@mickep -- fantastic! i love it! thank you. (reminds me of my old mechanics textbook, which was conventionally typeset in metal, and is one of the books that made me realize i was interested in how they were created.) have set it aside for my tugboat column. (@JosephWright -- the link you posted is worthwhile too, though not quite as much fun as this one.)

How do I get the maximum width of a single character allowed in text mode? Is there such a length defined?

@marmot Wow, next step 333,333!

@CarLaTeX Grazie but I do not think I will be that long-lived... ;-)

@marmot You got 133K in 2 years and 5 months, you need only 3 or 4 years to reach 333K, even if you are 100 you can reach it :)

@CarLaTeX Long-lived on this site...

@marmot Oh no, please don't add yourself to the list of the left users! :'(

@CarLaTeX Not at the moment but indeed the questions start to become more and more repetition after some while, which may well be the reason why some left. It may not be some fight, it may just be a natural process. I promised the TikZ people to do something, which I really want to do over summer, and then we will see...

@marmot :'(

@bp2017 ?

@UlrikeFischer, @DavidCarlisle A quick look at the latest hooks check-ins would be appreciated

@DavidCarlisle Er .. yeah ... no idea

@DavidCarlisle I suppose the answer is "the width of the paper"

@CarLaTeX or wider if you don't mind an overfull box

@DavidCarlisle Or if you print the char over more pages :)

@JosephWright is this the expected outcome:

\documentclass{book}
\usepackage{l3hooks}
\begin{document}
\ExplSyntaxOn
\hook_new:nn{modA}{hook}
\hook_gpush_all:nnnn{modA}{hook}{modB}{code1}
\hook_gpush_all:nnnn{modA}{hook}{modB}{code2}
\hook_use:nn{modA}{hook}
\ExplSyntaxOff
\end{document}

@bp2017 if you are using classic tex and tfm fonts the limiting factor isn't the page size but that you can only have 16 lengths in a tfm

@UlrikeFischer It occurred to me on my cycle ride it should be gadd not gpush ...

@DavidCarlisle I don't think that applies here, you can only have 16 heights and 16 depths, but 255 widths. But the dimension in a TFM file can only be upto 2048 or 4096 units, so that gives some upper limit. But I think in most cases, the page is significantly smaller.

@UlrikeFischer Well, that is one question: do we want over-write and if so should there be a warning?

@JosephWright the question is more about the repetition. Should code2 be executated twice?

@UlrikeFischer Ah, good point

@UlrikeFischer Hmm, I suppose there should be a \seq_gremoved:cn before the gadd (internally)

@UlrikeFischer I'll add a test for that

@MarcelKrüger ah I should have checked, just thought of tfm constraints as I drove in:-) so in that case I don't understand the question at all:-)

@JosephWright I would say it depends on the meaning of the order. If modeB adds code1 and then modeC adds code2 and modeB adds code3, does mode B expect code3 to replace code1 or that code1 is deletated and code3 appended after code2?

@UlrikeFischer At present, I'm going with 'each entry is at the end of the hook'. Dealing with ordering needs some new seq functions. I plan to mail the team list in a bit

@bp2017 TeX doesn't seem to have any significant limits here, but past 10 times the fontsize, the Type 1 font renderer on Windows gives some trouble. If the font doesn't set some special values, everything above 32 times the font size can no longer be represented in a Type 1 font.

@UlrikeFischer @JosephWright hook_gpush? I just updated and looked in l3hooks.dtx am I looking in the wrong place?

@DavidCarlisle I've just switched it to hook_gadd

@JosephWright ah:-)

@DavidCarlisle push was wrong ...

@UlrikeFischer you are so far behind the times.

@JosephWright that would be consistent with what we did for callbacks, getting rid of the priority mechanism, but allowing you to deconstruct and reconstruct the list in a different order if you really needed it

@JosephWright yes, probably. I guess I will have to adjust my hook ideas a bit. Adding two values should be a wrapper around "set a property code-hook". But I will have to go for walk and some shopping to think this through ...

@DavidCarlisle Exactly

@UlrikeFischer Cool: just thought before I start a long email thread ...

