« first day (2840 days earlier)      last day (982 days later) » 
00:00 - 21:0021:00 - 00:00

12:01 AM
So I have math equations in my paper
I would like to display the Mathematica/Maple/Matlab code on mouseover(well some way)
so my pdf document still looks the same
@samcarter At least for Chromium Tampermonkey isn't needed to run these, you can install the userscripts as extensions by dragging them into chrome://extensions
is the standard way to do this in latex such that it translates over to pdf documents?
ping me if you have any ideas
@William "mouseover" effects in pdf tend to be highly pdf viewer dependent and don't really work well, are you sure you want pdf rather than html as your document format?
lol yeah I have decided to do HTML+MathJax at this point
but trying really hard to use Latex
@DavidCarlisle just trying to get a feel what latex can do I guess
not widely supported though in pdf reader unfortunately
@William latex can do lots of things but it can not generate pdf files that work like html.
12:06 AM
sometimes pdfs have notes
so maybe something like that
@William as I say, you can do some things but often they only work in acrobat and far less people use acrobat than they used to
Negative with HTML is well inline images really don't work very well
base64 images is well a really bad idea IMO
So then you have to decide pdf document one file
or interactive HTML document multiple files
and then I go well I guess I could just host the code on a server that the pdf links to
@William svg
svg yes if you have the luxury and time of creating those
mathematica doesn't export svg very well
@DavidCarlisle hmm probably should do more svg research thank you
@William if you can get plots as eps or pdf they ought to convert to svg well enough
12:10 AM
so rasterize the image from mathematica and use a command line tool to convert to svg
hmm I think I have done this before but don't remember thank you
@William but these days does having a document in a single file help much? It's not as if you have to post it on a floppy disk
it helps for emailing. Yes its not that big of a deal but pdf has a gauranteed way to print
@William no not rasterize it, imagemagic or similar tools should be able to convert the path data from eps to svg
@William not if you fil it with acrobat-specific javascript extensions, it may not work at all
:P that is a fair point
@William I will grant you that printing pdf is more stable than printing html, but I don't think email is really a problem, I've email zip files containing a few thousand html files + assorted images and stylesheets and recipients manage OK.
12:16 AM
I'm okay with going with HTML+MathJax except for my frustration with printing
papers need pixel perfect accuracy
and length requirements
this shows what can be done
@DavidCarlisle This feels different. Maybe a good thing, but I'm much less inclined to browse questions now and must less likely to stay on the site for any length of time. It just feels icky and it is uncomfortably glarey. The exception appears to be chat. The three columns is also absurd. Half my screen is taken up by navigation and intrusive buttons, especially if I view questions in a particular tag. I'm mostly motivated to stop looking at it.
@PauloCereda Llongyfarchiadau!
@cfr "I realise the middle section is just the boring bit where the users chat about their interests and the sidebars are the interesting places where stackexchange can insert revenue earning links" :-)
@DavidCarlisle It doesn't matter if nobody stays to look at the revenue-earning links?
@DavidCarlisle pdf is the standard document type for academic papers and well because code isn't encouraged or a requirement if all you have is the pdf its really really difficult to work backwards sometimes
@cfr It's pretty bad actually Jon Ericson admitted it didn't look good in his comment under the answer where I wrote that.
@William maybe times have changed but when I was in academia all you'd have was the paper (on paper) so unless the text gave a link to a code repository the same was always true.
12:24 AM
sure but were are in the digital age
come on this stuff could help others
research that is nonrepeatable seems silly to me
@cfr But your avatar looks also threedimensional now. Nice! ;-)
@samcarter the new site is bad, but it's not that bad:-)
@William yes but latex can only make pdf, it can not redefine what pdf is or what non standard features pdf viewers support, so you are asking for features in the wrong forum, latex can make (with more or less effort, depending) more or less any pdf file, but it can't make a pdf file with all the interactivity of a web page because it's pdf.
pdf has/had this fuctionality just it was added to the other readers
but yes you are right
it wasn't*
12:52 AM
Is there anything that takes a subset of latex commands(unambiguous ones) and evaluates it through a mathematical software?
It wouldn't be to hard to write I would think
multy character names are a pain though
@William why are mult-letter names a problem? (in general it's easier to go in the other direction, from a computational system to latex, but several people have gone the other way, usually with highly constrained project-specific markup.
