« first day (2685 days earlier)      last day (2250 days later) » 

12:02 AM
@DavidCarlisle thanks David! I also realized that is 1 am here, that's why I was sleepy! Good night
 
@DavidCarlisle Amazzzzing. Not a single \zzz command. ;-)
 
@marmot tikz has too many z already
 
 
8 hours later…
7:57 AM
@egreg @DavidCarlisle @WillRobertson polyglossia does advertize the language tag. But it does it with \addfontfeature and so the question is if features set with \addfontfeature should survive a \normalfont and if not how language tags should be better advertized.
\documentclass[11pt]{memoir}

\usepackage{fontspec,xcolor}
\setmainfont{Linux Libertine O}

\begin{document}
In Turkish,  `\textsc{ı}' and `\textsc{i}' are different
letters.


\addfontfeature{Language=Turkish,Color=red}\selectfont
In Turkish,  `\textsc{ı}' and `\textsc{i}' are different
letters.

\addfontfeature{Language=Turkish,Color=red}\normalfont
In Turkish,  `\textsc{ı}' and `\textsc{i}' are different
letters.

\defaultfontfeatures{Language=Turkish,Color=red}
\setmainfont{Linux Libertine O}
 
8:41 AM
@TorbjørnT. In English keyboard how can I type your name? Then is it Torbjorn T.?
I saw you answered many of my question. But, when I comment to you I face a problem typing your name in my keyboard. So, sometimes I copy and paste your name, But I dislike mouse. So, if you permit me to your synonyms in English it is easy to me.
 
@alhelal Yes, that's fine. (But if it's to ping me in a comment or chat use tab completion, o instead of ø might not work then.)
 
@DavidCarlisle A TikZ answer?!
 
@PauloCereda It will be l3draw soon enough :)
 
@JosephWright ah. :)
 
@alhelal type @T and then you should be able to just select the user (or use the tab key) to finish the name
 
8:56 AM
@UlrikeFischer In my opinion the language tag for the main language should be added to the “family 0”, that is, \normalfont.
 
@PauloCereda yes if presented with a three line example in which lines 1 and 2 work and line 3 doesn't, my in-depth tikz knowledge lead me to the solution of copying code from line 2 to line 3. I may be able to transfer some of that knowledge to l3draw with @JosephWright's help.
 
@egreg But if one switch to another language it should be removed and the question is how this should be implemented.
 
@DavidCarlisle Which question are we talking about? Need to make sure I've implemented all of the requirements!
 
@JosephWright question in chat last night (scroll up a dozen lines)
 
@JosephWright it's a TikZ thingy not solved with TikZ thingies because the first TikZ thingy problem could be solved with a second TikZ thingy, but then the third TikZ thingy could make the other two TikZ thingy fail, so @DavidCarlisle wrote more TikZ thingies to avoid TikZ error thingies. :)
5
 
9:01 AM
@DavidCarlisle Ah, might need to do a bit more work to cover that!
 
@TorbjørnT. In comments all time the Tab completion not works.
 
@alhelal Then notifications wont work either (unless I've edited the post). If you're commenting on an answer I've written, I'm always notified, so you don't need to write @username.
 
@TorbjørnT. Thank you.
 
yo'
9:22 AM
Ok, so the referee of my paper called me TOMAS HAJDA. However, he gave a nice review, so let's forgive him :-)
 
@yo' should be yo' I assume:-)
 
@yo' At least it's better than Patrick's famous Mr. Gundla story. :)
Hello Marek,

>  Hi Mr. Gundla
>  Just want to say Im very dissapointed that no installing package of ConText exist.
>  I mean like you install things in Win – two cliks and Done!
>  I was studying ConText whole afternoon and i liked it – But – im very very sick of this kind of installing.
>  It just makes people like me to hate the whole thing about trying something else then MS Word.
>  I was eager to learn this  ConText ant than BIIIIIIG! Frustration cause i could not run it.
>  All those versions of packages - i wanted to shoot myself.
 
@PauloCereda oh wow I never realised that the context installation instructions were an exact copy of the latex ones.
 
@DavidCarlisle :)
 
yo'
@DavidCarlisle sure!
@PauloCereda oh
 
9:35 AM
@yo' a reviewer once corrected my name. :)
 
yo'
However, this minor mistake makes me guess the country of origin of the reviewer :-)
 
:D
@yo' ooh Mr. Hadja
 
yo'
@PauloCereda well, Tom HEDJA (literally) was written on my apartment rent contract once :-)
 
@yo' Isn't the French that capitalise the surnames?
 
yo'
@PauloCereda well, these were, moreover the Scots expats in France :-)
 
9:39 AM
@yo' ooh
 
yo'
But in spite of both my first name and surname spelled wrong, the contract was sufficient to get rent support from the government agency :-)
 
@yo' ooh :)
 
9:53 AM
@DavidCarlisle would you write in an offer "a monitor -- optionally" or "a monitor -- optional"? And is both valid or is one really wrong?
 
