12:00 AM
@Shalvenay Use this as your model: tex.stackexchange.com/q/116534/2693 there it's clear that you can definitely use a macro for the post-break continuation character, so the only bit to figure out is how to make it conditional.

12:14 AM
@AlanMunn yeah -- that's what I'm really not clear on

@Shalvenay Yes, I understand. But if you ask your question using a minimal example like this then you can make it quite clear to people what you want to do. I'm sure there's someone who might be willing to dig into the listings code to figure out a solution.

@AlanMunn yeah

5 hours later…
5:04 AM
Hi
I am using texmaker and when I click on run to generate the pdf file then it flashes an error message - pdflatex ~ synctex=1
any help!

2 hours later…
7:36 AM
@AlanMunn New avatar?

@Johannes_B It happens: on a forum you try to answer every question, that's not quite the same here

@JosephWright The questions have been closed too quickly, imho.

8:00 AM
Today an OP's accepted an answer of mine posted on May, 17th :):):)

@CarLaTeX I think I had an acceptance more than one year later after I posted the answer

@ChristianHupfer Better late than never! :)

@CarLaTeX Yes, where 'never' is often the case, however ;-)

@ChristianHupfer Unfortunately...

@Johannes_B: One of the pretty unique 'thesis.cls' questions again: tex.stackexchange.com/questions/376664/… ;-)

8:15 AM

8:33 AM
@ChristianHupfer This is lowercase thesis.cls, not Thesis.cls. Alas, i don't have hope.

@Johannes_B Rubbish is rubbish, regardless whether it is written with lowercase or uppercase ;-)
@PauloCereda: Which reminds me of "Upperclass twit of the year" competition ;-)

@ChristianHupfer True that.

@Johannes_B: And here is one of your other favourite classes: res.cls ;-)
1

I am using the popular res.cls for my resume (CTAN Link) I would like to put internal links in my file. However, the following does not successfully place the internal links: \documentclass{res} \usepackage{hyperref} \begin{document} \section{Section 1} \label{sec:part1} \hyperref[sec:part...

@ChristianHupfer Please do not make the template hunter angry on sununday. I am going to drink too much coffee which almost makes my heart explode.
COFFEE!

9:16 AM
@JosephWright you could do if you want, or I could but your probably have it in your head what you need to pass down...

@DavidCarlisle OK, I'll sort in a second (just doing the l3build split)
@DavidCarlisle Do we call it decode or decodearray?

@JosephWright decodeararray I think as it's what it was before and "decode" is rather a generic name for a rather specific thing

@DavidCarlisle Cool

@JosephWright still fighting travis ci :-)

@DavidCarlisle :)
@DavidCarlisle Set up a VM!
@DavidCarlisle Updated graphics stack off to CTAN

9:32 AM
hello

@SBM Hi

@JosephWright don't think that would help I have all the build working it's just deploying back to the gh-pages branch of another repo, which also works but the doc says "Deploying to GitHub Pages uses git push --force to overwrite the history on the target branch" which is fine for my test repo but means if I update the one document I'm building via travis CI I wipe the entire website, so I think I need to make the install git clone the project site update teh document and then push back
@JosephWright thanks

@DavidCarlisle Ah
@DavidCarlisle Not a problem: I'll add support for the same concept to L3 soon

@JosephWright I'm trying to understand why they only offer this --force option, reading the git documentation for that more or less just says make sure that to really want to do that:-)

@DavidCarlisle Yup: that's needed for example when you mess stuff up entirely and need to say 'right, everyone else forget about the history, I'm changing it' (I did that when we moved from svn-mirror to latex3 for the Git view of LaTeX3, as the old one was read-only and the new one isn't, and is done using an entirely different system)

9:51 AM
@JosephWright but it seems a dangerous thing to do on every edit, which is what the built in travis ci deploy to gh_pages thing does

10:04 AM
@DavidCarlisle Er, yes

10:19 AM
@JosephWright woo
@AlanMunn Great job, Alan! I will add texindy to the official rule set. :)
@AlanMunn: in other news, what happened to your avatar? :)

1 hour later…
11:21 AM
@Johannes_B: I sometimes tell new users, who currently use a standard class, how to use scrlayer-scrpage not only because I now scrlayer-scrpage better but mainly because they could later switch to a KOMA-Script class. That's not fair, because fancyhdr is a good package. And I should tell them, how to use either fancyhdr or scrlayer-scrpage.
I would prefer if fancyhdr would have better support for the KOMA-Script classes so that I never again have to say: "Yes, you can use this, but note that it could break features of the class you are using." Maybe someone could write a fancyhdr interface to scrlayer that does support also all page head/foot features (and the font element feature) of the KOMA-Script classes.