@DavidCarlisle well because for larger equations the formatting is everything
well just comparing
\sum _{\text{testing}=0}^{10} \text{testing}^2
its not a problem that prevents you from doing anything but not how I would design the language
I see why they did it this way
I think
@William well I wouldn't use \text for a variable name
okay what then?
it is still a command probably
Mathematica has a BoxData form which can be converted between the 2 essentially
you wouldn't want to write it though
Q: Two letter variable names

mbqIn math mode, TeX assumes "hidden multiplication" (so that two variable names put nearby have invisible multiplication between), so that, AFAICT, expressions like $ABC$ are rendered with small distances between letters. Now, what to do if I have a two-letter variable name, like for instance TP or...

okay macros for them
@William \text is for text (embedded natural language) i'd use \mathrm not that would help conversion, but if you want to use conversion then use an input something like \rangesum{testing}{0}{10}{{#1}^2} or some other prefix form that is more amenable to calcuation.
1:08 AM
@DavidCarlisle mathjax doesn't like that but interesting. Why do you say rangesum is more amendable?
oh rangesum
is your own syntax
yeah I'm trying to work backwards fro mthe display form to something that is evaluatable
@William mathjax would accept a definition of that just as tex would, but the problem with traditional sumation notation is you never know what you will find at the bottom, \sum_0^{10} i^2, \sum{i=0}^{10}i^2, \sum_{i\in[0,10]} i^2, ... all mean the same thing, so you need heuristics to parse the form and guess the structure
You are right Mathematica
gets around this by having only a set number of options
@William in our manual we have (~300000) math expressions all marked up in mathml from which we derive print forms but also code (C or fortran mostly) it is far easier to start from a constrained system and generate the display than the other way round, although of course authors prefer to use an unconstrained form and expect that "something" will understand it
I mean sort of easier
mathematica boxdata does both
if you have the display form in mathematica you also have the evaluatable code in many but not all cases
mathematica let's you enter the computational form and then it converts to a display form that goes backwards
@William as I say many projects successfully use latex as computational input but they specify the subset they use and can handle. the real killer is spacing is ` A \hspace{-2em} B` A times B or B times A ...
1:18 AM
do you know of a parser for latex that makes this feasible
to a layman
I am looknig one sec
@William it varies I'd be tempted just to use tex
@William but it's also 2 in the morning and I'm going to be, nice chatting but I'm off now:-)
night late here also
sorry first day with latex noobie questions
but I have it running now
1:57 AM
@William We all were newbies once!
2:11 AM
Hi 👋. I know this is the wrong community, but here I know more people and I feel better :). With all this topic of changing style I discovered that in Math.SE the zoom trigger of MathJax is a bit ugly (Prob. 1). Where can I report this issue or how can I solve it? Also there is a problem with some equations (Prob. 2). Again take that link, and please see the image below:
I think the two problems are in fact one, but I don't know how change the configuration to restore back the old style. Any ideas?
P.S: question
5 hours later…
6:46 AM
@manooooh you probably can't get the old style back, just report it at the meta site for math.se or the main meta.se
7:21 AM
If that's not the best place to report it, please let me know what is.
Are we/many of us really boycotting SE for the new design (tex.meta.stackexchange.com/a/7799/117050)?
@Skillmon I won't be, at least not at the moment: the problem with 'let's set up an alternative' is that someone has to do the 'back end' work, and the pool of TeX people is already small and they are busy ...
If so, who's boycotting and who not? Are there any real chances that there will be another site where we can help people? I feel like boycotting only makes sense if there is a viable alternative, after all one of the greatest "recent" developments in the TeX world is in fact the higher QA rate established by this site.
@Skillmon The problem is partly setting up a site that has the design and 'back end' support: the bit of TeX-sx that is free is the content, the old design belongs to the Powers.
@Skillmon I think @DavidCarlisle has it spot on in various comments on the meta thread from the Powers: it's not change per se that's the issue, it's the fact it is a pretty rubbish new layout. It doesn't help when we are told 'this is in your long-term interests' when clearly it's not.
7:36 AM
@JosephWright I don't care much for the old design. A new design would be fine, if it just doesn't look awful. In fact I'd prefer the looks of a gopher space over the new design (but I'm a heavy terminal user -- firefox is one of the few GUI programs on my production laptop).
@JosephWright I was thinking whether I should ask to make TeX.SX usable via terminal browsers like lynx or the like.
@Skillmon To a good first approximation, no-one uses Lynx ;)
@JosephWright I know.
