« first day (1577 days earlier)      last day (2197 days later) » 

12:22 AM
@DavidCarlisle I finally went with \tikzmark; perhaps tomorrow I'll rethink things with a fresh mind to see if I can avoid TikZ altogether.
@GonzaloMedina avoiding tikz is always a worthy goal:-)
@DavidCarlisle Agreed! It's powerful, but sometimes overused. I include myself in those who tend to overuse it :-)
Tomorrow I'll try the \pdfsavepos approach.
Any community opinions supporting me suggestion?
A: Notifying the user when they are about to post a "give me the codez"-question

WernerThis is going to be difficult, since it doesn't fall within the generic realm of StackExchange - site-specific features. Instead I propose we create a generic question that has a please-do-this-for-me flavour and answer it in a generic way, pointing to some FAQs about usages for packages, other ...

@Werner I'm somehow not "for". But I'm not sure I'm able to explain why... :-/
1:08 AM
I gotta love package bugs :-(
1:21 AM
@yo' not mine I assume
@DavidCarlisle not this time
@yo' As I suspected!
but it's really annoying: geometry doesn't give me symmetric result for a twoside document if I ever use \newgeometry. And I don't even know what is causiing it because a very simple MWE works like charm. But my twocolumn switching and stuff does not.
ah, maybe I got it.
@yo' I never have much faith in newgeometry:-) why do you want to change the page size mid document?
@DavidCarlisle because the frontmatter is partially in twocolumn and we want it slightly wider.
1:25 AM
@yo' I'd use a list environment like quote but with -ve indent it's much less intrusive
I still don't know why I get three different values of \evensidemargin with two layouts, but I'll probably live with that
@DavidCarlisle sounds like a mess....
@yo' compared to newgeometry that has to make scary changes everywhere quote has a 2-line definition and \begin{quote}...\end{quote} just works and the stuff has larger margins
@DavidCarlisle well, my env is a group, so I could even just increase \hsize manually, no?
it'll be tough to center the line that divides the columns, but it'll be ok
@yo' Not if you have any nested itemize etc as if you just change hsize they won't know. that's all quote does but it also changes \linewidth and \@totalleftmargin and other stuff so any latex display construct knows where to start.
> \quote=\long macro:
->\list {}{\rightmargin \leftmargin }\item \relax .
l.6 \show\quote