@JosephWright dangerous topic ;-).

@barbarabeeton It is fantastic indeed. The same guy previously wrote at the same web page about how he reconstructed the Byrne setting of the Euclid's Elements. This is also very nice, I think. Except for the geometry part, the way he generates the Lettrines is just fantastic.

@mickep github.com/jemmybutton/byrne-euclid :)

> Alan Turing Receives a (Late) Obituary From the NYT

Hello

If a macro is very long

how can I \show it?

in the log file, it is shown with ...

of which I want to see details

@Yai0Phah it's better to look at the source file anyway where hopefully it is better indented, and has comments?

It is very hard to do so

@Yai0Phah but why do you get ... here for example you can see all of \document which is quite a big macro:

> \document=macro:
->\endgroup \ifx \@unusedoptionlist \@empty \else \@latex@warning@no@line {Unus
ed global option(s):
\@spaces [\@unusedoptionlist ]}\fi \@colht \textheight \@colroom \textheight \v
size \textheight \columnwidth \textwidth \@clubpenalty \clubpenalty \if@twocolu
mn \advance \columnwidth -\columnsep \divide \columnwidth \tw@ \hsize \columnwi
dth \@firstcolumntrue \fi \hsize \columnwidth \linewidth \hsize \begingroup \@f
loatplacement \@dblfloatplacement \makeatletter \let \@writefile \@gobbletwo \g

@Yai0Phah do you have an example?

I am looking at a code in tex4ht

It's doing things like

/edef~{/def/noexpand/eorem:syle##1\th@##2 !*?: {##2}}

I need to trace back to these ##1 and ##2

sadly

I am trying to resolve a conflict

@Yai0Phah one of these files presumably?

$ grep -r 'eorem:syle' generic/tex4ht
generic/tex4ht/amsart.4ht:   \edef\:temp{\expandafter\eorem:syle  \meaning\:temp!*?: }%
generic/tex4ht/amsart.4ht:/edef~{/def/noexpand/eorem:syle##1\th@##2 !*?: {##2}}
generic/tex4ht/amsbook.4ht:   \edef\:temp{\expandafter\eorem:syle  \meaning\:temp!*?: }%
generic/tex4ht/amsbook.4ht:/edef~{/def/noexpand/eorem:syle##1\th@##2 !*?: {##2}}
generic/tex4ht/amsproc.4ht:   \edef\:temp{\expandafter\eorem:syle  \meaning\:temp!*?: }%
generic/tex4ht/amsproc.4ht:/edef~{/def/noexpand/eorem:syle##1\th@##2 !*?: {##2}}

Yes, in fact I am looking at amsthm.4ht

It seems that the code is used to override \@thm

@Yai0Phah so why \show that will just show a mangled version of what is in the file?

you are probably best to make an example and ask a question on site (I haven't used tex4ht in years)

\def\noexpand\eorem:syle##1\th@##2 !*?: {##2}

this is used to parse string after \eorem:syle and omit the part before \th@

It is called by

This parameter depends on the content of #1

since it is related to the callee

@Yai0Phah as I say you are better to ask on the question/answer site rather than here in chat, chat doesn't really work for code level discussions.

Yes, so I only want to know whether it is possible to show verbosely by \show

I tested it there and it showed me a lot of "..."

details are concealed

@Yai0Phah can you show an example, I just tried a long definition:

*\def\a{aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa}

*\show\a
> \a=macro:
->aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.

@Yai0Phah ah, you hav elong csnames?

@Yai0Phah but even there the csname is shown in full in the \show output you only get ... in the context line information.

I just added

here

and then, htlatex gives me

OK, I give up. This seems to be very complicated.

I just report the bug

@DavidCarlisle aah

I mean

ooh

@PauloCereda You've just doubled your vocabulary? :)

@yo' he can also say quack

@PauloCereda Indeed, lovely that it is on github.

@yo' pretty much. :)

@yo' would you mind giving me an opinion? :)

@PauloCereda go on!

Is there any known issue with backref not working correctly?

@PauloCereda indeed

Hi

Can someone tell me the solution to this problem that also includes \acrshort, \acrlong, \acrfull and \Glsfirst?

