Isn't even three people quite a small number for all the moderating duties? It seems to me that there must be a lot of work, even for a not-so-big site.

@FaheemMitha No, there's not too much work: as I've said in the current meta post, @StefanKottwitz and I are on top of the work, it's the policy that's important (having at least three mods)

@FaheemMitha We have around 5-10 active flags most of the time, many of which are 'more than 20 comments': on that one, we've traditionally been quite laid-back (a lot of discussion is focussed on the questions-at-hand, so we leave the comments in place)

@FaheemMitha Lots of the smaller network sites have only three moderators

@FaheemMitha For example, from stackexchange.com/sites#traffic, Database Admins (dba.stackexchange.com/users?tab=moderators) has similar traffic to us and three mods

@JosephWright I don't know how is the moderator work, so I believe you when you say it's better to have only 3 mods. I appreciate @AlanMunn because I think he's a wise and calm person, moreover he is Australian, so he is from a different timezone. But, unfortunately, he is also another Anglo-Saxon man. I would prefer a person from another culture.

@CarLaTeX I'm saying that it's one thing one could consider - in the end, the nature of the makeup of the moderation team is down to what the community as a whole feels works best

@CarLaTeX I'm really only trying to say that we likely wouldn't want a large number of moderators - one could argue for four or five, but probably no more than that (we had five when the site was in beta)

@CarLaTeX The key is that there are two separate questions - how many moderators are needed, and who they should be

@JosephWright I think at least 4 mods are needed to have a woman or a person from another culture...

@JosephWright OK, I thought it would be more work. But I suppose crazy disruptions are not so common here. Or people picking fights with each other.

@CarLaTeX I thought Alan was British.

@JosephWright Are the "too many comments" flags automatically generated by the system?

@FaheemMitha The culture is the same :)

@CarLaTeX Culture? What culture?

@FaheemMitha Lol

@CarLaTeX One for a meta thread I think

@CarLaTeX There are of course two aspects: the 'appearance' and the practicalities. I do see entirely that however fair the mod team is, having them all as white men from Europe is ... potentially problematic

Does being a mod here significantly involve making "policy"? For some value of "policy"?

The site seems like it's been doing things the same way for quite a while, and is mostly quiet.

@JosephWright Yes, of course, I'm not saying you are sexist or racist, but being all men of the same culture could be a problem in understanding certain behaviors.

@CarLaTeX I've no idea who might stand when we get a formal notice, so it's hard to judge: the implication is I guess that the top person in a vote will be a guy?

@CarLaTeX All quite true

@CarLaTeX I will add that probably 90% of the work is dealing with 'comment is out-of-date' or 'more than 20 comments' statements - basically 'housekeeping'

@JosephWright The top person will surely be @AlanMunn. He is known, appreciated, and suitable for the role. That's why I think we need at least 4 mods.

@JosephWright Fortunately, the problematic cases are few.

@CarLaTeX Well you've made a good argument for someone else:I suspect you are not the only person thinking the same

@CarLaTeX More mods might lead to a more diverse team, but there's no certainty - it's complicated, and as I said, we don't know at present who will stand

@CarLaTeX Yes. We get very few 'personal' issues, and of those we've had, they've largely been between european guys

@FaheemMitha You're right! They told me @AlanMunn is British living in Canada. I confuse him with Will Robertson

@PauloCereda very poetic!

@samcarter ooh

@PauloCereda I wonder which poetic metre this is :)

@samcarter let's open a siunitx issue :)

@JosephWright ^^ ooh

@samcarter we need to get those version numbers going :)

@PauloCereda yeah! We don't want to wait too long for the cake

@samcarter ooh

Do you know how 490 Roman soldiers laugh? ... XD

@yo' LOL

@PauloCereda one of those rare shakespeare lines that does not use iambic pentameter

@yo' hahaha

Wow an entire concept for me! tex.stackexchange.com/q/616790/101651

@DavidCarlisle ooh

@CarLaTeX :D You should write a package for this!

@samcarter I'll ask @Skillmon!

@CarLaTeX ooh

@CarLaTeX @Skillmon I would use this font for the maggi environment images.creativemarket.com/0.1.0/ps/2604852/600/840/m1/fpnw/wm1/…

@samcarter funny font!