@UlrikeFischer If I understand the context then "a monitor -- optional" (or more likely to write "and optionally a monitor")
 
@DavidCarlisle It is in a title (table head) not a sentence.
 
yo'
@UlrikeFischer "a monitor (optional)"
 
@yo' could work too. I will ask them (that's more a layout than a language question).
 
@UlrikeFischer If I understood this correctly, zref won't work with hyperref. Did I get this right?
 
10:00 AM
@UlrikeFischer ok then optional not optionally and either a dash or parenthesis as @yo' suggested
 
@JohnDorian No. I have built very complicated \longref commands which works fine with hyperref. And as zref and hyperref are from the same author it would be quite bewildering if they were incompatible.
 
@UlrikeFischer that's right because tex authors never make packages that are incompatible with their own packages.
 
@PauloCereda I'm struggling to get the fourier/utopia font combination to work in ConTeXt, and I just say I feel with Marek at the moment :)
 
@mickep :D
 
@DavidCarlisle well some of them build complicated depency systems to keep them compatible ;-).
@mickep I hadn't the time to try (and try hard not to do -- i will never finish my project when I start on this) but are you aware that you didn't wrote in your question what doesn't work?
 
10:14 AM
@UlrikeFischer Oh, I thought that was clear from the output in the pdf file. But maybe I should have been more clear. I'm right now trying to understand how the LaTeX fourier package works, and then maybe it is possible to see what I did wrong. But you have a point, I should write an email to update what I was hoping for... I hate to bother Hans with these issues, though...
 
@JosephWright Interesting feature: If say ifpdf.sty is not available I (in memoir) throw an error, so they user now know that this is not being emulated anymore. If found I run \RequirePackage. But here is the problem with this: this does not tricker the miktex package auto install! Is there a \RequirePackage with custom error msg upon error?
 
@daleif Please I am finishing the thesis, keep memoir as it is for a couple of days. <3
 
@PauloCereda s/finishing/starting/ surely
 
@DavidCarlisle oi
@DavidCarlisle finishing = starting iff |text| = 0. :)
Wait a minute...
 
@daleif you could do IfFileExists{ifpdf.sty}{Requirepackage...}{custom error} but then you are executing far more code in the guards than if you'd just copied in the half a dozen lines of the package and defined \ifpdf directly.
 
10:24 AM
@DavidCarlisle This is what I'm doing right now. I'm removing the suboptimal/bad emulations memoir has of the ifpdf, ifetex, ifxetex and ifluatex packages and force users to use the provided external packages. Just wanted to let them know why if they update only memoir, that their doc might no longer compile (you have users with very minimalist setups)
 
@mickep well to see the output one must first repeat your installation. That's a lot work. And imho it is okay to write such questions to a mailing list -- Hans doesn't need to answer if he doesn't want to.
@daleif You could run the tests with the custom error and then load the package outside/after the test.
@daleif But I doubt that many users really care for a custom error here.
 
@UlrikeFischer I agree that it is too much to ask anyone else to install the font to help. This is why I posted the contents of the three files. But, yes I agree with you, it is perhaps not so easy (for someone else than Hans) to see directly what could be wrong. In any case, I'm at home today, and will spend some more time to try to understand...
 
@mickep to help one probably will have to install them -- if only to test solutions -- but it shouldn't be necessary only to understand what the question is about.
 
@UlrikeFischer You have a point, sometimes one just tries to be too nice.
 
10:46 AM
Now just replaced with \RequirePackage
 
11:02 AM
@DavidCarlisle is the following still a proper test for whether luatex85 is needed
\ifx\pdfextension\@undefined\else
\RequirePackage{luatex85}
\fi
I don't know if ifluatex has tests for the version build in
 
yo'
11:30 AM
@DavidCarlisle @JosephWright Can I have an English question, please? Which is correct?
..., which follows from the Strong Approximation Theorem~\cite{foobar}.
..., which follows from Strong Approximation Theorem~\cite{foobar}.
 
@daleif well I suppose so but loading luatex85 by default is not the intended usage it is supposed to be used as a stopgap measure for document authors to use if they want to use packages that were not updated.
@yo' the
 
yo'
@DavidCarlisle ok I thought so just wasn't sure, thanks a lot!
 