> \endquote=\long macro:
->\endlist .
l.7 \show\endquote
so what do I do to make the correct margin?
add \leftmargin-1cm somewhere I suppose
1:32 AM
change that one arg to eg \leftmargin-1cm \rightmargin \leftmargin and both margins will decrease by 1cm
ah this is not going to work. I need to expand to the right on even pages and to the left on odd pages :-/
@yo' ah I thought it was for titlepage stuff where you know where you are. asymmetric changes over multiple pages gets you back to newgeometry:-)
@DavidCarlisle I do know.
I could as well probably just mess with the geometry parameters manually just before envoking \twocolumn[]\relax
@yo' well I don't care as it's your problem and my bed time:-) Night
@DavidCarlisle I go to bed, too, I just hate leaving these things unfininshed, it gives me nightmares
good night
2:17 AM
@yo' Lourdes in France. :)
@PauloCereda I got it then :) nice mountains
@DavidCarlisle Both are miracle-related. :)
@yo' <3
@egreg: England's rugby team in a nutshell. :) ^^
@PauloCereda you're mean :D
7 hours later…
9:34 AM
@yo' <3
@PauloCereda :-)
(btw, not spending much time here today, I've got some eye pain, and I think it's the "monitor eye")
@yo' Oh no.
@PauloCereda yes :-(
it sucks, tomorrow's a deadine for one conference (already submitted, so good), in 1 week's time is a deadline for another conference (will take some time)
so I hope it'll be better soon
@yo' You'll be feeling better. :)
@PauloCereda well now I'm a pirate, I should get an arara
9:38 AM
I have a doctor's appointment on Friday, my eyesight is getting worse and worse.
@yo' arrr
@PauloCereda :-( is it serious? do you wear glasses?
@yo @yo' @yo' and a bottle of run. <3
@PaulGessler Scuba ducks. :)
@yo' I do wear them. :) For 14 years. :)
@PauloCereda I wear them, but not for so long time
@yo' Welcome to the club. :)
@DavidCarlisle Did you see the c.t.t. post from 'GL'?
9:43 AM
@PauloCereda well, it's been a while (8-9 years I think)
@JosephWright no, not followed ctt for years, will look...
@JosephWright \plus ?
@DavidCarlisle He points out that xparse adding 'command X declared on line ...' isn't that helpful as the line refers to some package (usually) rather than to the user's code
@DavidCarlisle I was thinking of taking the point and removing the 'line context'
eeee macarenaaaaa
@DavidCarlisle GL is a 'regular' and is not keen on expl3 (or indeed LaTeX in many ways: he's the author of tabu)
@JosephWright must be a different message I was reading the top message from him about parsing plus in lengts
9:58 AM
@DavidCarlisle The most recent one about \noexpand? I'll find a link to the one I mean
@JosephWright oh got it a couple of messages earlier
@DavidCarlisle Yes, that would be it
@JosephWright I must say I don't really have much sympathy for the redefinition logging. when you load unicode-math and get told a thousand math characters have been redefined, what's the point? If left to me I'd turn them all off by default and have a \tracingall command that turned them back on.
@DavidCarlisle I'm also minded to do that, I have to say :-)
@DavidCarlisle You can guess probably who on the team likes this stuff :-)
@JosephWright Oh I know that:-)
10:35 AM
@DavidCarlisle this.
10:47 AM
@DavidCarlisle, @JosephWright: couldn't we forget all about this and have a nice cup of tea and biscuits? :)
@PauloCereda no go back to your vim site if you want to chat. This site is for serious tex^H^H^Hcricket discussion only.
@PauloCereda I started wearing glasses at 10. :(
@DavidCarlisle excuses excuses excuses. :)
@egreg Oh. :( I started with 16.
@DavidCarlisle: Tell @JosephWright to use this: whatthecommit.com
Reading a paper in Science: we have on the first page in an equation ln in italics!
@JosephWright Burn it! :)
10:59 AM
@PauloCereda Need to read it first, I'm afraid
@JosephWright Oh no! :)
@PauloCereda They've also suggested the said equation only applies to one set of units, which is also misleading
@PauloCereda A shame as the lead authors are (unsurprisingly) well-known and 'high power'
@JosephWright I'm Brazilian, I'm used to this situation. :(
@PauloCereda Well the journal office should have picked this up
@JosephWright Oh.
@DavidCarlisle: See @egreg speaking Spanish: tex.stackexchange.com/a/229733/3094 :)
11:27 AM
@PauloCereda same as Italian anyway isn't it?
@DavidCarlisle Don't you dare. :) I already suffer a lot when people think our capital is Buenos Aires and we speak Spanish. :) Oh and someone once mentioned our currency is Brazilian dollar. :)
@PauloCereda Minor differences between European languages are insignificant if you are a master of many languages and can speak 中國 and العربية as well as easy languages like italiano
@PauloCereda Of course! Brazil's capital city is Montevideo!
11:43 AM
@egreg I kill you. :) But I must say it's original. :)
@PauloCereda Asunción?
@egreg Lima. :)
@egreg: on a more serious note, any hints on how should I achieve this? tex.stackexchange.com/questions/229394/… My first approach, perhaps?
@PauloCereda Probably so.
@egreg Grazie. :)
1 hour later…
1:07 PM
@JosephWright pull request!
@DavidCarlisle just saw that in my inbox :-)
2:07 PM
Dirty question: do I need a visa for France?
@DavidCarlisle Yes I saw that: need to think about it
Dirty question two: do I need a visa for UK?
Dirty question three: can I take a picture with the queen? :)
@PauloCereda No, if you use your Italian passport. For taking pictures with Her Majesty The Queen you need a special permit from @DavidCarlisle who's friend with her.
@egreg oooooooh!
2:55 PM
@egreg she doesn't meet with vim users.
@DavidCarlisle I suspected it: it's well know she's a long time Emacs user.
@egreg Oh no!
3:49 PM
Hiho :-)
When you go to http://write-math.com/experimental/ you will notice a new "get PNG" button. What would you like more:

1. The PNG gets displayed in the browser (as it is currently done)
2. The PNG gets uploaded (anonymously) to imgur.com (stackexchange puts the images there, too)

This feature can be used by people who ask "how can I make this symbol with LaTeX" questions. They draw it, they save the PNG, they paste the PNG to the question.
(And similar to the automatic search which happens when you write the question, you might just find the symbol you looked for while drawing the symbol)
4:05 PM
@moose 1 seems enough to me, people can always get the png from the browser and do whatever they want with it (that would be a bit easier if you gave the png a zzzzz.png url as the file save default is a bit better then at least in firefox)
@DavidCarlisle Ok. I will try to get this. (By the way, I just removed MathJax. I have to work a little bit to make sure that STIXGeneral is included and to make sure that the user can give feedback about the suggested symbols, but I think this will replace the default version tomorrow)
4:22 PM
@DavidCarlisle Ok, I've got it. The file gets now automatically downloaded as 'symbol.png' (tested with Chrome 40) when the user clicks on 'get png'
4:37 PM
@moose thanks
4:49 PM
As a tourist I can enter and stay in France for 60 days using only a passport. A visa is required for more than two months.
@PauloCereda Brazilian passport?
@DavidCarlisle Yep. :)
Same for UK, apparently. :)
I'm not a dangerous duck. :)
@PauloCereda What happens when you came to France on 1st of July and are there at end of August? (is getting a visa for France / the UK complicated for you?)
@moose I just took a peek at the French ambassy website. :) I can easily go for a proper visa, but I just need a couple of days. :)
@moose If I need to stay for 60 days, I need to prove I have enough money for it. :)
Wow international laws are complicated.
Seems I forgot to make \bm a math command and @karlkoeller uses that fact:-)
Wow I would never have guessed the construct \bm{$h$} worked (and I wrote it:-) Actually even now I'm not quite sure how it worked. \boldmath as in my answer would be much safer and quicker:-) — David Carlisle 5 mins ago
5:04 PM
@PauloCereda Oh yes. A few months ago I went to the US to study. I had a scholarship, I only went there for 4 months but I still had to prove I have (1) about 1775 USD/month (2) I really want to get back. I needed to go to the embassy, fill out a lot of forms, pay a lot for the visit to the embassy and for SEVIS ... (if you're interested in details, I wrote it down here: martin-thoma.com/interact/#tocAnchor-1-18 to help other students who got the same scholarship)
@moose Thanks! :) This will be my first trip abroad, I'll go to Spain, but we will probably spend a day or two in France and/or Portugal. :)
@PauloCereda it's a conference?
@yo' Yep. :)
then don't forget to get business travel insurance.
Submission in TeX! Hell yeah!
5:10 PM
@PauloCereda Which conference?
@moose good question. :) Let me check.
5:29 PM
@DavidCarlisle A bug in your package?
6:19 PM
@karlkoeller undocumented feature?
6:33 PM
@PauloCereda -- we will require a report on your experiences in portugal, specifically, whether you can understand mainland portuguese, or, more importantly, whether the locals can understand you! (have a great time.)
@barbarabeeton :-)
It's funny to say "mainland Portuguese", it's like "mainland English", where the "mainland" is order of magnitude small than the "auxland"
@yo' Probably...
7:06 PM
@yo' -- true enough, but around where i live, there are several varieties of portuguese spoken quite extensively -- "mainland" (suppose one could refer to it as "ur-portuguese") and several varieties from the islands: cabo verde, madeirense, sao miguel, and other azorean dialects. (no brazilian that i'm aware of.) most of these are mutually intelligible, but there are idioms that "don't travel". (the food from the different areas is also distinctive, and all yummy!)
@barbarabeeton you gotta love the remark about food :)
@yo' -- i'm pretty much omnivorous. i won't starve (and in fact, will enjoy what i order) in lots of languages. all i require is that it be well prepared, made from naturally occurring ingredients, and (sadly) not too spicy-hot (results in "pizza effect", which is not fun for several days afterwards).
@barbarabeeton I'll have to google "pizza effect" I think. But that all means you would enjoy Czech food :-)
I probably have not done as much research on this as I should, but: is there any way to achieve this? vvv

