the layout of the book. What part of the design is it that you have in mind that would be most challenging to do in TeX?

@AlanMunn if it is that [APP/3B2] then the mathematics (but not the page layout) is essentially set by TeX. (@Nasser)

Hi folks. Is there any way to write an oval symbol other than generating it using TikZ or similar?

In this particular case I need to circle a word in a document.

There are simple TikZ recipes on the site. Just wondering if there were other options.

@FaheemMitha \oval perhaps?

@DavidCarlisle Which package is that in?

@FaheemMitha standard latex, no package.

Something with adjustable parameters would be nice.

@DavidCarlisle Oh.

There is no such thing as Plain LaTeX, but @DavidCarlisle uses it.

I have a question about tex.stackexchange.com/a/193554/3406

What does the e in

\node[draw,ellipse,minimum height=5pt,minimum width=10pt](e){};

mean, and can someone refer me to relevant documentation in the TikZ manual?

@FaheemMitha It looks to me like a name of the node (section 3.7).

@mickep Ah. Thank you.

@JosephWright: Hi Joseph, I just uploaded filehook to CTAN. Sorry for letting you waiting.

@MartinScharrer Cool: thanks!

@DavidCarlisle, @UlrikeFischer ^^^ We are good-to-go: do you want me to include the ltluatex change

@JosephWright yes I think so (I hope to do hyperref today to match (but I broke xr-hyper at present:-)

@DavidCarlisle OK, I'll cherry-pick, re-tag, etc.

@DavidCarlisle what is broken?

@UlrikeFischer I just realised it was always broken so I'm wondering of I should fix. I had a test file in which referenced two external documents one of which used hyperref and one which did not, that doesn't work in the current public version (probably never worked) It can't be too hard to make sure the re-constructed \r@zzzz macro always has 4 not two fields...

@UlrikeFischer I'll get something working and then probably ask you and @egreg to have a look later (it also includes xcite)

@DavidCarlisle this rings a bell. I think there was either a question or an answer that mentioned that both documents must use hyperref.

@UlrikeFischer xr-hyper is so old, and it's basically just the text of an email I sent to Sebastian that he stuck in a .sty file and distributed:-)

@Skillmon we got a lot of replies for duckuments, mr. rabbit! :)

@Skillmon there we go: twitter.com/aminophen/status/1179492050577674240 twitter.com/wtsnjp/status/1179588372165099520 twitter.com/aminophen/status/1179597882183143424 twitter.com/aminophen/status/1179600893961834497 twitter.com/doraTeX/status/1179606110593925125

@PauloCereda regarding the page counting I showed it to skillmon: chat.stackexchange.com/transcript/41?m=51939288#51939288

@UlrikeFischer thank you!

@DavidCarlisle it would be nice, if it could be merged with xr and then work with/without hyperref.

@UlrikeFischer that's what the comments in the email said in 1993 :-) But it's harder than it seems...

@AlanMunn your Japanese skills reached the Japanese TeX community:-) twitter.com/aminophen/status/1179492050577674240

@PauloCereda thank you!

@UlrikeFischer yes, and again, thank you very much for it!

@UlrikeFischer I have an even more secret plan than the secret plan to fix the xr-hyper != xr issue.

@DavidCarlisle we need an extended label system. zref has a xr-module and with distinct properties it would be much easier to ignore or use something.

@DavidCarlisle ?

@JosephWright trying to build pdflatex-dev.fmt:

(.../tex/latex/l3kernel/expl3.ltx
(.../tex/latex/l3kernel/expl3-code.tex

! Package expl3 Error: No expl3 loader detected.

@UlrikeFischer If I answered your ? it wouldn't be a secret

@DavidCarlisle but what is a secret that only one person knows? You need two to have a leak.

@UlrikeFischer Perhaps everybody knows it except you!

@DavidCarlisle oh

@UlrikeFischer ooh

@DavidCarlisle ooh

@DavidCarlisle a plan? sounds familiar...

@Skillmon it's so nice how we involved a lot of people in this adventure!

Rabbit adventures

Hi! What is difference between packages pdflscape and lscape?