I have no idea how macros work. For me, that's Chinese and I don't have time to learn Chinese now.

So, if someone could kindly provide me the solution, I would really appreciate.

@Skillmon if it's not been updated since an email I sent to Sebastian 25 years ago, it's lucky it works at all...

@DavidCarlisle CTAN version is from 2016 (ctan.org/pkg/backref)

@mickep -- Thanks! This too definitely goes into a tugboat column. I'll have to do some more research.

@Skillmon that was just noting Sebastian died in the doc:(

@Skillmon what went wrong?

@barbarabeeton Good to hear!

@yo' I was thinking of releasing the source code of my transcriptions of church songs on GitHub. Should I add a license header?

@DavidCarlisle undefined control sequence error with natbib and citation in caption.

@DavidCarlisle having a hard time creating an MWE.

@DavidCarlisle, @MarcelKrüger, I was looking to find the widest glyph in a font in pt units (including fontspec-compatible fonts), if that's possible. Because I have a function that defines line widths in \parshape until last line becomes very narrow (0pt in some cases, which doesn't make sense) and I can't decide on how wide the last line should be (at least allowing for widest character supported by the font seems like a good idea).

@bp2017 which characters are to be considered? Every char in the font? I'd say not doable. Every printable char in the ASCII-table? Seems like an easy loop. Why not simply use 1em though?

@bp2017 going down to a single character sounds like an unusable small length can't you simply take some measurement like 5em or something and not go smaller than that?

@Skillmon what's the undefined command?

@DavidCarlisle \caption@xref

@Skillmon \protect\caption ?

@Skillmon /usr/local/texlive/2019/texmf-dist/tex/latex/caption/caption.sty:\DeclareRobustCommand*\caption@xref[2]{%

@DavidCarlisle it is just in a figure* environment (with two \subfigures before it)

@Skillmon actually I mean to say \protect\cite ?

@bp2017 As @Skillmon wrote, for traditional fonts containing up to 256 characters you can just loop over the font and use eTeX's \fontcharwd to get the width of every character. For other fonts you have either XeTeX or LuaTeX. In LuaTeX you can use Lua to iterate over the font or you could try to access the bounding box directly, for XeTeX you can do something similar to small fonts, iterating over glyph indices instead, but it will be neither fun nor fast.

@DavidCarlisle already tried taht.

@DavidCarlisle I have no idea what causes this.

yyyyyyyyyyyyyyyyyyyyyyyyyyyyyy

@DavidCarlisle What do you want to say with that?

Ä oh flip where is the at sign now.... I am trzing to ping Ulrike

@DavidCarlisle have you changed your OS's language?

@DavidCarlisle @ is Alt Gr + Q

@DavidCarlisle or just copy and paste it :)

@UlrikeFischer so it is

@DavidCarlisle and y and z are exchanged

@Skillmon zes I know that everztime I sat at a machine in mainy and tried to zank text the terminal would suspend...

@DavidCarlisle :)

I think I had better revert the settings...

@DavidCarlisle giving up on recreating that backref error.

@Skillmon \caption@xref is imho from the caption package. check the brf file if it crept in.

@PauloCereda A pretty crappy one, IMO.

@AlanMunn ouch

@PauloCereda I mean, better than none, and I like the idea that the Times is doing this, but I didn't like the obit very much.

@DavidCarlisle For some reason FF has causing my Mac to do this on me. It's very annoying, and has only started happening in the past month or so.

@DavidCarlisle The verz same thing happened to me in Berlin, so it's not citz dependent. :)

@AlanMunn I think FF set some new defaults recently...

@PauloCereda It's very annoying. I have quite a few keyboard layouts active, but I don't want them to switch when I open particular web pages, especially when the pages aren't actually in the language of that layout.

@AlanMunn Do you use FF sync?

@PauloCereda No. I'm not completely certain of the cause, because when it happens I don't notice it immediately, so I haven't been able to reliably reproduce it.

@AlanMunn I use sync, so it seems that FF respects my settings, even after a fresh install and login.