I would like to make

  test \\
  test 2

Behave like

  test \\
  test 2

until the specified size is reached, after which it behaves like

  long test here \\
  test 2
I want the column to grow to a certain point, and only after that begin to wrap lines.
@PaulGessler that's quite a tough thing to do
7:19 PM
@yo' -- well, i just googled "pizza effect" and what is reported there is mostly different from what i know it as. locally (providence, ri), "pizza effect" means that the cheese on the top insulates the sauce underneath, so that when one bites into a fresh slice, the sauce erupts and burns/blisters the roof of one's mouth. too late ... and i don't know of any effective remedial measures. in the case of foods with high scoville unit content, it's a chemical burn; same unpleasant result.
@yo' yes, I can't begin to think how it might be done. It's way over my skill level, that's for sure. I wasn't sure if it's been done before and my searches missed something.
@barbarabeeton ah, the "lasagne effect" :D
@PaulGessler for a one-line table it's easy, but with anything more you'll need the longtable approach of writing to the aux file. It knows to be complicated.
@yo' ok, that's what I feared. Probably "too broad" for a question on main then, I think.
@PaulGessler I don't think so. But you've got only one @DavidCarlisle to answer it.
@yo' er what? answer what? I just come in and see my name:-)
7:32 PM
@yo' I thought too broad because it would basically require a new package, but I suppose there is precedent for that. :-) Maybe David has ideas.
@PaulGessler this @David
Sorry to bother you again @David.... is your "holiday" still ongoing this week? :-D
@PaulGessler tabulary can do something similar
@PaulGessler nah back at work
how do you assign a l3dim to a l2dim in proper L3 syntax?
@barbarabeeton Will do. :) I heard that language is something taken literally, so if you say I'd like a pizza, they will reply, So call again when you've made your mind. :)
7:37 PM

I would like to make


  test \\
  test 2

Behave like

  test \\
  test 2  test 2  test 2  test 2  test 2  test 2  test 2
  test 2  test 2  test 2  test 2  test 2  test 2  test 2
  test 2  test 2  test 2  test 2  test 2  test 2  test 2
  test 2  test 2  test 2  test 2  test 2  test 2  test 2
  test 2  test 2  test 2  test 2  test 2  test 2  test 2
@yo' Huh?
@yo' @PaulGessler ^^
@JosephWright \setlength \headwidth { \dim_use:N \g_ctuthesis_geometry_textwidth_dim }
@yo' You don't need \dim_use:N there as it's a context where TeX expects a <dimen>
@DavidCarlisle oh, you're right. I always forget about tabulary. I just need to have it set the target width on one column rather than the entire table. I'll take a look at the code when I get home tonight. Thanks for the recommendation.
7:39 PM
@yo' I'd use \dim_set_eq:NN \headwidth \g_ctuthesis_geometry_textwidth_dim myself
@JosephWright ok, but other than that, this is the prefered way?
@JosephWright ok. that's what I wasn't sure about
and should I keep \renewcommand \headrulewidth { 0pt } or better replace it by the L3 variant?
@yo' That one I'd keep: you don't gain anything replacing it
@JosephWright yep, seems to me so :-)
@JosephWright or even \headwidth \g_ctuthesis_geometry_textwidth_dim :-)
@DavidCarlisle that's what I hhad before, and then I thought: wait, that doesn't look right surrounded all these colons and underscores :D
@JosephWright so then also \dim_set_eq:NN \headwidth \textwidth I suppose
7:47 PM
@yo' Yes
@yo' As @DavidCarlisle obliquely points out, this is really about making the intention clear as TeX provides a 'short cut'
@SeanAllred nicely done with the XKCD reference. :-D
And then get the answer on this page somehow – either by directing them to this site or by quoting the necessary stops to a solution. xkcd.com/979Sean Allred 12 mins ago
@JosephWright after reading unformatted sourcecode of so many function as it's provided by \tracingmacros=1, this is nothing. It even stands on its own line ;)
@JosephWright includeinrelease seems stalled again
btw @JosephWright Is there any other way how to type a literal : in expl3 context than this?