@Schweinebacke One could argue as well that the KOMA classes should be rewritten/changed to be usable with fancyhdr ;-)

@ChristianHupfer This would mean to remove features like \setkomafont{pagenumber}{…} and options like headsepline from KOMA-Script or to patch a loaded fancyhdr a lot which could break after a update of the package. Mostly making a package compatible with a class is more easy.
The first releases of KOMA-Script where compatible with fancyhdr only because they did not have had such features. scrpage (the predecessor of scrpage2 the predecessor of scrlayer-scrpage) has been designed and implemented only because of such incompatibilities of fancyhdr.
BTW: KOMA-Script classes are usable with fancyhdr. A lot of users use them together. But it's not a perfect combination. At least not for new users. Experts are able to use fancyhdr with support for all KOMA-Script features. But they have to do it on their own.

11:40 AM
@Schweinebacke There is also titleps ;-)
@Schweinebacke Do you want to rewrite that wikiboooks page? In short mentioning all three packages?

@Johannes_B And some more. BTW: The very first KOMA-Script releases had (one or even some) additional page style package(s). Each provided only one page style alternative. scrpage was also an answer to create one new package per each page style question.
@Johannes_B I don't like to write wikibook pages. I'm not good in writing English manuals. And I need too much time for them. And if I'd do, I'd feel responsible for the result for ages. No, thank you.

@Schweinebacke The german wikibook? ;-)

@Johannes_B The German wikibook is a mess. And same problem: It need a lot of time and I would feel responsible but cannot fulfill my own pretension (don't know whether or not this is the correct term, in German: kann meinen eigenen Ansprüchen nicht gerecht werden bzw. sie nicht erfüllen).

@ChristianHupfer There are worse things than unaccepted answers: tex.stackexchange.com/q/367858/101651 :):):)

@ChristianHupfer Oh no. Try to being compatible with everything is as bad as trying to be please everyone. It leads to fat, complicated and at the end unmanagable code. Due to the history (scrpage, scrpage2, scrlayer etc) it is already complicated enough. I don't want more "if then" cases.

11:53 AM

@ChristianHupfer LOL

@UlrikeFischer Well, I don't try to use best of both worlds then. One should choose one approach and drop the other one, each on a personal decision

@UlrikeFischer Have you seen the code in scrkernel-tocstyle.dtx? Extremely optimized code to avoid having duplicated lines of code. But I you'd try to find out what the result would be or where to change or add something …

12:08 PM
@Schweinebacke I know what you mean. I feel responsible as wll, so i do some edits every once in a while. I even extended the spanish wikibook on title pages without knowing the language. :-)

@ChristianHupfer Sorry I don't understand what you mean here ;-) I don't have to drop one of KOMA and fancyhdr. They work together. But I want don't want KOMA to e.g. silently add \usekomafont-commands to fancyhdr code. It is quite enough if it documents the "API": \setkomafont{pagenumber} will afffect elements that uses \usekomafont. Then I can add them myself if I want.

@Schweinebacke The code of the KOMA classes got too complicated for me. I get goose bumps seeing some answers that implement what a user wants in a KOMA-clean way. If i canot figure out what is happening, what does an unexperienced user think?

@UlrikeFischer No one is forced to use both, however

@ChristianHupfer No, but what has this to do with your remark that "One could argue as well that the KOMA classes should be rewritten/changed to be usable with fancyhdr "?

12:23 PM
@UlrikeFischer It was a remark about @Schweinebacke's complaints that fancyhdr is not (fully) supporting KOMA ... It's justified to say that fancyhdr should be changed as is equally justified to change/adapt KOMA

@ChristianHupfer KOMA-Script has been adapted in some aspect: It has it's own page style package(s).

@ChristianHupfer No sorry the relation is not symmetric. The class is the frame work, it offers tools and API's which users and package can use or not. So e.g. for headerlines KOMA could tell package authors "the option set this boolean and this values, and you can use them if you want". But it should not patch packages or package code.

@JosephWright I got bored. :). It may go back to the other one at some point.
@PauloCereda Thanks, but I assume the rule I wrote isn't general enough for inclusion in the official rule set. But I don't know texindy at all so I don't know what all the parameters do.