You can now discuss PGF/TikZ on Matrix: matrix.to/#/#pgf-tikz:matrix.org

@HenriMenke YAAAAY

@CarLaTeX :) But a nice environment for a package should maybe not use \@hangfrom (that is really "instable" from a user's point of view) but built upon \trivlist, I was just too lazy.

@CarLaTeX -- Also soup mixes. History here: nestle.com/brands/allbrands/maggi An ancestor, perhaps?

@CarLaTeX -- While it would probably be a "good thing" to have a female person among the mods, there are really very few of us. A quick scan of the first five pages of users (36 shown per page on my screen) finds an average of no more than two per page, or about a half of one percent. And of those shown, the ones I'm familiar with are all very busy, even "overcommitted". I haven't got any good suggestions.

@barbarabeeton maybe, it's a Swiss company and the Maggi's come from the Maggia valley (that of Lake Maggiore)

@barbarabeeton Let's see the candidacies!

@Skillmon ooohhh

@CarLaTeX Pretty sure that white women from the same places that the white men come from have the same mentality. I mean, you've pretty much lumped together around 200 million people (only counting the males) in three different continents as the "same culture" based on skin colour. Sooo.....

@Plergux Indeed, I said a non-European woman

@CarLaTeX Well, a non-European woman can still be North American or Australian.

@Plergux Let's say "non-AngloSaxon woman", then

@CarLaTeX That's a nice idea in theory, but in practice there aren't many candidates who are sufficiently active on TeX.SX to take that role.

@CarLaTeX >20k rep (i.e. trusted user) should be a requirement for becoming a moderator.

We already have a duck mod. :)

@CarLaTeX do you mean EU-citizen with European? Just wanting to make sure I understand the implications for the British.

@CarLaTeX Ok, then I'm gonna ask "What is an anglo-saxon woman?" There haven't been Anglo Saxons in this world since around 1000 BC. I suppose I see what you're getting at but by the same token, if I want to learn about South American culture, do I just go to Spain?

@HenriMenke ^^ "our unit tests work" :)

@PauloCereda bad unit tests.

@Skillmon No, I intend geographic Europe, we already have a British mod

@CarLaTeX (I know, was just the poor attempt of a joke)

@Skillmon :)

@Skillmon :P

@Plergux In Italy we call Anglo-Saxon the British, German and similar people, still nowadays. A south American woman should be good. Spain is still Europe, so there is the problem of the same timezone as the current mods

@Plergux mod could not be found-o :D

@PauloCereda :þ

@Plergux -- Don't forget Portugal. There's also Cabo Verde or the Azores. But honestly, I can't name a single woman TeXer in South America. It would please me to see more. There are some capable ones in Japan. but not active on tex.sx. India?

@CarLaTeX So would I count as Anglo Saxon then? (Icelandic, white as mayonnaise)

@Plergux ooh may-o :D

@Plergux I think you do

@CarLaTeX Germans are definitely not Anglo-Saxon.

@HenriMenke That's how we call them

@Plergux Icelanders are Scandinavians (I guess?).

@CarLaTeX Maybe, but it's wrong nevertheless.

@HenriMenke -- But some are Saxon (as was one of my grandfathers).

@HenriMenke Depends on where you are looking. Many places list Scandinavian countries as the same as the Nordics, even including Greenland, but I think the official definition is: Denmark, Sweden, Norway (and maybe Finland if they're being generous).

@barbarabeeton I think the overriding definition that would apply is perhaps one which one is not allowed to say any more.

@Plergux youtu.be/pynlaaLWspQ?t=130 (starts 2:10)

@PauloCereda XD Gotta love Brian Blessed. Just say it louder, sure that's more right XD

@Plergux I freaked out when I found out he voices Peppa Pig's Grampy Rabbit!

@PauloCereda LOL

@barbarabeeton But Saxons and Anglo-Saxons don't have a lot in common anymore. According to Wikipedia the latter emerged from the former back in the 5th century.

@HenriMenke -- Yes, I know that. Three of my grandparents and two of my great-grandparents migrated from (different areas of) Germany in the mid 19th century, no doubt to avoid the draft. (That part of the story didn't survive the next generation, or more likely was hushed up.)