\RequirePackage [ a4paper ,
	hmarginratio = \ctu@hmarginratio ,
	textwidth = \g_ctuthesis_geometry_textwidth_dim ,
	textheight = 0.75 \paperheight ,
	vmarginratio = \ctu@vmarginratio ,
	headheight = 16pt ,
] { geometry }
@DavidCarlisle Yes :-(
@yo' Afraid you'll need to use the \lowercase trick
7:52 PM
@JosephWright this works well.
@yo' in most cases you can just use : when do you need a catcode 12 :
@DavidCarlisle not here (I tried. Actually, I tried the hard way, getting unexplanable errors from geometry before I found the problem)
@yo' yes i saw the use and realised, it'll be doing a delimited argument, you could use \edef\zz{4\string:5}
@DavidCarlisle which would save me 2 Expl switches, right?
@DavidCarlisle with a space before : of course
@yo' no, none, if it worries you you can use \new_name_for_edef:N{4\new_name_for_string:N :5}
@yo' yes
7:57 PM
@DavidCarlisle that would be strange IMHO. I'll keep this small bit 2e so that it's clear it's a mistake of L3 :-D
@DavidCarlisle \tl_set:Nx \l_tmpa_tl { 4 \token_to_str:N : 5 } :-)
@JosephWright exactly what I said!
@PaulGessler ;)
@SeanAllred I so want to use xkcd.com/1179 as the response to a comment on a document I've put out for review at work, but I think I value my employment a bit too much to actually do it... :-)
@JosephWright I just learnt something about luatex hyphernation, did you know this?
@UlrikeFischer Interesting. It works if you make digits letters according to the hyphenation system (ie make them lowercase to themselves) luatex breaks 978-3-642-07617-6 if you add \lccode`0=`0 \lccode`1=`1 \lccode`2=`2 \lccode`3=`3 \lccode`4=`4 \lccode`5=`5 \lccode`6=`6 \lccode`7=`7 \lccode`8=`8 \lccode`9=`9David Carlisle 56 secs ago
8:09 PM
@PaulGessler HAHA! Yeah, I've been sneaking this format into everything
It's acting like a benevolent virus – people are actually adopting it
@SeanAllred not latex though!
@DavidCarlisle That actually might be changing
I seem to have set myself up in a similar situation as Jubobs (?)
@SeanAllred not in 2e.
@DavidCarlisle at least 2e has got the order correct, just the delimiters are non-compliant :-)
Spoke with my team lead and the topic switched to LaTeX, I mentioned that I'm not your average bear with it (I really don't want to toot my own horn, but the public's experience with LaTeX is usually less than ideal) and he said there might be an application within the company I'm working at now.
@DavidCarlisle Completely agree with the 2e thing though – LaTeX3 will make the whole thing so much easier to sell
in the figurative sense, of course
8:13 PM
@DavidCarlisle Sounds about right: hyphenation is for letters :-)
@JosephWright yes but pdf and xetex add a exhyphenpenalty in 123-456 but luatex doesnt
today there were two identical questions: latex \usepackage{graphicx} (on hold) and \usepackage{graphicx} (duplicate), both now closed. the second one was clearly an attempt by the author of the original question to add a mwe (he didn't know how to edit his original question, and was responding to a comment thereon).
some of the comments on the first question were good, but the statement of the second question was better, so i flagged it for moderator attention with the suggestion to close the first, point to the second. but before anything else could happen, the second was closed as duplicate, and the first shortly thereafter put on hold as unclear.
@barbarabeeton I initiated closing of the second as a dupe of the first because the MWE should have been added there instead of put in a new question. I didn't realize there were close votes going on in parallel on the other question; sorry about that.
there is actually a plausible answer in a comment to the first question -- that "The latest version of pdflatex will treat missing pictures as if you were using the draft option." i didn't know that -- if true -- but it's now impossible to add an answer to an on-hold question. if someone can verify the truth of the "draft" behavior, i will vote to reopen.
@PaulGessler -- i left a comment stating what i was doing; i think there was only one close vote when i did that, and there were few enough comments that it didn't just go into the "invisible extras" stack. but someone copied the mwe into the first question, so things are somewhat better there w.r.t possibly reopening. (but as i said, only if the suggested behavior with the new pdflatex really works.)
8:43 PM
@DavidCarlisle are you here ? I am no sur to understand your answer
								% Languages