Maybe it's worth a shot. :)

@PauloCereda But the problem is that this is happening at the OS level, even though it seems to be caused by FF. So I'm not sure that sync will help.

@AlanMunn oh

@JosephWright Sigh, I just stumbled over a class with lots of such definitions:

\str_set:Nn \g_language_str {english}
\str_set:Nn \g_english_str {english}

@UlrikeFischer Tut tut

@UlrikeFischer Why even do that?

@JosephWright well the language string is as the class has a setup key to change it, then it tries to use it with babel.

@UlrikeFischer I've found it ...

@JosephWright not so difficult ;-). I got a name to write to. Is tex.stackexchange.com/a/115909/2388 still correct concerning prefix registering?

@UlrikeFischer Yes (it's an alias for the team list)

@UlrikeFischer already did, and yes, it crept in there. But I have no idea why, and can't reproduce it.

@JosephWright as for the registration process of L3 code prefixes. Is that intended for real code, or also for 2e packages? For instance my ducksay doesn't provide any functionality for other packages, but uses the L3 naming scheme internally.

@UlrikeFischer and the exact same code didn't throw these errors with an early TL2017...

Is it possible to create an animation (in a pdf) from a sequences of images?

@Skillmon well if you have code, you could sent to me.

@UlrikeFischer it's not my code, sorry, I can't share it.

@Skillmon can you share the log?

@nbro The animate package? See tex.stackexchange.com/q/466361/2693 for an example. (I have no direct experience with this package.)

@nbro sure, you can even create videos: vimeo.com/337320777

Well, that video is funny :P

Btw, I asked a more precise question here: tex.stackexchange.com/q/494736/63097.

@UlrikeFischer ok, I know how I can solve the issue.

@nbro I think this is a duplicate of tex.stackexchange.com/q/117543/2693

If it is a duplicate, I will delete it. Let me check

@UlrikeFischer or I don't. I have no idea what fixed it, but now it's working

@AlanMunn The accepted solution, if I understand correctly, produces an animation from pdf pages

I am not sure how to adapt it to my case

@nbro it should work with png too.

@Skillmon Anything using L3 naming, ideally

@UlrikeFischer there was a file in the downloaded archive that contained the erroneous backrefs, after deleting that file consecutive runs didn't throw any errors and the bibliography contains th ebackrefs.

@JosephWright in that case you'll get mail soonish.

@nbro I don't see what the problem is: \animategraphics[...options..]{frame-rate}{base-name}{first}{last} should do what you want.

@nbro This doesn't depend on beamer.

I am getting a lot of errors while following this tex.stackexchange.com/a/401900/63097

@JosephWright should I also register prefixes which are not meant to be published to CTAN (at least for now)?

I have 7 png files

However, not in the same directory as my tex file

So, I am doing \animategraphics[width=\linewidth]{12}{images/animations/jtvae}{1.png}{7.png}%

where images/animations/jtvaeis the directory where the png files are placed

@Skillmon Anything generally-available would be best registered: if its really for a limited audience then no, don't worry

Now, I am getting a ! Missing = inserted for \ifnum. error

@nbro That's not the way it works. Just use 1 and 7.

Yes, I tried first that, but I got a ! Package animate Error: None of the files

@JosephWright you got mail (at the modules@latex-project address)

@nbro All of you files should be jtvae1.png etc. with the syntax you gave. Or are they name just 1.png etc. and jtvae is the folder?

@Skillmon Received

So, how can I create an animation with those 7 files (as you can see from the directory structure above)?

@nbro Ok, so rename the files with some base name. img1.png etc. and then use images/animations/jtvae/img}{1}{7}

where the tex file is at the same level of the images folder

@nbro Or possibly just add a trailing / to your base name (untested) without changing the file names.

Now, it compiles, but it only shows the first image

@nbro Viewing is viewer dependent. Try Acrobat reader. You may need to click on it to start the animation.

If it is viewer dependent, I cannot do this

I thought it could work in all cases, but, indeed, the pdf format has some limitations

@nbro Most viewers don't implement things perfectly w.r.t. what Adobe will do.

I should have thought about this before

Thanks anyway

Unfortunately, I cannot provide this animation (because I don't know which viewer my readers will have access to)

Maybe there is another way of doing it?!