@Skillmon Sure, I get that: I meant that we'd still need someone to do the design work ... it's all doable, but it's not a trivial task
@JosephWright that's correct, I'm afraid.
As for the backend there seem to be alternatives, e.g. scoold.com, see also meta.stackexchange.com/questions/2267/stack-exchange-clones.
7:50 AM
@TeXnician I meant more the server(s), etc., although of course yes, one does need the stack
@JosephWright Yes, of course. There are many things that one would need to consider: hosting, server-side software, design, data migration etc. So this is nothing which could happen at a moment's notice.
@Skillmon no:-)
@Skillmon if it looked like comp.text.tex in the the emacs gnus newsreader I'd be happy:-)
8:43 AM
@TeXnician setting up a server doesn't take that long. A friend of mine and I set up a gitea server (after MS bought github) in two days while primarily working on other stuff. I'm still unsure whether I should transfer my packages to the new server or not (due to bug reports is easier with github as there are more registered users, compared to two on our server...)
@Skillmon Well, I have some experiences with servers too. Setting up a Gitlab instance is about 6 hours, setting up a webserver is about 2 hours and setting up a mailserver really depends, but what I meant is more or less the problem that hosting with all its extra tasks like making backups requires money and that is a crucial thing in the open source world.
@TeXnician yep, I agree on the money part :)
1 hour later…
9:55 AM
@DavidCarlisle thanks! I will do it later
@cfr Cwac! <3
@Canageek Thanks, pal! :)
@DavidCarlisle ^^
@PauloCereda For the record, regarding openjdk on the mac: After looking around a bit more, I found that I can just unpack the openjdk distribution (for macOS, of course) in /Library/Java/JavaVirtualMachines/, and then the built-in java command (i.e., /usr/bin/java) Just Works. 8-) This is not well explained on the openjdk web site: They just point to oracle's docs, which merely tell you how to install from their .pkg files.
@HaraldHanche-Olsen Cool! :) Now that you mentioned it, this is how sdkman does: it downloads the chosen JRE/JDK into a managed folder and handles symlinks as to make the chosen vendor/version available. Great finding!
I never thought too much about it, but it makes sense! :)
@PauloCereda Those aren't proper floppy disks – real floppy disks have no hard shell, and they're way bigger.
@HaraldHanche-Olsen Oh you mean those big ones!
10:11 AM
A floppy disk is a disk storage medium composed of a disk of thin and flexible magnetic storage medium encased in a rectangular plastic carrier. It is read and written using a floppy disk drive (FDD). Floppy disks were an almost universal data format from the late 1970s into the 1990s, used at first as a primary data storage mechanism, and later mostly as a file transfer system as part of what became known as "sneakernet". Work on a drive that led to the world's first floppy disk and disk drive began in 1967 at a San Jose (CA) IBM facility, and introduced into the market in an 8-inch format in...
@PauloCereda The one on the far left, yes.
@HaraldHanche-Olsen ooh
11:01 AM
@Skillmon well I don't see myself quitting friends and tex questions for good only because of a site design (which looks quite ok to me after I added a few of the css we developed). And I don't want to come back in disguise -- @samcarter or @DavidCarlisle could put on masks but it wouldn't work with my avatar. So threatening to quit would be only a bluff and I'm a quite bad at such games. Beside this I don't really believe in "after an exit everything will be much better".
@HaraldHanche-Olsen wow, who 3D-printed the save icon?
@UlrikeFischer neither am I, as I don't see any viable alternative and answering TeX questions is such a great way of procrastinating.
user image
@UlrikeFischer ^^^ You can also have masks
11:17 AM
@samcarter ooh ;-) but I think the mask of the black duck should be yellow (and it should look in the other direction ;-).
@samcarter ooh a ninja duck
@samcarter stored for karneval ;-). I hope I don't forget it ...
@UlrikeFischer ^^^ For karneval
@samcarter ooh
11:32 AM
23 hours ago, by Harald Hanche-Olsen
@PauloCereda Shall we keep going until we get to ω?
@PauloCereda Maybe there needs to be a special duck for Brazilian carnival
@UlrikeFischer just like spy vs spy ;-)
@PauloCereda did you get a mail yesterday?
@UlrikeFischer I did, and I am working on it. :)
11:47 AM
@PauloCereda great ;-)
12:02 PM
@samcarter you could add it to this list
@DavidCarlisle I would have expected to see a link to the menu as reply to chat.stackexchange.com/transcript/message/46055954#46055954
@UlrikeFischer: vimeo.com/284348495 :)
@PauloCereda great!!