\lipsum \cite{companion}

9:03 PM
@Maïeul oh hi I just went to the chat for that thread, but here works just as well.
@PauloCereda the awesome power of MS paint provides another on site answer.
@DavidCarlisle Not in the manual I can see: report to Hans/Taco?
yes, so I try to understand how this solution work
I am not sur
@JosephWright you're their favourite bug reporter
@DavidCarlisle It seems so
@JosephWright I could if you prefer:-)
9:05 PM
the dimens linked to an insert
@Maïeul yes
is the max dimen that allow for an insert in a page
@DavidCarlisle I guess I'll hack down to a plain (LuaTeX) demo and ask them
isn't it ?
@Maïeul not he 1000/factor it's the maximum height of \box\footins but of it's bigger than \vsize it's really a no-op as tex doesn't put more that vsize in an insert register anyway. plain tex sets it to 8in which is less than vsize:
$ tex
This is TeX, Version 3.14159265 (TeX Live 2014) (preloaded format=tex)
**\showthe \vsize
> 643.20255pt.
<*> \showthe \vsize


> 578.15999pt.
<*> \showthe\dimen\footins

9:10 PM
@DavidCarlisle <3
@Maïeul latex just doesn't do anything other than copy some code from plain (even has Knuth's original comment still) so it is left at 8in but \textheight in article is less than that so it makes no sense really.
hum, the problem in my case is that I have change to 0.25 \textheight
@Maïeul if you make sure that \dimen\footins is less than \textheight a split footnote will never take a complete page and there will always be at least one line of main text
and then the problem is still
the footnote continue
@Maïeul your MWE was single column so it is hard to be specific, but whatever column arrangement you have the idea is the same, you should make sure that a split footnote is at most a line less than a full column
9:14 PM
it was, in my real example
but as I have to make it minimal
@Maïeul minimal while still showing the problem is good:-)
it's very hard to explain
this topic help me
I was thinking the problem was both with \clearpage and \newpage
but not
@Maïeul eledpar seems to be in tl2015 so you could make an eledpar example and stick it here use lipsum package or some other text generator to keep it small;-)
@DavidCarlisle i stick in some year ago, you said "too complex code" ;-)
but I know now how eledpar work
so, to minute
pastebin.fr/38769 is perfect
@Maïeul well a simple example helps to understand the problem, but once that done, the examples can get more complicated:-)
@Maïeul Ok, will look:-)
9:19 PM
pastebin.fr/38770 dosn't work
so the problem is \newpage vs \clearpage
I don't know exactly what is the different
but I am not sur to need clearpage
newpage is maybe enough
the problem
on which I take so many time
@Maïeul oh yes, it certainly is;-) \clearpage is mainly designed to flush any pending floats, and latex does not support footnotes on floatpages, so \clearpage forces out the footnotes in a similar wayto it forcing figures and tables to be on float pages even if specified with [h] or [t] \clearpage basically means "dump everything you have now and start from a clean state in a new page"
is just because I dind't know the differnece between them
I am furious
very furious !
against me
@Maïeul so probably best you could do is make \clearpage be defined as if there are pending floats \originalclearpage \else \newpage That will give you the behaviour you want unless \clearpage is really dumping out figures
(and Peter Wilson, which is the original)
oh, in my case there is not float inside eledpar system
@DavidCarlisle Reported
9:29 PM
and I can change the call to \clearpage to a call to \newpage
@DavidCarlisle If I ever meet Hans or Taco it might not go so well!
@Maïeul In that case \def\clearpage\message{don't be silly: there are no floats to clear}} and your problem is gone
@JosephWright are you going to TUG2015:-)
@DavidCarlisle Still unsure, but I don't know if they go!
@JosephWright they used to when I used to. Not sure what that means for TUG2015.
@DavidCarlisle They are not on TUG2014 list: tug.org/tug2014/program.html
9:31 PM
on, no, just do pastebin.fr/pastebin.php?diff=38772 in the eledpar code
@DavidCarlisle I minimised as far as
\hsize=1.1cm %