@nbro not in pdf, not that works in all viewers.

@DavidCarlisle With regard to your comment at tex.stackexchange.com/questions/494741, it makes totally sense to load just one or two languages. However, I would like to evaluate the current (babel) language. Is there a way to conditionally include package sub-files either on loading time (evaluating all loaded babel languages) or on a on-demand basis?

Be aware: Training a single AI model can emit as much carbon as 5 cars in their lifetimes

(I know that you're like not concerned with it)

I just need to spread awareness (given that I am involved in AI)

@mhchem yes you must be able to test at begin document which babel languages are defined, I'll have a look or perhaps someone here will answer (I don't use babel much:-)

@nbro Although true, and mainly with very fine grained training that has dubious accuracy improvements, the model needs to be trained once in its life, so depending on what it's doing it may be a small part of the overall impact.

@mhchem as for command names if you know you are in (say) french then \hpstatement{P378} just needs to expand to (say) \csname hp-french-P378\endcsname which has been defined by \@namedef{hp-french-P378}{Utiliser \textellipsis{} pour l'extinction.}

@mhchem plus a bit of code for error checking, and saving the command as default for later use.

@AlanMunn Ok, but suppose that all students in the world train several models per semester, etc., or, e.g., suppose that you train a model with different combinations of hyper-parameters

@nbro :) True, but that's not the kind of models they're talking about, I think.

@nbro I'd be more worried about Bitcoin mining. arstechnica.com/tech-policy/2017/12/…

@AlanMunn isn't that what the students are really doing in their spare time?

I haven't actually fully read the article yet (but I will do it later, including the associated research paper). However, we need to be aware of these issues, which are more relevant than ever

@DavidCarlisle In the early days of bitcoin maybe, but the days of the individual miner are long gone.

@nbro Yes, there are hidden costs in all sorts of activities for sure.

we had an important visitor here for the last few days who explained that all these global warming/climate change/carbon use stories are fake news, so there is nothing to worry about.

@DavidCarlisle I could also \def it directly, couldn't I? \def\hp-french-P378{Utiliser ...}

@DavidCarlisle I couldn't find any "list of all loaded languages" in the babel manual, for now. I guess, this could be a proper question at SE.

@mhchem no you can't use - or digits in command names directly (not without changing a lot of defaults and breaking things)

@DavidCarlisle Is conditional \include a thing that works? Or is the \include executed before the \if around it is executed, rendering it kind of useless?

@DavidCarlisle I don't really think so. I think that the weather is really even if slightly changing every year (I hadn't noticed this like 10-15 years ago, but 15 years ago I was young :P)

@DavidCarlisle Ah, that's the reason.

@mhchem that would result in defining a macro \hp that must be followed by -french-P378, a call like \hp-german-P378 would then throw an error.

@mhchem using \include for small chunks of text sound very wrong.

@mhchem you would want input not \include but I can not guess what you mean by the question about conditional, you can include anything in a true or else clause of a conditional, nothing special about \input ?

@nbro You may need to hang around chat to get the local humour. :)

@mhchem I'd put it in a branching test, not inside the \if...\fi block, like so: \if<conditional>\expandafter\@firstoftwo\else\expandafter\@secondoftwo\fi{\include{<stuff if true>}}{\include{<stuff if false>}}, but as long as your files don't contain a whole chapter, I'd use \input instead of \include here.

@Skillmon edit that to \input before anyone copies it:-)

@DavidCarlisle provided that \include is the correct macro here, of course (e.g. the files containing a whole chapter). Of course I wouldn't advice to use \include for anything else!

@DavidCarlisle better, or does it need more changing?

@nbro you don't think POTUS is an authority on the scientific background to the climate?

Hm...

@Skillmon :-) (in the context the files are being loaded in the preamble and contain fragments of text...)

@DavidCarlisle AHHHHHH! that came too late, now the \include thingy sticks with us for all eternity :( There will be even more support requests on Overleaf where I have to tell users that \include shouldn't be used for <insert something wrong here>

@DavidCarlisle Statements like "Well, the United States right now has among the cleanest climates there are based on all statistics." are especially authoritative.