As far as I know, the difference is that "pdflscape adds PDF support to the landscape environment of package lscape, by setting the PDF /Rotate page attribute. Pages with this attribute will be displayed in landscape orientation by conforming PDF viewers" (from CTAN). With other words: with pdflscape pdf viewer rotate presented page for easy reading of page contents on your screen. However, at printing you need to set printer to "auto portrait/landscape" option for correct printing.

Am I correct?

@PauloCereda and all just to get ducky content!

@Skillmon indeed! This is what community is all about! <3

@Zarko well yes except if you use the pdflatex option lscape does that as well (pdflscape should have been a feature request to lscape, not a new package but it's too late now) (I should probably just make them both the same)

@DavidCarlisle, yes, I see somwhere this, but so far I didn't test this possibilities. If I use this, do then also need to set printer as at pdflscape?

@Zarko you shouldn't have to do anything for printing in either case as far as I understand it

Oh hmm I don't know. @UlrikeFischer what effect does \pdfpageattr /Rotate 90 have on printing?

@DavidCarlisle, i just tested your suggestion. It gives on screen the same result as lscape, i.e. not rotate page (I use sumatra for pdfviewer, if this matter) as do pdflatex.

@UlrikeFischer Hmm, it's not actually tested yet ;)

@DavidCarlisle Yup :)

@Zarko did you explicitly use \usepackage[pdftex]{lscape} ? As far as I can see that injects the same code as pdflscape does in pdftex.

@JosephWright when I add \def\ExplFileDate{2019-09-28}% at the begin of the ltx is goes further, but then I get an error in line 11564 of expl3-code that \file_full_name:n in undefined - which is quite odd as the command is clearly defined earlier.

@DavidCarlisle the reference says that it also affect printing. But I never did run tests here. I assume that the settings in the printer driver matters here too.

@DavidCarlisle, ups. In test accidentally I use ˙xelatex engine. With pdflatex works the same as pdflatex. I only need to test printing this result.

@Zarko the page attribute /Rotate 90 code in lscape (which was suggested by a user) is unfortunately hooked to the [pdftex] option so you need to use it explicitly. In the usual cases where pdftex has been detected automatically or the option was specified to graphicx and lscape just uses the same back end, then it is not invoked. It ought to test for which back end is being used rather than which option was used, which basically is exactly the pdflscape package

@DavidCarlisle, after rethinking printer settings ... of course one should select printing option Orientation: to "auto portrait/landscape". Than both version work on the same way.

Only difference between them is now, that pdflscape works with xelatex too but [pdtex]{lscape} not. I throw error: "Undefined control sequence. `\endlandscape ..."

@UlrikeFischer Like I said, I've not tested yet: trying to get my ideas together, hence the branches

@Zarko yes sure, same as graphicx, using [pdftex] would be wrong with xetex, it would be [xetex] except I don't think xetex existed last time lscape was edited, I could add it but then the code would be identical to pdflscape

@DavidCarlisle, than you for explanations. Now all is clear to me.

by

@UlrikeFischer I see the problem: there's one place where we need to avoid messing with some 2e stuff when pre-loading

@egreg Hi, have you know the faktor package? (if we speak about beauriful typography)

\makeatletter
\input{expl3.ltx}
\makeatother
\documentclass{article}
\ExplSyntaxOn
\tl_analysis_show:N \document
\ExplSyntaxOff

@UlrikeFischer ^^^

@UlrikeFischer Doesn't work in document body just yet ... have to fix backend loader ...

@DavidCarlisle, @UlrikeFischer, @egreg, @PhelypeOleinik, @MarcelKrüger Do we want a test in the 'pre-loader' so it can't be loaded other than as part of format-building? It's tricky to set up ...

@DavidCarlisle, @UlrikeFischer, @egreg, @PhelypeOleinik, @MarcelKrüger Also, I have set up pre-loading so it requires TL'19 XeTeX/LuaTeX: OK?

@JosephWright yes

@JosephWright yes.

@JosephWright I guess if someone tries to load in another place it would give lots of curious errors?