@samcarter ^^^^^^^
12:19 PM
@UlrikeFischer @PauloCereda Amazing!!!!!!!!!!
12:43 PM
@DavidCarlisle I found the culprit of the overprinting comments: my beloved "ignore user" browser addon -- guess I'll have to choose between reading comments and ignoring certain users.
@DavidCarlisle I have removed my comments from your answer tex.meta.stackexchange.com/a/7806/36296
@CarLaTeX Maybe the behaviour you describe in tex.meta.stackexchange.com/a/7810/36296 is the implementation of the "be-nice" campaign: no post will be deleted if nobody can access the button :)
1:08 PM
@UlrikeFischer Same here: I don't particularly like the design changes, but I think over time they'll get the worse of the rough edges sorted (particularly if like me you turn off the left-hand sidebar), and the set of people here is already good
@samcarter On the topic of "Make TeX.SX look nice again!", do you know how to implement a script that removes the border around the 'Active', 'Featured', 'Hot', 'Week', 'Month' buttons? This was a point raised by @Skillmon and Joe Friend tagged it 'status-bydesign'... see tex.meta.stackexchange.com/a/7796/128068
@JosephWright I personally don't mind the left side bar: I normally have only one window on top and in full screen, and then I like a large left margin. But the right bar could be smaller.
@UlrikeFischer Sure: depends on how you use your PC
@samcarter It seems disabling 'border: 1px solid transparent;' does the trick. But I can't work out how to make a Tampermonkey script to do this...
@Milo Try .s-btn {border-style: none !important}
1:14 PM
@Skillmon I think you've got to ask yourself if that will work for people who want answers: they already hit TeX-sx a lot as it has lots of existing answers
@UlrikeFischer Thank you! Will post this on Make TeX.SX look nice again! and credit you :)
@JosephWright Well there'd be a phase where people who want answers wouldn't get them as fast, but after a bit of time the new site could take over completely. Especially if we migrate all the existing QAs from TeX.SX. (all just theoretically speaking)
@samcarter OK I deleted my reply as well, thanks
With my new background image I get now this:
@JosephWright what's that for a chemical reaction? ^^^ ;-)
@UlrikeFischer I am curious: what did you choose as background?
1:21 PM
@samcarter the one with the duck. I was so used to it ...
The oxidated duck. R.I.P.
@UlrikeFischer A classic!
1:37 PM
@UlrikeFischer -- some of us are unable to install the changes because our office workstations are "locked down" -- can be used flexibly, but nothing can be installed. (my laptop is so woefully antique that even getting some things to run at all is problematic. it must be replaced before i retire early next year.) i guess i'll just continue to whimper.
@barbarabeeton even in that setting I'm surprised you can't use a user-script nothing needs to be installed system wide: they are just javascript files in your own area (and for some browsers an add on such as greasemonkey, which again can just live in your area) But I'm not using any of those scripts anyway, I think we need to see what others see:-)
2:03 PM
@DavidCarlisle seeing all the different screenshots I doubt that I can see what other see anyway;-). And I just played around with a screen reader and the thing is reading text I never noticed on my screen before.
@DavidCarlisle I saw that you wrote about mathml last night. I'm currently collecting ideas how to convert a sensible subset of latex-math to mathml with as less fuss as possible. If you have something can you sent it to me? There is no need to hurry, currently I even don't know if there is a way to store it in the pdf so that the screenreader could do something with it.
2:36 PM
@samcarter So it's a feature, not a bug! :)
@UlrikeFischer Ross gave a talk at a tug meeting (I guess Damstadt) about getting mathml into the pdf tagging, for conversion tex-mathml I'd use latexml or tex4ht or mathjax all of which make a reasonable attempt
2:59 PM
@DavidCarlisle Ross imho tags every single symbol. That's not really pratical, but I don't know yet if one can store an equation as a whole. Thanks for the list.
@UlrikeFischer well if you tag the equation as a whole that isn't mathml (but is I think what he was suggesting in the most recent plan of action) tagging every symbol isn't that impractical, it is what mathjax is doing everytime anyone types any tex into math.sx
3:13 PM
@DavidCarlisle the problem if you tag in the pagestream you also have to handle all the things people add to make an equation look "nice", like colors, spaces, special fonts etc.