12:43 PM
@UlrikeFischer I've another opinion, but as I said: If you don't like fancyhdr, don't use it, if you don't like KOMA, don't use, if you want to use both, live with the consequences. It's a personal decision

@ChristianHupfer @UlrikeFischer @Schweinebacke I don't understand why fancyhdr should support KOMA at all. Does it claim to work with all classes? Should it work with memoir? Or should memoir work with scrlayer-scrpage? DIfferent things for different purposes.

@Johannes_B Exactly

@ChristianHupfer A general new user doesn't know about the differences. They just use what google presents first.

@Johannes_B That's correct. But it is also the reason why it is not enough to explain how to use fancyhdr (or titleps or scrlayer or scrlayer-scrpage) to describe how to change page styles.

By the way, i never noticed the fancyhdr documentation has a completely different title and deals with floats as well.
@Schweinebacke And that is the point where a wikibook gets tricky. First of all, before we deal about page styles, you need to know what document class you are using. Memoir? Go to page 8. KOMA? Go to page 12? standard? you can choose between ABC on page 15, XYZ on page 18 or go to page 12 and use the method that is suited for KOMA. if you use titlesec however, forget about page 12 and use XYZ fro page 18.

12:50 PM
@UlrikeFischer Agree. And this is the reason why KOMA-Script classes only warn and one of the reasons why scrhack is not integrated into the classes.

@Johannes_B They are trapped and are lost forever -- abandon all hope ;-)

@Schweinebacke I know a class that has \PreventPackageFromLoading to avoid loading fancyhdr

@Johannes_B scrlayer – the basic package behind scrlayer-scrpage does also deal with floats. And the \iffloatpage in fancyhdr is a very simple \if@fcolmade #1\else #2\fi (similar for the other switches). The user has to use it himself to deal with float pages. In scrlayer-scrpage a user also could add float page layers and non-float page layers to the page styles. But fancyhdr bundle has also package extramarks. But AFAIK you can use this also with scrlayer-scrpage.
@Johannes_B Yes. But I don't think, that KOMA-Script classes should do so. Maybe even the warning should be less strong since fancyhdr has been changed to not longer use \it.

@Schweinebacke Agreed.

1:05 PM
@Johannes_B BTW: The ncctool has nccfancyhdr.sty. But I don't know if some should use them, e.g., if he/she uses, e.g., nccbook.cls. So a wikibook can never keep in mind to explain every best combination.

@Schweinebacke It could be worth mentioning thesis.cls and Thesis.cls <- @ChristianHupfer
;-)

@Johannes_B You should see my thesis.cls I've used for my diploma and PhD thesis ;-)
\NeedsTeXFormat{LaTeX2e}[1995/12/01]
\ProvidesClass{thesis}
\endinput

@ChristianHupfer I always have to live with the consequences of my decisions. I like fancyhdr as it is easy and fast to use, but I also like KOMA as it has various useful interfaces -- e.g. to change the fonts of the sectioning commands.

@ChristianHupfer By the way, i still get occasional visitors for CreativeCv.

@ChristianHupfer That's @PauloCereda 's entire thesis.

1:16 PM
1

I have been using TeX Live on Ubuntu 14.04 and the package classicthesis with no problems until now. I have recently reinstalled Ubuntu 14.04 and TeX Live with it. I followed the guide on how to install "vanilla" TeX Live (very good guide). But this very simple piece of code had a problem \doc...

Close as unclear?

@AlanMunn I wasn't aware that @PauloCereda is writing a thesis. I thought he is writing a book about procrastination ;-)

@AlanMunn I will have to rewrite it anyways because the v4 format is more robust and less error-prone, don't worry. Also, I do believe Nicola had a plan for a texindy rule in the past.
@AlanMunn oi
@ChristianHupfer oi

@PauloCereda Ok. I'll leave it to the experts. The project level config files for latexmk are nice to know about though too.
@PauloCereda :)

@PauloCereda <3

@AlanMunn er arara
@ChristianHupfer <3

1:25 PM
@PauloCereda I will use all tools at my disposal. :) I usually use latexmk for final runs of documents.

@AlanMunn Oh I think you meant the new v4 feature, the project level config. My bad. :)
@AlanMunn: what happened to your avatar?

@PauloCereda No I was referring to Herb's answer.

@AlanMunn oh sorry, I made a confusion.

@PauloCereda I got bored.

@Schweinebacke There should options for package/class writers to silence such warnings. When I load the fancyhdr package I know what I'm doing. \expandafter\let\csname fancyhdr.sty-@before\endcsname\relax works but as it is not documented one can't be sure if the command name will stay in future koma versions.