@JosephWright er perhaps, team list I suggest

@UlrikeFischer Well no, that's rather the thing: it will still load

@JosephWright but probably only if it hasn't been already loaded.

@JosephWright now the format building was successfull. But for every document I get directly ! LaTeX Error: Missing \begin{document}. ;-(.

@UlrikeFischer Oh ... I'll keep fiddling about!

@JosephWright Ok, but we have to remember to document/communicate the new requirements.

@JosephWright it seems to be some line break or some other invisible char ;-(.

@JosephWright it doesn't like the everyjob message. Shouldn't this be in a \typeout or something like this?

@yo' I saw it and tried to forget about its existence. ;-)

I've seen this error before, but I'm not sure what to make of it.

pdfTeX warning: pdflatex (file ./upstox_form_dates_removed.pdf): PDF inclusion:
 found PDF version <1.7>, but at most version <1.5> allowed

This is when including PDF files in LaTeX files.

@FaheemMitha You are trying to include a PDF 1.7 in a PDF 1.5. Use \pdfminorversion 7

@PhelypeOleinik Thank you. I should have mentioned that the error is presumably coming from \includepdf.

@PhelypeOleinik I should include that in my LaTeX file?

@FaheemMitha Yes, preferably at the very top (or before anything is written to the PDF)

@PhelypeOleinik Yes, I added it. And it makes the warning disappear.

Where is this documented? In the PDFTeX manual? Somewhere else?

@FaheemMitha Yes, pdfTeX manual

@PhelypeOleinik Ok. Thank you very much.

Does it make a difference if I don't include it? Aside from the warnings?

@FaheemMitha it might or it might not. pdftex generated PDF says at the top that it is pdf version 1.5 then it finds later on you include something claiming to be 1.7 it can not go back so it includes it anyway, whether that matters or not depends...

@FaheemMitha Someone else will have to answer that (^^^). I don't really know the PDF specs

@DavidCarlisle So PDFTeX doesn't "support" PDF 1.7?

@PhelypeOleinik Ok.

@FaheemMitha some files claiming to be 1.7 are really 1.5 or older just that whatever generated them can do 1.7 so in that case it is all fine, if it really uses 1.7 features then it depends on which viewer you use, your document then claims to be 1.5 overall but uses 1.7 features.

@FaheemMitha It does. PDF 1.5 doesn't support PDF 1.7

@FaheemMitha it doesn't really need to support anything in that sense but if you set pdfminorversion it will say it is generating 1.7 so including 1.7 is then fine and if a viewer doesn't support that it will refuse the file with a sensible error rather than giving an error mid document with some unexpected command

@DavidCarlisle I'm confused. But let me just ask - what PDF version does PDFTeX output by default?

texlive is rather conservative in its setting so that any pdf generated by any of its tools can be read by all of the tools that claim to read pdf.

@DavidCarlisle If I understand correctly, the problem that arises is "internal" to the document I'm trying to include. Is that correct? It's declaring itself as 1.5 but then has 1.7 features?

@FaheemMitha whatever version pdfminorversion is set to (this doesn't change what is generated except one line at the top of the file) by default the initex settings used for pdflatex and pdftex formats set this to 1.5 in texlive 2019

@DavidCarlisle Ok. So 1.5 by default.

Though pdfinfo thinks the file in question is 1.7.

@FaheemMitha well yes again pdf files always declare their version, it doesn't mean they actually have any new features.

Can someone please give some comments (bit of review) for the following code to get a PDF's page count?

@DavidCarlisle I thought the issue was that the file was declaring itself as 1.5 but actually using 1.7 features.

\ExplSyntaxOn
\char_set_catcode_other:N \%
\cs_new:Npn \__duckuments_set_to_page_count:Nn #1 #2
  {
    \ior_shell_open:Nn \g_tmpa_ior { extractbb ~ -O ~ #2 }
    \ExplSyntaxOff
    \char_set_catcode_other:N \%
    \char_set_catcode_letter:N \:
    \ior_map_inline:Nn \g_tmpa_ior
      {
        \__duckuments_set_to_page_count_aux_a:Nw #1 ##1%%Pages: \q_stop
      }
    \ExplSyntaxOn
    \char_set_catcode_comment:N \%
    \ior_close:N \g_tmpa_ior
  }
\cs_set:Npn \__duckuments_set_to_page_count_aux_a:Nw #1 #2 %%Pages: #3 \q_stop

@FaheemMitha no the issue is that the file you are including declares itself to be 1.7

@DavidCarlisle Oh.