@DavidCarlisle and with "as a whole" I meant that I would like to add some reference /Alt = whole mathml of the complete equation around the equation instead of tons of /mo <</MCID 8 /ActualText<FEFF2062>/Alt( times )>>
3:35 PM
@UlrikeFischer ah make mathml at that level but put it all in a single of tag thing, I think Ross was claiming in his talk that using the tex sources as the actualtext was actually what people wanted (and is presumably easier) but i'm interested in the mathml version but it may need to be in a couple of weeks
@DavidCarlisle I made some tests regarding "tex source in the /Alt" text. github.com/u-fischer/tagpdf/blob/master/source/examples/…. The math read out is not so bad, but a structured mathml would probably be better -- if there is a accessible tool/workflow which can extract and use it.
@AdamLiter oh my, I have to revisit this one too!
@PauloCereda I think I might have some time this weekend where I can help. I can try to look back through my emails with you, too. I'm pretty sure you already had a working knitr rule with arara 4.0 at some point in time ...
@AdamLiter I think it is just a matter of using getCommand to properly invoke knitr...
4:46 PM
How about reopening tex.stackexchange.com/questions/34406/… ? Couldn't the LaTeX team suggest synonyms with a correct spelling for the time being as a first step and then deprecate the old spelling in the next step, and then remove the wrong spelling altogether? After all, LaTeX is not that dependent on whatever Adobe did in the past...
Sorry for just shouting in that room, but has there been an update of the fp package?
@PauloCereda Yes, I think so. But then one will also (I think?) need custom (pdf|xe|lua)latex rules that only use the file base name and not the file name, since the standard workflow with knitr is to edit a .Rnw file, compile the .Rnw file to a .tex file with knitr, and then compile the .tex file to PDF with a LaTeX engine.
@marmot don't think so. Why?
Because of https://tex.stackexchange.com/q/445454/121799 It seems that both `min` and `max` do not really work with `\usepackage{fp}
\usetikzlibrary{fixedpointarithmetic}` ...
@PauloCereda Or is it possible to inject the correct file name into the (pdf|xe|lua)latex invocation using getBasename and the files parameter?
5:01 PM
@AdamLiter It is also possible. :)
@UlrikeFischer Yes, they are certainly related. Thanks! Maybe you should announce that on the main site ...
@user49915 actually, we probably could, have more space than we had in 1993 for duplication of names. But a question here isn't the place to suggest a change, either the latex-l list or a github issue would be better.
@marmot l3fp is a lot more accurate
@DavidCarlisle Well, if you could turn that in an answer that solves that problem, that would be great. Turns out that I sort of rediscovered what you and @egreg already knew. Bamboo's workaround also works.
@PauloCereda Would it look roughly like this? (I'm just now starting to look at the `arara` 4.0 docs 😊):

% arara: pdflatex: { files: [ getBasename(currentFile) + '.tex' ] }

I'm not sure how you're supposed to do string concatenation.
5:08 PM
@marmot I know nothing, I just saw a notification of a vote on that answer and looked to see what it was about and saw it was same issue as this, I had no recollection of it. It would probably take a bit of work to hook l3fp in the back end of pgf arithmetic, @JosephWright would have been a person to do it but he's opting for the simpler option of re-writing the whole of tikz-pgf so it sits naturally on top of l3fp.
@AdamLiter Oh no, we don't support methods inside parametrized directives. :( I need to understand better what we want to achieve. :)
@PauloCereda Okay, sounds good. Perhaps I'll send you an email this weekend. I have some other stuff I need to work on today. Thanks for the quick discussion though! 🦆
@AdamLiter My pleasure! Sorry for not providing a direct answer, but I am quite in a hurry these days as well.
@DavidCarlisle As long as @JosephWright makes sure that there will be 3D marmots, I cannot see anything wrong with his attitude ;-) But seriously, are you still checking if one of your answers gets upvoted? (I would understand if you checked if someone upvoted your questions, though.) I guess than I have made your life a bit more eciting in the past. ;-)
@marmot I don't always check but I got a silver badge and i didn't remember the question at all so i just had a look and realized it was same issue as you'd been discussing here (which I had been ignoring:-)
5:15 PM
@DavidCarlisle If you mean github.com/latex3/latex2e, then I would submit the issue there.