@AlanMunn lies, damned lies and...

@DavidCarlisle You can't have it both ways...

Oh there is a cute little spider on the wall behind my screen.

@DavidCarlisle I was talking more about the concept of inclusion and conditionals, not actual code \include and \if. // My question was: An \input command will be evaluated just as any other command? So, having a never-executed \input will not load anything. No pre-loading, no early-execution-on-parsing or similar?

My cute little new spider friend is leaving me :(

@mhchem no, it shouldn't.

@mhchem tex never looks ahead, it is just a macro expansion language, expanding macros and executingthe one non expandable token at the head of the input buffer

@DavidCarlisle Thanks. Well, I remember (from some years ago, when I dived into the TeX book) that there are several stages. Some things being evaluated on parsing the code, some on executing the code. (I am pretty sure, I am using the wrong words here, please excuse.)

@mhchem I see egreg has posted the \@namedef version (\tl_set:cn same thing basically)

@DavidCarlisle he just copied from your chat message, I assure you!

@Skillmon as always

@mhchem yes as I say expandable commands expand and finally you come to a non-expandable command like \hbox or whatever that is executed but tex never (simplifying slightly) parses ahead it just expands until it has a nonexpandable command which it executes. it is not like a compiled programming language where an entire program is parsed and analysed before code is executed.

@DavidCarlisle Good to know. I had a quick glance in the TeX Book. It is indeed a multi-stage process. Characters are combined to tokens and these are evaluated then. And the exact order of that is so complex that it fills 4 pages full of "danger bendy road" signs.

@mhchem there are (as in any language) corner cases that make the full spec a bit long but the actual execution model is pretty simple really (simpler than a compiled language in many ways)

@DavidCarlisle Okay. I remember that I needed the help of the gurus for things like redefining cat codes temporarily and messing with expansion order in order to get things done.

@DavidCarlisle Why should I include the language files at the beginning? Why wouldn't "load when needed" work? I remember which language is loaded and when a new statement is requested, I \input the definitions before using one of them?

@mhchem most packages with language-specific texts make the languages options of the package so \usepackage[french,german,english]{hpstatement} you could instead pick up any options passed to babel, or you could try to input the definition file the first time \hpstatement is used in a given language but that can be tricky if the first use is in a tricky place like a section heading, but any of those could be made to work

@DavidCarlisle Can you give me a pointer about the section heading? The heading is written into an external file (was it .toc?) and re-used in the table of contents, right? I have to make sure that what gets written into the file is not something expanded half-way into something that cannot stand on its own, right?

@AlanMunn -- Would it were only humor. When visitors to the Marianas Trench see trash on the bottom with printing on it, and little bottom crawlers have the signature of carbon 14 from nuclear tests in their guts, I really think humans have gone too far.

@mhchem something like that yes, captions of course are similar basically anything called a "moving argument" in the latex book. however so long as your files make global definitions and prevent themselves being loaded twice inputting even in those places is probably safe.

@barbarabeeton Our current capacity to destroy the planet is unprecedented, that's for sure.

@DavidCarlisle Thank you so much for your time! This is the Internet (the Internet community) as it should be!

@mhchem :-)

I've just got an equations badge by telling OP how to redefine \( to be \begin{equation}... Spooky

@PhelypeOleinik I got a tikz badge by telling people how to use picture mode to avoid tikz, so it seems Ok to me if you get an equations badge for that

@DavidCarlisle Putting it like that, makes sense :-)

where is my name?

@DavidCarlisle Presumably on your passport?

@marmot hmm, surely I could have an OBE for services to duck recipies

@DavidCarlisle What are you gonna do with an Overly Bold Emblem?

@DavidCarlisle The musical choices seem to be mildly anti-Brexit themed, :) with Costello's Oliver's Army and M.I.A.'s Borders. Good choices.

@DavidCarlisle There is a certain David George Thomas BLOOMER from Carlisle (Cumbria). assets.publishing.service.gov.uk/government/uploads/system/…

@Skillmon, @DavidCarlisle, @MarcelKrüger, thank you.