@Skillmon duckuments?

@DavidCarlisle Ok, I reread what you wrote. I think I understand now.

@FaheemMitha a package of mine.

@Skillmon Funny name.

@FaheemMitha it makes documents with ducks :(

@egreg I have a colleague who insists on having I/J with raised I and lowered J. I know there's no way around it. However, his typist used faktor. I try to diminish the tragedy by changing \diagup to \big/ (or other scales), but I don't think this will go very well... I especially love \faktor{\left(\faktor{A}{B}\right)}{\left(\faktor{C}{D}\right)} in inline math :-)

@DavidCarlisle you could use it for menu cards.

@UlrikeFischer not really, there is no pineapple pizza option

@DavidCarlisle make a feature request ...

@DavidCarlisle Seriously?

@FaheemMitha for some definition of serious, yes.

Building LaTeX2e 2019-10-01

@DavidCarlisle Huh.

@DavidCarlisle quack

@DavidCarlisle oh no

@UlrikeFischer you have mail:-)

@DavidCarlisle ooh

@yo' Awesome output, I'm sure.

@DavidCarlisle answered ...

@UlrikeFischer we can blame @egreg for getting us in to this mess

@egreg we can blame @UlrikeFischer for getting us in to this mess

@UlrikeFischer but what do you think of the [nocite] idea? it makes \externaldocument a bit over-loaded with three optional arguments, but might work out OK.

@DavidCarlisle Were you blaming Ulrike back in 2013 too?

@FaheemMitha apparently not:

@DavidCarlisle Per Philip Larkin, always a good choice.

Speaking of egreg, it looks like you are catching up. Unless I'm hallucinating.

en.wikipedia.org/wiki/This_Be_The_Verse, in case anyone doesn't get the reference.

@FaheemMitha I doubt it, I rarely reach rep cap these days and he's 250000 ahead:-)

@DavidCarlisle Ah. It must be those magic mushrooms, then.

@DavidCarlisle yes, I think making the cite code optional is a good idea. But probably even better the other way round: if someone wants it, it should be activated explicitly.

@UlrikeFischer I was trying to be consistent with the xr in tools that's gone to ctan today:-) but perhaps you are right

If anyone is looking for pleasant evening listening, I recommend A-ha.

@DavidCarlisle oh ;-)

@UlrikeFischer ooh

@DavidCarlisle OOH ULRIKE IS RIGHT

If I need to read in a file, how sure is it that I can assume normal catcodes? Should I manually set the category codes of digits to other, or is it safe to assume they are?

@Skillmon is your name Heiko?

@DavidCarlisle ooh the German bloke

@DavidCarlisle no, it isn't. So I shouldn't...

@Skillmon so in that case assume normal catcodes

@Skillmon to test if luatex for example most people do \ifx\directlua\@undefined and call it done, Heiko's ifluatex test does:

%%    ifluatex-test1.tex, ifluatex-test2.tex, ifluatex-test3.tex.
%%
\begingroup\catcode61\catcode48\catcode32=10\relax%
  \catcode13=5 % ^^M
  \endlinechar=13 %
  \catcode35=6 % #
  \catcode39=12 % '
  \catcode44=12 % ,
  \catcode45=12 % -
  \catcode46=12 % .
  \catcode58=12 % :
  \catcode64=11 % @
  \catcode123=1 % {
  \catcode125=2 % }
...\endgroup%
\begingroup\catcode61\catcode48\catcode32=10\relax%
  \catcode13=5 % ^^M
  \endlinechar=13 %
  \catcode35=6 % #
  \catcode39=12 % '
  \catcode40=12 % (

Meanwhile, a giant tomato in a metro

@Skillmon and so works in a couple of theoretically possible different catcode regimes:-)

@PauloCereda or, it could be a very small train.