@user49915 yes (do you think it would help much though?) I would guess that most people who need them just type the characters directly rather than \textg-whatever (but personally it seems a reasonable request to make, whether or not we decide to do it I can't say yet:-)
@DavidCarlisle You got a silver badge? Really? Congrats! Are you going to open a bottle of good honey liquor? ;-)
@marmot Necromancer:-)
@marmot no, perhaps I'll have pineapple pizza though.
@DavidCarlisle because it's a Necromancer badge? Some rotten pineapple and ham? ;-)
@DavidCarlisle Good point. It never occurred to me that one can type them in directly, but one can indeed, you are right. Still, these characters are neither present on my keyboard nor in the lower part of the ASCII table, so, probably, I would continue using the long description \guillem..., similarly to how I use dozens of other characters from the texdoc symbols-a4 document despite their availability in UTF-8.
5:33 PM
@user49915 yes for math symbols I agree the \asciiname form is possibly always going to be useful but for languages using those quotes I'd have thought that people would want something simpler, just as in english you use `foo' rather than \textquoteleft foo\textquoteright I seem to remember that even before utf-8 input babel offers a shortcut "< or some such.
@DavidCarlisle isn't ` part of TeX syntax?
@user49915 we did consider using the correct spelling back then but decided to follow adobe, ltoutenc says:
% The naming conventions in the kernel are not what we would use if we
% were starting from scratch\dots\
% Those defined by DEK (like |\ae| and |\ss|) or by the \TeX{} Users
% Group Technical Working Group on multi-lingual typesetting (like
% |\th| and |\ng|) have short names.  Those which were added to the
% kernel in 1993 and early 1994 are named after their Adobe glyph
% names (like |\guillemotleft| and |\quotedblbase|).  Unfortunately,
% this naming scheme won't work for all glyphs, since some names (like
@Skillmon well not really, but it is in the font metrics of most tex fonts.
@DavidCarlisle Why not use the proper names, but alias the adobe names to point to the same thing?
@DavidCarlisle Really, "sigh". Anyhow, thank you, nice to know. To me, this text seems more internal, considering the implementation of LaTeX rather than its interface to the users. Since the LaTeX team is acutally very active nowaday, they could use the chance to touch also that part.
@DavidCarlisle As for the reporting to github: do you think that I should still provide the full minimal example and the log? (No problem with taking time for that, but it would seem to me kinda cluttering the real issue...)
@user49915 we would probably never remove the old name (would just break existing documents) and as I say I think most people would not be using the name in documents so adding an extra name is not a clear win. surely any document would have something like \newcommand\q[1]{guillemotleft #1guillemotright} ... \q{foo} so the bad names only appear once, most likely in a package file
@user49915 probably no need (you can always say i said you didn't need to, to save Frank telling you off for not providing one:-)
5:50 PM
@user49915 I use either csquotes (\enquote) or keyboard shortcut to insert «». I would never type \guillem...
@UlrikeFischer the name is too long for you to type, being German not used to long words?
@UlrikeFischer I need the ASCII-only variant if you have to stick to ASCII for some reason (say, automatic file processing by external ASCII-only tools). I prefer the ASCII-only variant if I know I view your document in an ASCII-only viewer. Both happens to me seldom (once every couple of months), but it does happen. I think like time I ran into the ASCII vs. utf8 issue was when I used the listings package.
@DavidCarlisle I probably could handle \frenchleftdoublequote but I can never remember if it is \guillemotsleft or \guillemotleft or \guillemetleft or ...
@user49915 well then I would \enquote. Semantic markup is always better than inserting explicit symbols.
@UlrikeFischer Also, the left/right distinction is tricky, as Norwegian uses them «like this» while Danish (and German, I think) »like this«.
Perhaps one should adopt the unicode names and say \leftPointingDoubleAngleQuotationMark and \rightPointingDoubleAngleQuotationMark? Or perhaps not.
In fact, «» and »« do not always occurs within a macro. Consider
*typesetting* C++ code containing "operator<<()" and "operator>>()" in a proportional font rather than in monospace. Yes, I had this situation in the past, and I chose "operator\textless\textless" back then, but today, I'd go for "operator«".
6:06 PM
@user49915 but if you are allowing non-ascii in your c++ wouldn't you just want « in the tex document?
@HaraldHanche-Olsen yes, than another problem (but the various quote are even more difficult here -- that's why I never looked back when csquotes appeared.).
@CarLaTeX did you saw the new video?
@user49915 the long \textzzzz names were added just to give some name and followed the published adobe names to allow people to guess them, but there was never really any intention that they be often used in documents, the assumption was always that packages or document authors would set up some more convenient markup.