\lccode`\2=`\2 %
\lccode`\3=`\3 %
\lccode`\4=`\4 %
\clearl@dleftpage and \clearl@drightpage should not be called outside of the internal eledpar command
and there is no reason to have page
but as you could read, I didn't understand clearpage vs newpage
and I dind't try this way
@JosephWright it arrived
@DavidCarlisle I always report on the list in case it's something I've missed in the manual
@DavidCarlisle I wonder if they need a good test suite ...
@JosephWright we could add that to ours....
9:36 PM
When working in the real world and reading/experiencing a Word file subject to a novice user, how do you avoid the urge to gouge out your eyes?
@JosephWright if we're going to seriously support Unicode engines we'll need a lot of compatibility tests like that:(
so, so many time for a so, so stupid solution
I am furious agains me !
@DavidCarlisle Yes, of course
so @DavidCarlisle clearpage clear float + insert
that is the problem in my case
@DavidCarlisle However, we are meant to test our code not other peoples, in the sense that if they've got bugs we are limited (see expl3 tests)
@DavidCarlisle We certainly are ending up doing tests for XeTeX/LuaTeX development more-or-less by accident!
9:38 PM
@Maïeul yes latex doesn't really do inserts footnote just crept in from plain tex and almost all the output routine is based around handling floats (which don't use \insert at all)
@JosephWright yes but at least if we have the tests we know they have bugs, even if we can't do much about it.
Are floats / footnotes ideas that are going to be married together in the future?
to creep ?
@SeanAllred how many conversations can I multiplex:-) No most likely (as in \insert isn't much help, really. To unify \footnotes should make them into floats
@Maïeul They 'snuck in' – historically
@DavidCarlisle > To unify \footnotes should make them into floats
That would be logical, yeah
Though I can't say I really know what \insert does
@DavidCarlisle Yes, true, and what we already are building up :-)
9:41 PM
@Maïeul In other words, footnotes were inherited from Plain TeX
@Maïeul sorry. As Sean said. It looks from the code like they were not at the top of Leslie Lamports concerns:-) the code for footnotes is just a copy of the plain tex code.
@DavidCarlisle Issue is that LuaTeX in particular is still pretty buggy
@DavidCarlisle you mean of \footnote{}? I means of personnal note created by \newinsert
@SeanAllred but you need split footnotes, so that would mean you need split floats and that would mean rewriting most of latex, which was why there was a decision to have latex3 a couple of decades ago;-)
@DavidCarlisle And we've come full circle once again! ;)
9:42 PM
@DavidCarlisle You can guess what I'm going to say here
@Maïeul Yes the definition of \footnote in latex is very simple and just copied from plain
@JosephWright I refer you to the starred comment on the right
@DavidCarlisle :-)
ok, but in the case of eledpar
@DavidCarlisle It's like PMQs
it \newinsert\Afootins
etc etc
9:44 PM
@JosephWright I wish I was getting cash for questions
and \clearpage also \clearthem
@DavidCarlisle :-)
@DavidCarlisle Probably cash for answers but ....
@DavidCarlisle No-one from outside the UK is going to understand this at all!
@Maïeul There are very few latex packages that use \newinsert and latex's output routine has essentially no support for inserts other than the primitive behaviour that is built in.