@Plergux -- Regarding Scandinavia and Greenland, you're ignoring the Sami and Inuit. But I'm not aware of any TeX usage in those cultures.

LuaMetaTeX now supports Potrzebie units as dimensions. mailman.ntg.nl/pipermail/ntg-context/2021/103662.html

@barbarabeeton Well, I was only thinking geographically. Culturally I'm about as Scandinavian as a Big Mac.

\let\anglosaxon\german

quack :)

@Plergux ooh

@Plergux Ein Groß Mac, bitte

@PauloCereda LOL. Vollen sie mit das auch Freiheitskartoffelln, mein Herr? XD

@barbarabeeton While there certainly is an imbalance, the numbers might not be as bad as it seems. I don't know if there are any statistics for tex.se, but for stackoverflow there are reports which suggest that looking at the user names will underestimate the number of female users because they are more likely to choose a neutral username or even a male name.

@samcarter -- I'm certainly aware of that, but in fact I'm at least marginally familiar with nearly all of the inhabitants of the first five pages of users of tex.sx. It's likely that a higher percentage of women will appear later on, but if a certain minimum of activity is required for a mod, then likely candidates would have to appear in the earlier pages. It's still not going to be a large population.

@barbarabeeton “certain minimum of activity is required for a mod” I think mods should have >20k rep on some SE site so that they are at least a little familiar with moderation.

@HenriMenke -- While what you say is true, I don't really know whether the >20k needs to be on the particular site to be moderated, or just a total overall on the .sx platform. I'd be happy to be proved wrong, and find that there *is a knowledgeable and willing candidate out there.

@barbarabeeton I don't think it's necessary to be a trusted user on the to-be-moderated site, but having seen and used the regular moderation tools before is essential, I believe. In fact, it might even be beneficial if a mod was an “outsider” so that they are truly impartial and objective.

@HenriMenke -- While familiarity with moderation tools is certainly a good thing, there are times that I find it hard to determine whether a question fits on the site. (I'm not talking about questions that are "solved in comments".) If a moderator can't determine that almost without thinking, then it's not likely to win respect.

@barbarabeeton Can you tell who is male and who is female by looking at the profiles? The user's sex is not a required field. And most people don't say.

@Plergux I suspect that @CarLaTeX is uncluding in the same group so-called European settler societies. A polite way of describing places where Europeans came, saw, liked it, and proceeded to murder the inhabitants and steal their land.

@barbarabeeton There is at least one company (or used to be) in India (Kerala) called River Valley Technologies. But I don't know if that has any representatives on TeX SE.

@FaheemMitha -- I'm in the fortunate position of having met a number of these people in person at various meetings, and even though a lot of others hide behind pseudonyms, they admit to having created packages, and CTAN doesn't hide authorship. There are other useful connections too. Not perfect, but not entirely "blind".

@barbarabeeton I see.

@FaheemMitha Yes, well you will excuse me if I resent the insinuation that this makes every white person in the world guilty of racism by proxy through all eternity. Even if they live in these "conquested" places such as Canada or Australia. You can't choose the place you're born or what colour you are. All you can do is try to be good and hope people don't judge you for it.

@Plergux I wasn't suggesting that every white person in the world was guilty of racism by proxy. That wasn't implied, nor could you infer it from what I said, which was merely a factual statement.

@FaheemMitha -- I am not sure. But the woman I know in India who uses TeX actively is not in Trivandrum. (There was a TUG meeting at River Valley, and I met quite a few of the Indian TeXers there.)

@barbarabeeton Are they active on TeX SE?

@FaheemMitha Obviously not. It wasn't you who was insinuating it.

@Plergux Sorry, I don't follow.

@FaheemMitha Never mind. Doesn't matter.

@FaheemMitha -- Not to my knowledge, but I could be missing some.

@barbarabeeton Is River Valley an actual river valley? Or is it just a name?

Dropping out for a while. Going to a concert.

@Gus Wouldn't it make the package usage itself invisible outside the group?

@FaheemMitha I am not sure, and I'm not sure what you mean

@Gus Anything within the group is restricted scope. I would have thought (but have never tried, so don't know) that would apply to the package itself.