@DavidCarlisle ooh a plot twist

@DavidCarlisle with midgets next to the normal sized tomato?

@DavidCarlisle but this seems so very robust... :)

@Skillmon or a wrong \begingroup

@Skillmon express service to Lilliput

@PauloCereda deploying new version of duckuments. Should support any engine with full feature set as long as extractbb is allowed in a shell-escape. Do you want to post that on Twitter?

@Skillmon sure thing mr. rabbit!

@PauloCereda thanks.

@Skillmon done! :)

@DavidCarlisle A couple?

@FaheemMitha at most

@PauloCereda and uploaded to CTAN.

@Skillmon word is already spreading in the Japanese community. :)

@JosephWright Petra strikes again ;-)

@UlrikeFischer Yes ...

@UlrikeFischer Re-running the build: will be sorted in an hour or so

@JosephWright worked :-) (with a format)

@UlrikeFischer Yay!

@UlrikeFischer I think it's mainly done: I did say almost all the work was in the prep

@PauloCereda 6666 coming up soon …

@HaraldHanche-Olsen ooh

@JosephWright I must say your Twitter header is really classy. :)

@JosephWright twitter.com/zr_tex8r/status/1179685229759361024

@DavidCarlisle ;-) so which driver do they use? Now I'm confused too.

@UlrikeFischer basically it's like latex so dvips or dvipdfmx, but they (somewhat surprisingly) expect dvips default as that is the way it has always been

@DavidCarlisle for all the variants?

@UlrikeFischer certainly the default is dvips for all the variants as we never thought to add anything to graphics.cfg, which is why I got pushback from the suggestion to add something now.

@UlrikeFischer did you see the reply twitter.com/aminophen/status/1179692996314329088

Hey friends, only hyperref has a way of clickable link?

Or is there any other package?

@PauloCereda well you don't need a package at all of course

@DavidCarlisle really?!

@PauloCereda that is true for every package, no package has private access to the primitives. There is also code in l3backend...

Managed to fix with \usepackage[implicit=false,hyperfootnotes=false]{hyperref} :)

@UlrikeFischer xr-hyper looking OK to check in at least ? (probably won't push to ctan until the weekend)

@PauloCereda currying:

@DavidCarlisle I see I get mentioned :(

@DavidCarlisle Currying was really invented by Schönfinkel – Curry merely popularised it. For some reason, nobody calls it Schönfinkelisation. (Oh wait, Frege did it first?)

@DavidCarlisle :)

@UlrikeFischer From the GitHub comments, I think a change in expl3 would be appreciated (@DavidCarlisle)

@JosephWright gh comments? did I miss them (but changing the default seems requested)

@DavidCarlisle github.com/latex3/graphics-cfg/issues/1#issuecomment-537660976 and 'reactions'

@JosephWright ah yes sorry,

@DavidCarlisle I'll probably do that as part of the current work

@DavidCarlisle I had a quick look and it looks ok.

@DavidCarlisle yes :-)

@PauloCereda to fix what?

@UlrikeFischer hyperref broke a couple of things here. :)

@PauloCereda yes but which? implicit=false will disable a lot of things, so why using hyperref at all?

@UlrikeFischer Just to get a clickable URL, that's the only thing I really needed. :)

It seems that Stuff is happening with SE. Is anyone here tracking it?

@FaheemMitha Huh?

@FaheemMitha In what way?

@JosephWright This was just posted in U&L Chat. News to me.

@FaheemMitha Ah

@FaheemMitha The relicensing is all a bit of a mess: conditions from day 1 should have allowed them to use newer versions of the CC

@FaheemMitha No idea about any firings

@JosephWright Mmm. It all sounds a bit worrying.

Reading more now.

@FaheemMitha There's been some mod-only comment, but as 'we' tend to leave the rest of the network to their own devices, I'm in the dark

@JosephWright plural is imho wrong. One mod was removed.