1:30 PM
@AlanMunn ooh
@AlanMunn: speaking of boredom, one of my favourite movie lines is from Al Gore in a Futurama film: Finally, I get to save the Earth with deadly lasers instead of deadly slide shows.

@PauloCereda Where's Al Gore when we need him?

@AlanMunn Hunting manbearpig, I believe. :)
(Another South Park reference)

@PauloCereda whooooosh

@AlanMunn LOL

@UlrikeFischer Such options would become part of most templates and would make the warnings useless.

1:32 PM
@AlanMunn EXCELSIOR

@PauloCereda Watch it. Pretty soon my 'whoosh'es will have a certain southern Italian flavour...
@PauloCereda Until I run out of battery.

@AlanMunn ooh

@AlanMunn :)

@JosephWright now I can do PR's directly to the new repo, right? :)

@PauloCereda Yup
@PauloCereda I expect to follow a slightly more 'modern' workflow with this new repo

1:42 PM
@JosephWright I wonder what will be outcome of the meeting regarding this new modern vibe :)

@Schweinebacke templates can silence the messages anyway. Either with the low-level code above or with the silence package. It wouldn't get worse if KOMA would offer a command to filter them.

@UlrikeFischer For most template doctors it is big difference whether there would be a documented options or command or whether they have to find out a more tricky workaround.

2:01 PM
@Schweinebacke These template doctors don't read the documentation ;-).

@UlrikeFischer A template a day keeps the doctor away? :)

2:14 PM
@AlanMunn: what about that, I was not crazy! github.com/cereda/arara/blob/master/rules/texindy.yaml
@AlanMunn: Nicola is the reason why arara is still alive today. :)

@PauloCereda So this is an undocumented feature? :)

@AlanMunn It's a new v4 rule, I blame the lack of documentation. :)

@PauloCereda And as I said, mine is way too simple. But I guess it will work for the OP.

@AlanMunn Sure. :)

@PauloCereda reformers will be locked up in chains and left in the tower

2:25 PM
@PauloCereda, @DavidCarlisle opensource.guide makes interesting reading (got pointed there from GitHub for various reasons ...) :)

@JosephWright do we follow any of the guidelines:-) ?

@DavidCarlisle Some of them ;)

@JosephWright yeh just now, got travis ci to update an html file without wiping the entire directory

@DavidCarlisle Interesting that some of them are things we've done for years, some reflect differences in thinking between some of us, etc.
@DavidCarlisle Good :)

@DavidCarlisle oh

2:31 PM
@JosephWright I realise now why it uses --force in its default deploy rule, it does not check out the existing repo at all, its model of "deploy to gh pages" is start with git init, make some files, force push whatever you made. That seems not particularly useful, but it's what it does. However getting the script to clone, then push is easy enough but you have to handle more of the authorisation stuff "by hand"

@PauloCereda H. I don't think that @Johannes_B would agree. More "a template a days calls the doctor to stay" ;-)

@UlrikeFischer ooh :)

@UlrikeFischer Yes, that's true :-)

3:23 PM
@JosephWright PR in progress. :)

3:34 PM
@PauloCereda s/PR/Thesis/

@DavidCarlisle that too

@PauloCereda really? wow? I hope you never finish that, if you did what would we chat about?

4:17 PM
@PauloCereda I ordered a Sony DPT-RP1/B Digital Paper for reading your thesis.
@PauloCereda They said: "Delivery estimate: We need a little more time to provide you with a good estimate. We'll notify you via e-mail as soon as we have an estimated delivery date."
But it's worth waiting for. :-)

@StefanKottwitz ooh

@PauloCereda Proof:

@StefanKottwitz ooh :)
@StefanKottwitz: speaking of purchases, I need to poke you about TUG 2018. :) Ideas, ideas. :)

@PauloCereda That's my solution to finally read all letter sized two-column tugboat journals digitally :-) I can't wait to get it.

@StefanKottwitz :)

4:25 PM
@PauloCereda I started already with long term ideas for TUG 2018. But don't hesitate to poke me about further ideas!

@JosephWright I wonder when we last uploaded grfguide.tex :(

@PauloCereda I will be in Miami during July. Want to visit me? It's not as far away as Europe. ;-)

@DavidCarlisle Yes, thought it was the recent checkin but have literally just run the build script with the current release of l3build and found it is missing there too

@StefanKottwitz I cannot enter the US.
@StefanKottwitz Cool!