I.e. outside the group, it will appear that the package was never invoked.

What does that mean functionally? What does it mean to invoke a package and for it to appear invoked? Do you mean that the final document will not contain any output emitted within the group?

@barbarabeeton oh no! Someone is looking behind my pseudonym! :)

@Gus Just try it. Write \includepackage{something} and wrap it in a group. Then try to use the package outside the group.

I'm not trying to include the package, I'm just trying to do \alglanguage{blah}.

longtable, to take a random example. But there are hundreds of choices.

@barbarabeeton oh, have fun!

@Gus You'd need to include algorithmicx first, right?

And your use case is to have the layout for only part of the document?

@Gus the result of {...} and \begingroup...\endgroup is (practically) the same, except that \begingroup...\endgroup can't delimit macro parameters. But when used as a group the only difference is that one of them can't be matched by the other, so {\endgroup and \begingroup} are both wrong.

@FaheemMitha exactly

Ah, if that setting is separate from the actual \includepackage{algorithmx} invocation, then it might work.

@Skillmon , you mean I can do e.g. \textbf{blah} but not \textbf\begingroup blah\endgroup

@Gus I suggest just trying it. As I understand it, stuff defined inside the group only lasts within the group.

@Gus at the end of a group in TeX all local assignments made inside that group are reverted. Global assignments are not affected by such scopes. I don't know whether \alglanguage does use a local or global assignment, so you have to test that.

@Gus correct.

Is there some way to persuade \immediate\write18 to return an error? Or not?

If not, I'll be reluctantly forced to use Lua.

@Gus as a rule of thumb: In the document you are free to use \begingroup...\endgroup or {...} but for programming you should use \begingroup...\endgroup for scoping and {...} for arguments, that makes the code more readable.

@FaheemMitha use case?

Assuming lxx isn't an actual command on the system. (Command name randomly chosen.)

Here it happily executes it without complaint. With pdflatex -shell-escape...

@FaheemMitha well, it prints sh: line 1: lxx: command not found and system returned with code 32512 on the terminal/in the log. But it doesn't throw an error.

@Skillmon Yes, it prints that to the terminal. But can I make it throw an error?

Specifically, I want it to stop execution.

@FaheemMitha in pdfTeX this doesn't seem built in, take a look here: tex.stackexchange.com/questions/13226/test-success-of-write18

@Skillmon Though I don't see it in the log. Just on the terminal.

@FaheemMitha oh, you're right. I wasn't aware of that.

@Skillmon Thank you. That's helpful. Though it seems like a fair amount of work.

Actually, that seems to be testing for something different than I'm after. It's checking if the \write18 call is allowed to be executed. Which seems effectively equivalent to checking for whether -shell-escape was passed as a parameter. At least, I'm not aware of another way to allow the \write18 call to be executed.

The link here seems to be broken.

The second one. Amazingly, the Google Groups one still works.

Incidentally, Lua doesn't have great support for this either, but I can up with this not very nice bit of Lua code.

local function exec_cmd(command, stdout_message, stderr_message)
   local pipe = io.popen(command)
   local stderr = pipe:read("*all")
   pipe:close()
   local filename=os.tmpname()
   local f = assert(io.open(filename, "r"))
   local stdout = f:read("*all")
   f:close()
   os.remove (filename)
   if stdout ~= nil and stdout ~= '' then
      texio.write_nl(string.format(stdout_message, stdout))
   end
   if stderr ~= nil and stderr ~= '' then
      error(string.format(stderr_message, stderr))
   end

Sadly, Lua doesn't have anything like Python's subprocess library. Which does everything you could want.

@FaheemMitha yes/no. The answers cover both, checking whether write18 is enabled but also show ideas how to check whether they actually worked or not, and to get back the results. You could write a shell one-liner that generates a file that tells you whether everything worked or whether there was an error (in shells you can typically use $? to get the exit code of the last used command).

@Skillmon OK. I didn't read through it that carefully.

@FaheemMitha very primitive approach:

\documentclass[]{article}

\begin{document}
\immediate\write18{lxx foo; echo $? > \jobname.shellstatus}
\begingroup
\openin0=\jobname.shellstatus
\read0to\myshellstatus
\closein0
\expandafter
\endgroup
\ifnum\myshellstatus=0
\else
  \GenericError{}{The shell had an error}{}{}
\fi
\end{document}

@Skillmon Thank you. To be really useful, it should return the error message.