@JosephWright @DavidCarlisle Yeah I'm completely lost
but I'm okay with being lost :)
@SeanAllred PMQ= weekly argument in parliament "Prime Ministers Questions"
9:46 PM
@DavidCarlisle i.e. getting paid to ask questions all day – gotcha
@SeanAllred At PMQs it's common to ask a 'standard' opening question (the rules make this advantageous), so the PM has a set of 'ready to go' answers of the form 'No. 3'
@DavidCarlisle indeed, and, because I am a philologue which touch a little bit in LaTeX, am the only maintener of one these package... for my great misfortune
@SeanAllred No there is a current potential scandal of MPs taking money to ask questions on things
@DavidCarlisle There was the Al-Fayed one as well: brown envelopes etc.
@DavidCarlisle oh my
That kind of getting paid
9:49 PM
@JosephWright I'm sure no other country's politicians ever have corruption allegations.
I haven't been keeping on top of the news since the new job
@DavidCarlisle Nah – we're infallible and incorrigible.
@SeanAllred a small tip, if you are going to be fallible don't do it with someone who is really a TV reporter with a hidden camera.
@JosephWright -- i think these days they mostly concentrate on the context meetings. and sometimes to bachotex (both taco and hans gave presentations at bachotex 2014).
@barbarabeeton Yes
10:25 PM
@DavidCarlisle thanks in any case. You'r conversation help me to find where was the problem. Very strange. 2 Years on this problem, and the solution is so easy...
This might be basic and easily searchable in the book, but in case someone can give a good and fast answer: how does \iffalse work?
Does it go, without expansion, until the next \fi but watching and matching \iftrue and \iffalses with \fis along the way?
It it's not that, but easy to explain, I would like an explanation :P
@Maïeul ;-)
@Manuel more or less it passes over tokens without expansion but matching if/else/fi until it gets to the matching \else or \fi
Okey, So \let\foo\iffalse would be “taken in account” while \def\foo{\iffalse} would be “hidden”, isn't it? \iffalse\foo\fi
@Manuel not just \iftrue \iffalse but als \ifx \if \ifnum ec
10:35 PM
@Manuel \foo would be hidden \def\foo{\iffalse} would be seen as an \if (and is a common source of error if people put such things in a conditional region)
@DavidCarlisle Yes, I meant \foo not the definition itself. Thanks for the fast assistance :)
10:54 PM
@DavidCarlisle that's why you have \footrue and \foofalse in the first place :D
@DavidCarlisle I am something very strange. If I use \insert directly I have a problem
but if I do a \vbox and insert the \unvbox ... that wroks
I don't understamd
very very strange
ok, I understand why
the vbox lost some floatingpenalty
@yo' exactly
11:10 PM
so, so stupid. Many thank to you @DavidCarlisle. YOu have forced me to try to make a clear MWE, and so the problem appears when making it
11:57 PM
@egreg: I saw you just answered something on Math.SE...
...now please answer my question:
Q: How to calculate the median age of a population?

WernerI want to calculate the median age of a population. Consider the following example data (downloaded from Statistics Canada's CANSIM database, Table 051-0001): Table 051-0001 Estimates of population, by age group and sex for July 1, Canada, provinces and territories, annual (persons unless otherw...

@Werner Is this math? ;-)
@egreg Fundamentally, yes.

« first day (1577 days earlier)      last day (2197 days later) »