@JosephWright ctan doesn't seem to have the file at all so I'd guess it's never been in since we switched to l3build and/or latex-graphics

4:33 PM
@StefanKottwitz By the way, the SELECT ALL button on LC jumps to the top of the page. If you can reproduce it, can you have a look at it when you got time?

@DavidCarlisle Could well be
@DavidCarlisle Will be fixed by the end of my train journey later today

@JosephWright I tihink I'm going to have to install texlive in my travis setup (I suppose it's not that much slower than apt-getting the ubuntu one anyway) it seems to come with texlive 2009 but I want the stuff to fix the epoch so pdf shows no diff if all I have done is tinker with the travis.yml file. so need texlive 2016 or 17 I think

@DavidCarlisle Yes, sounds sensible

@Johannes_B It works for me, I tested on latex.org/forum/viewtopic.php?f=46&t=29874

@DavidCarlisle With caching it's not an issue

4:36 PM
@StefanKottwitz Probably my outdated system then. Sorry for the noise :-)

@JosephWright The scary thing is it would work and be fast enough to appear sensible but building an entire clean VM from scratch and installling dozens of software packages just so I can convert one xml file to one html (or pdf) file and check it in seems scarily over-engineered but perhaps I'm just not used to a "modern workflow" :-)

@PauloCereda I searched for Miami TeX.SX users, guess who I found

@StefanKottwitz LOL

@DavidCarlisle Probably, but Travis-CI is mainly for testing
@DavidCarlisle Have you tried the Trusty images?

4:58 PM
@JosephWright I thought the banner meant I was using them already, but I see I'm using "precise" do you know what trusty has for tex?

@DavidCarlisle trusty packages? :)

@PauloCereda stick to windows, this linux talk is beyond you

@DavidCarlisle oh :(

@DavidCarlisle packages.ubuntu.com/trusty/tex suggests 2013

@JosephWright so I won't bother then, simpler to steal your yaml for caching tl2017 :-)

5:04 PM
@DavidCarlisle ooh

@DavidCarlisle Sounds like a plan
@DavidCarlisle I think everyone using Travis-CI for TeX testing is going to do the same
@DavidCarlisle, @PauloCereda I'm thinking for l3build using a slightly more 'web 2' workflow: I'm really going to try to open issues for most changes

@JosephWright web 2? :) I am lost. :)

@PauloCereda Ok, 'Agile'

@JosephWright oh no, jargon. :)

@PauloCereda Don't worry, I don't understand any of it

5:08 PM
@JosephWright Who does? :)

7 mins ago, by David Carlisle
@PauloCereda stick to windows, this linux talk is beyond you

@DavidCarlisle /sob

@PauloCereda glad to be of assistance

1 hour later…
6:16 PM
@DavidCarlisle @Moriambar @barbarabeeton So you are ready for the reverse double backslash bug?
0

I used this to make pdf files from 8 images I had. I think with little bit modification this will work for you. My latex doesnt have arrow that you're looking for. So you have to add them. \\documentclass{article} \\usepackage[utf8]{inputenc} \\usepackage{graphicx} \\usepackage{subcaption} \\set...

1

I've configured Mathjax on my Github pages site and it can render some equations properly but not others. It can render properly the following equations: \$\frac{1}{n^{2}} \$ \$J(\theta) = \frac{1}{2m}[\sum_{i=1}^m(h_\theta (x^{(i)}) - y^{(i)})^2 + \lambda\sum_{j=1}^n\theta^2_j] \$ But...

Found on SO though.

@StefanKottwitz -- oh, nasty word, nasty word. i've seen that before, actually. the processor that i need to use in preparing ams web pages, in particular for the faq, does exactly this in the initial posting. (if the content is corrected, or something added, it isn't repeated, so if it's caught right away, the posting will be correct. but if it's not, ... i'm not registered on so. who here is?

@barbarabeeton trying a sql data query on SO looking for this timed out :-o
My first intention was to verify if SO was affected by the same issue, so I found the other issue

@StefanKottwitz -- definitely depressing. @JosephWright -- did you catch this? the powers really need to look into how backslashes are being handled.

6:33 PM
@StefanKottwitz ok. Lately I've not been here 'cause I tried writing a bit. That did not go well
@barbarabeeton I fear the worse

@StefanKottwitz -- i think the second item you link to could have been correct when it was answered peter krautzberger does know tex, and he is to mathjax equivalent to what @DavidCarlisle is to latex.
@Moriambar -- so do i, but i'm trying to ignore that right now. i'm still trying to rewrite my tug2017 paper so that it can go into the proceedings issue of tugboat; you will observe that i'm woefully behind schedule ...