@FaheemMitha on a Unix:

\documentclass[]{article}

\begin{document}
\immediate\write18{lxx foo 2>\jobname.shellerror; echo $? > \jobname.shellstatus}
\begingroup
\openin0=\jobname.shellstatus
\global\read0to\myshellstatus
\closein0
\openin0=\jobname.shellerror
\global\read0to\myshellerror
\closein0
\expandafter
\endgroup
\ifnum\myshellstatus=0
\else
  \GenericError{}{The shell had an error:^^J \space\space\space\myshellerror}{}{}
\fi
\end{document}

Does putting a lua file in ~/texmf/tex/luatex make it globally visible within the TeX installation? This appears to be what I'm doing, but I don't recall where this is documented. In the LuaTeX manual?

@Skillmon Thank you. I'll try that.

@FaheemMitha that's part of KPSE (in TeX Live).

@Skillmon Is that separate documentation?

Kpathsea? As in Karl's Path Search?

OK, looking at the manual returned by texdoc kpathsea.

@FaheemMitha yes.

This section from texmf.cnf, is confusing.

% LuaTeX.
TEXINPUTS.luatex          = $TEXMFDOTDIR;$TEXMF/tex/{luatex,plain,generic,latex,}//
TEXINPUTS.luajittex       = $TEXMFDOTDIR;$TEXMF/tex/{luatex,plain,generic,latex,}//
TEXINPUTS.luahbtex        = $TEXMFDOTDIR;$TEXMF/tex/{luatex,plain,generic,latex,}//
TEXINPUTS.luajithbtex     = $TEXMFDOTDIR;$TEXMF/tex/{luatex,plain,generic,latex,}//
TEXINPUTS.dviluatex       = $TEXMFDOTDIR;$TEXMF/tex/{luatex,plain,generic,latex,}//
TEXINPUTS.lualatex        = $TEXMFDOTDIR;$TEXMF/tex/{lualatex,latex,luatex,generic,}//

I don't understand all these different options. And is luatex a format? I thought latex and plain were the formats?

@Gus \begingroup makes a semi simple group (Knuth's little joke, I'm sure)

@FaheemMitha see the format= banner on line 1 of every log file: This is LuaHBTeX, Version 1.13.2 (TeX Live 2021) (format=lualatex 2021.9.20) 26 SEP 2021 21:29

@DavidCarlisle So luatex is a format, and lualatex is also a format?

I thought format and engine were separate.

@FaheemMitha there is nothing special about TEXINPUTS you can use that /format nottaion with all the kpathsea variables. fro example pdflatex and pdflatex-dev are just different releases of the same thing with the -dev version being the test release of the main version, but by having TEXINPUTS.pdflatex-dev you can have a separate input path for pdflatex-dev

.... so it pulls in the development version of classes and packages from texmf-dist/tex/latex-dev if there is anything there before looking in the main tree

@FaheemMitha the luatex format is plain tex built in luatex

So the term format is just used generically? As opposed to engine {PDFTeX, LuaTeX} vs format {LaTeX, ConTeXt}?

@FaheemMitha well, each format has to be built against the engine you want to use it in, as a format is essentially just a dump of the engine's memory.

as meant there it is a specific file, eg the pdflatex format is /usr/local/texlive/2021/texmf-var/web2c/pdftex/pdflatex.fmt but the .suffix does not relate to the format but to the program name.

@FaheemMitha so you can have the "same" format as in plain TeX vs. LaTeX, but build it for different engines. And to somehow separate this, someone has decided to use different names as in luatex vs pdftex and lualatex vs pdflatex. The first two are plain TeX (but for different engines) whereas the latter two are both LaTeX (but for different engines)

@Skillmon OK, I see.

@Skillmon --

@Skillmon -- Not to worry. I won't divulge your secret. And the concert was fun. One of the offerings was a concerto for cello, but there was a second (unheralded) cello that was consistently playing a duet with the soloist. Only in the cadenza did the soloist play solo. I really preferred the duets.