@UlrikeFischer And some others have resigned. And others have suspended activity. Including our historically most active mod (in U&L).

@JosephWright Well, TeX is a bit special. But do you have contingency plans if something happens to SE?

I'd be sorry if that happens, but if the VCs take over, there is no saying.

@FaheemMitha Yes, there are fall-back plans of course

@FaheemMitha @StefanKottwitz has a server farm in his house ;)

@JosephWright You'd need some software to run it, too. I wonder how hard the chat room software would be to build from scratch.

Maybe someone will do a Torvalds with SE.

@FaheemMitha Stefan has a 'v1'-like set up running goLaTeX, etc.

@FaheemMitha stefans Q&A sites run on an open source platform

@FaheemMitha Chat servers are easy to run

@JosephWright Oh.

@DavidCarlisle Can you remind me what they are?

we could all go back to c.t.t

@DavidCarlisle But you wouldn't get to post duck pics there.

@FaheemMitha several including latex.org/forum

@JosephWright 'v1'-like?

@FaheemMitha excellent

@FaheemMitha It looks like a v1 StackOverflow set up

@JosephWright Oh, version 1.

Also, Gilles resigned as mod.

@DavidCarlisle, @UlrikeFischer Re-uploaded :)

@JosephWright at this time Petra is perhaps not active anymore ;-)

@UlrikeFischer Probably not ... but I suspect they will all get installed tomorrow

@UlrikeFischer Hopefully that's the last time I have to build the zip files on my PC

@JosephWright the next challenge will be to do this for the oberdiek package ;-).

@UlrikeFischer I should dig out my "split" branch for that....

@DavidCarlisle which probably means that one needs to get rid of this hobsub stuff?

@UlrikeFischer Yup

@UlrikeFischer Strip things down, yup

@JosephWright regarding challenge what to answer about the pkgloader issue?

@UlrikeFischer 'No'?

@JosephWright ;-)

@UlrikeFischer 'Ask FMi;?

@UlrikeFischer well I had a mechanism that built hobsub but assuming sibling checked out repos but just dropping it is another alternative

@JosephWright "no" I think (we are adopting too many packages:-) (@UlrikeFischer)

@DavidCarlisle It's not just package number: really it's not the approach we want to encourage to solving package-order issues

@JosephWright well there is that as well yes. (I can't see how it would scale to manage ctan-level numbers of packages)

@DavidCarlisle, @UlrikeFischer The plan of course is that we pick up packages, integrate the ideas into a stable core, make the older packages into stubs ...

@DavidCarlisle Exactly

@JosephWright context mole, that you are

@DavidCarlisle Yoda you are ;)

@DavidCarlisle I must say I never used the package and had to look up what it does. But the pkgloader-recommended.sty was an interesting read - basically saying "loading hyperref is complicated".

@UlrikeFischer Yeah, but with the l3pdf plan, etc., we are moving to fixing that by having support in the kernel

@JosephWright yes. And then we will have to convince the packages and classes to use it ;-)

@UlrikeFischer Yeah, but if it's 'hidden' inside hyperref ...

@JosephWright right, then they have no chance to do it themselves.

@FaheemMitha some more similar setups: texwelt.de and texnique.fr and latex.net.br/osqa

@FaheemMitha German, French, Brazilian Portuguese - no English to not compete with TeX.SE

@FaheemMitha So yes, there are backup ways to continue.

@FaheemMitha chat? there's Openfire and other XMPP based open source ways

I'm getting the following after a tlmgr update. Only with lualatex, not pdflatex. I have no idea how to start tracking it down:

@JosephWright ooh iPad

@dedded that is an old luatex primitive that was removed in luatex 0.85 (4 years or so back) when did you last update??? you could add \usepackage{luatex85} which is a list of compatibility definitions to define the old names

@AlanMunn ^^ :)

/sad trombone

@JosephWrig

@JosephWright I just updated (Texlive 2019 on a Mac). I did tlmgr update --self, then tlmgr update --all. I just tracked it down to the newtxmath usage and posted a MWE on the main page. All it is is a four-line .tex file that uses newtxmath and the word hello.