@user49915 -- but those aren't guillemets. they're really double less-than or greater-than -- \ll or \gg (although those do require math mode).
@barbarabeeton You mean the symbols I typed in into this very chat?
@user49915 -- exactly, but nested as a single symbol.
6:14 PM
@barbarabeeton These ones are the guillemets to the best of my effort: « (U+00AB), » (U+00BB). I apologize if I typed in something similarly-looking earlier.
@user49915 no I think barbara meant if you want a << for C++ it would be better to use U+226A ≪
@DavidCarlisle -- yes, that's exactlly what i meant.
@barbarabeeton although << is neither a quote nor a less than operation in C++ it's more of a double headed arrow (it means shift) so neither is quite "right" :-)
@DavidCarlisle -- okay. then that sounds like a candidate for a submission to unicode. there are plenty of unicodes for apl, so why not for c++?
Thanks for driving my attention to them. I was not aware that these symbols ≪≫ are also available as Unicode symbols, but you are right, they are, and, at least "operator≪" looks better in my browser than "operator«". But the former one occupies more horizontal space, which you may not like.
Anyhow, that's a matter of font choice rather than the LaTeX shorthand. Unicode probably does not have a proper symbol for the "output operator".
6:22 PM
@barbarabeeton well the symbols in apl were defined to be symbols (single characters) but the C++ operator really is <<
@DavidCarlisle Yes, really it is "operator<<". And it really looked ugly to me if typeset exactly like that in non-monospace font in a paper or an article. Given Barbara's suggestion, I'd say "operator≪" in a single-column output (or when I have horizontal space) and "operator«" in a double-column output (or when I'm short on horizontal space).
@user49915 you should be consistent. Don't change amid a document.
@Skillmon Of course.
@user49915 it just sounded as you would and I felt required to stop madness :)
@DavidCarlisle Alternatively, one may use the left guillemet « for commenting: the source code you typeset is on the left, and, on the right in color, after «, you typeset the explanation. E.g.:
iseseq(x◦y) = and(iseseq(x), iseseq(y)) « the concatenation of two sequences is empty iff both constituents are empty.
(...with colored part including and after «. I apologize for poor typesetting here in this chat.)
6:43 PM
@Skillmon You are right. Yes, I got you.
6:57 PM
@DavidCarlisle I decided to bother the folks at github anyway; I apologize a lot about that.
@UlrikeFischer do you mean the one of the TUG meeting? Yes, of course!
@user49915 that would be me with a different hat then:-)
@CarLaTeX Is it just me or is your logo broken on the very left panel?
@DavidCarlisle Hm, as far as \guillem{e|o}tleft is concerned, your hat would probably lie on the left side ;-).
@CarLaTeX No I mean the starred one on the right from Paulo.
7:18 PM
@UlrikeFischer I didn't see it, gorgeous, LOL
I wonder how long it'll take when the chat looks ugly :( Trying to read the stuff on the main site is painful :(
@user49915 @egreg just agreed with your suggestion, so now I'll have to disagree.
@Skillmon apply some of the css suggested in meta. My background image shows now the duck chess font and looks great ;-)
@UlrikeFischer I don't feel like installing Greasemonkey just for one site, my Firefox is cluttered enough by the 200 open tabs.
@Skillmon you can probably apply user css without any addon, something like this, I may try later ffeathers.wordpress.com/2013/03/10/… (that's a bit old but was first thing that came up)
7:25 PM
@Skillmon -- how do you manage to get 200 open tabs? mine "shuts down" when i get more than about 30. (well, "shuts down" isn't exactly accurate. every one goes all black, and i have to shut everything down to recover. bummer.)
@DavidCarlisle Well, he only partially agreed. To be 100% complementary, you can disagree with his part but agree with everything else.
@Skillmon Can't firefox load a user.css?
@barbarabeeton dunno, having many open tabs is something I do all the time. It started when I encountered Pentadactyl the first time. Now I can't live without so many tabs. Sadly Pentadactyl ceased working, so I'm using the slightly inferior Tridactyl. It doesn't perform that bad.
@UlrikeFischer probably it can, but I never tried up until now :)
@UlrikeFischer usercontent.css, yes, here is more modern doc (@Skillmon) probably should post a non-user script answer to that meta question kb.mozillazine.org/index.php?title=UserContent.css
actually that's old doc again
@DavidCarlisle does it work (so did you try already)?