@barbarabeeton I see and understand
anyway, should you need me I'll be around, just ask ;)

@barbarabeeton @StefanKottwitz it's not clear that the gh one is wrong (I'd have to check)

@DavidCarlisle -- i suspect that what's actually on github is correct, that the backslashes gpt doubled on so ... and that there are more ... any idea what other sx or so sites might have tex code in questions?

1

I have latex string of a system equation: \\left\\{ \\begin{array} { l } { y = - \\frac { 5} { x } } \\\\ { y = - 5} \\end{array} \\right. but i don't know how to input or convert it into wolframalpha. Anyone know how to do it?

Quoting by an additional backslash can have a proper reason within another language. But then, I guess, for all backslashes, not just for some
I don't find further such posts with a doubled backslash though, just good quoting backslashes

7:33 PM
@ChristianHupfer This could have been MDT :-) tex.stackexchange.com/questions/376725/…

7:50 PM
@Johannes_B Is there any difference actually? ;-)

8:28 PM
@DavidCarlisle Issue with grfguide.tex was actually pretty recent

@JosephWright yes must say it seemed surprising (not least because the ctan checks for pdf doc without source have been there before I think)

@DavidCarlisle Also as I think this was one of my test cases when getting l3build usable with LaTeX2e

@barbarabeeton it isn't clear the doubled \ is wrong (and I haven't had chance to check the github pages markdown syntax)

@DavidCarlisle I read over some of those 'best practice' things on my train journey: we probably do a surprising number of them :)

@JosephWright haven't had time to look today but I saw the landing page you linked to, I have a feeling I read those before in some version, it looked familiar

8:34 PM
@PauloCereda Discrimination! Look at the rule that I found for the rabbit hash mayor election: "Candidates may be biped, triped or quadruped and must be able to chase a rabbit from their home to the Rabbit Hash town center within one hour’s time. (Ducks and geese are still excluded)." (rabbithashhistsoc.org/rabbit-hash-seeks-new-mayor).

@DavidCarlisle -- please tell me how \\documentclass can be correct. also, to quote from the second example that @StefanKottwitz quoted, "It can render properly the following equations: \$\frac{1}{n^{2}} \$ ". since it's stated in the answer that using  rather than \{...\] works, this one can't be correct either. and @StefanKottwitz has quoted a third one, where there are four backslashes in a row, in a string that begins with \\left. something has definitely gone flaky.
i won't blame a misguided mass update, but certainly the original posters wouldn't have used those backslashes in that way.

@barbarabeeton the question is about typing into a wiki page so the question is does \[ have meaning in the wiki in which case it will need \\[
@barbarabeeton the first one with all the \\  is clearly wrong but it could just as easily be user error there is no indication that it's been changed by any back end update

@DavidCarlisle -- okay, i'll buy that, but i've found the wikis here that are allegedly created to inform the production group how to code certain things show double backslashes in the "regular output". all of which says to me that wikis are probably not really suitable for guiding (la)tex input. i'm simply depressed by the whole mess.

@barbarabeeton it's confusing and people have been getting it wrong since before wikis and before stackexchange, look at any shell script running sed over tex input, the number of \  you require depends on which shell you are using, whether you use single or double quotes around the expression, the phase of the moon and several other factors, basically everyone just doubles them until they work, getting the quoting right in wikis isn't any different.

8:52 PM
@DavidCarlisle -- okay, i'll buy that too. however, having been burned by a tool required here for preparing web pages, that definitely (reproducibly) doubles backslashes without warning, i'm gunshy. the problem has been demonstrated to the person in charge of maintaining the thing -- who refuses to do anything about it since there's a workaround, namely redoing the input. i dislike unpredictable tools.

@barbarabeeton it's probably predictable, check the phase of the moon!

@DavidCarlisle and the pound exchange rate

@StefanKottwitz monotonic decreasing functions typically don't generate apparently random behaviour
3

@DavidCarlisle It's about speed and acceleration, not direction.
Well, tomorrow it may be our politicians causing unpredictable stuff.

9:22 PM
@StefanKottwitz \let\tomorrow\the_past ;-)

@ChristianHupfer missing \$

@DavidCarlisle No... I did not want to write \thepast` ;-) It's pseudo code

9:39 PM
@ChristianHupfer I thought it was expl3 but just one underscore is way too few