7:37 PM
@PauloCereda I procrastinated ... 🤪 github.com/cereda/arara/issues/112
@PauloCereda No rush to respond. Take your time! I know you're busy.
getCommand('Rscript', '-e', 'library(knitr); knit("' + file + '")');
@AdamLiter ^^^ try this. :)
@AdamLiter I will try to provide a comprehensive answer for you tomorrow. But for now, try:
identifier: knitr
name: knitr
- Adam Liter
- name: knitr
  command: >
        return getCommand('Rscript', '-e', 'library(knitr); knit("' + file + '")');
arguments: []
@PauloCereda Thanks! That works. I should have asked you sooner. 😔😜
@AdamLiter No, it is my fault. I should explain things better. :( Do not worry, I will cover every aspect of your issue. :)
cout << "Hello world!";
@PauloCereda Lol, I also don't know anything about Java, so it's probably me ... But thanks! I'll update the issue sometime today with more info on the other options to knit().
7:47 PM
@AdamLiter No worries, we will write a nice knitr rule!
@manooooh Some folks would typeset it as cout≪"Hello world!" or cout«"Hello world!", depending on the fonts used and the horizontal space available.
@DavidCarlisle I don't get it to work :(
@user49915 I am the person who wants the best for the programmer ;)
@Skillmon no my first attempt didn't work either
@manooooh I am the person who wants the best for the user ;)
@user49915 where would you put the open braces in a function? `void myFunc(...) {` or `void myFunc(...)
@user49915 ya... ehm.. huh... yeah!!
@Skillmon probably I won't do it then, cosmetic customization that require you to turn down security features aren't much fun.
@DavidCarlisle still not working :( I think I give up for today.
@UlrikeFischer @PauloCereda I'm watching "Never Say Never Again" right now :)
8:00 PM
I want you to know that you are the most :)
@samcarter oh :)
@samcarter we too ;-). Btw I added the new background image for my laptop.
@samcarter but the music is wrong ;-(.
@PauloCereda Something today put me in the mood for 007 - I wonder what that could have been :)
@UlrikeFischer :) Glad you like the chess logo!
@samcarter ooh a secret inspiration
@UlrikeFischer Luckily I now know which video to watch for right music
@PauloCereda The best kind of inspiration.
8:28 PM
I see the maths community is not happy about the new look: math.meta.stackexchange.com/questions/28842/…
8:38 PM
Good evening to everybody. I have upvoted all your answers and lot of comments on meta. The new site is very horrendous.
I hope that something will change as soon as possible. I am waiting for your instructions and in mass we will boycott the site :-)
^^^ An update has been made to the site theme. Blue tags are now green. And the background colour of the content area is now off-white.
@JosephWright -- and their "new look" question has ever more downvotes than ours. (although as a proportion of participants, ours may be higher.)
@JosephWright LOL
@DavidCarlisle they are working. The filename must be userContent.css not UserContent.css. After that I got some of the css working.
The only thing our dear Joe Friend does is edit answers... He is definitely unfriendly
8:55 PM
@manooooh I'm very confident he misspelled his name during profile creation. There slipped an r into it.
@Skillmon after you set the preferences in about: ?
I just learned about the adjective egregious.
@DavidCarlisle no, I reverted it to the default.
@JosephWright find "Joe Friend"; 3 of the 4 times he commented was for edit messages... and there are several answers!!
@Skillmon hmm my file was userContent.css but it didn't seem to work, i may try later.
@-moz-document domain(tex.stackexchange.com) {
    .post-tag::before { content: "{"; }
    .post-tag::after { content: "}"; }
    .post-tag {
        border-width: 0pt !important;
        color: #444 !important;
        background-color: transparent !important;
        padding: 0pt !important;
    div#sidebar { max-width: 30%; }
    div#mainbar { min-width: calc(70% - 24px); }
    .question-hyperlink {
        font-size: 18px;
    .s-btn {
            border-style: none !important;
@DavidCarlisle current contents of my userContents.css. At least the tags in curly braces are working.
8:57 PM
@Skillmon phew no duck images
@DavidCarlisle reduced sidebar width doesn't seem to work. Questions' font size doesn't change either. But the borders around the newest | featured | ... thingy are gone.
@Skillmon the widths must be subject to a @ media query and I'm not sure of the priority of those with respect to @-moz-document I may try again...
00:00 - 21:0021:00 - 00:00

« first day (2840 days earlier)      last day (982 days later) »