@daleif it would be a good thing if the default usage of memoir in lualatex did not load the pdftex compatibility by default, packages that have engine dependencies really need to have a luatex-specific code path rather than just getting away with luatex being close enough to pdftex
 
@yo' Where's @JosephWright with those lovely chemistry paper titles that no one can ever read them? :)
 
@yo' perfectly readable as beta
 
yo'
@DavidCarlisle sure
 
@yo' people who don't know their unicode codes shouldn't be reading the paper.
 
@yo' Actually a bit funny that the title is about some kind of substitution. :)
 
yo'
11:52 AM
@mickep yeah, a beta-substitution :-)
 
@yo' you should have stuck to ß :-)
 
yo'
@DavidCarlisle is that the German thingy?
 
Mar 3 at 15:37, by David Carlisle
@egreg or if we get bored of annoying the Italians by discussing pineapple pizza, we can call it beta and annoy the Germans
 
12:05 PM
@DavidCarlisle how do you suggest we do that? I'm note quite sure what you mean by "not load the pdftex compatibility by default". There if no equivalent to \ifpdf in \ifluatex, so users who maybe want to use several engines on their doc should do what exactly?
 
yo'
Why I so enjoy the usage of words such as straightforwardly?!
 
12:36 PM
@daleif ifpdf is OK, I mean you shouldn't load luatex85 in the class
 
@DavidCarlisle I think we added it so that we still could get the document size written directly to the PDF. Though I have not checked what else registers memoir uses that need luatex85 (other than \pfdpageheight/width), there was also the problem of users having old luatexs
 
@daleif take the example of graphicx if that hadn't been updated then it wouldn't have worked after luatex 85, but a user could load ltluatex85 and then luatex woudl emulate pdftex enough to make graphicx default to the pdftex back end and that to more or less work, however it only just worked and is not a stable way to build the system going forwards. Now graphics has a luatex.def back end just as it always had a pdftex.def back end, and so any luatex specific stuff (notably lua code) can .
.. go in theer without being tied to always working via a pdftex emulation
 
BTW: using the geometry package, can anyone make this work under lualatex (luatex85) does not seem to do anything
\documentclass{article}
\usepackage[paperwidth=10cm,paperheight=10cm]{geometry}
\begin{document}
test
\end{document}
 
@daleif exactly so \ifluatex then set \pageheight and \pagewidth not \pdfpageheight and \pdfpagewidth no need to emulate the pdftex way
@daleif I'll check, it used to work, I have asked roughly every 6 months that geometry be updated not to requure luatex85...
@daleif this works for me with luatex
\RequirePackage{luatex85}
\documentclass{article}
\usepackage[paperwidth=10cm,paperheight=10cm]{geometry}
\begin{document}
test
\end{document}
 
@DavidCarlisle I don't even think memoir has a luatex branch in the layout fixing code, will fix that. Does not seem that we use that many \pdf... macros, there is a little in memoir.cls and a \pdfstringdef in memhfix.sty
 
12:45 PM
Creator:        TeX
Producer:       LuaTeX-1.07.0
CreationDate:   Thu Mar  8 12:44:18 2018 GMT
ModDate:        Thu Mar  8 12:44:18 2018 GMT
Tagged:         no
UserProperties: no
Suspects:       no
Form:           none
JavaScript:     no
Pages:          1
Encrypted:      no
Page size:      283.465 x 283.465 pts
Page rot:       0
File size:      2763 bytes
Optimized:      no
PDF version:    1.5
 
@DavidCarlisle I'm using TL17 frozen, that gives me an A4 sized doc
Creator: TeX
Producer: LuaTeX-1.0.4
CreationDate: Thu Mar 8 13:46:10 2018
ModDate: Thu Mar 8 13:46:10 2018
Tagged: no
UserProperties: no
Suspects: no
Form: none
JavaScript: no
Pages: 1
Encrypted: no
Page size: 595.276 x 841.89 pts (A4)
Page rot: 0
File size: 2770 bytes
Optimized: no
PDF version: 1.5
 
@daleif hang on let me go back a year
@daleif oops i can't my tl2017 has luatex updated with a dev version I compiled from the luatex sources:-) but luatex 1.04 from texlive tl2016 gives:
Creator:        TeX
Producer:       LuaTeX-1.0.4
CreationDate:   Thu Mar  8 12:51:45 2018 GMT
ModDate:        Thu Mar  8 12:51:45 2018 GMT
Tagged:         no
UserProperties: no
Suspects:       no
Form:           none
JavaScript:     no
Pages:          1
Encrypted:      no
Page size:      283.465 x 283.465 pts
Page rot:       0
File size:      2758 bytes
Optimized:      no
PDF version:    1.5
@daleif with RequirePackage as first line? that's very odd
 
@DavidCarlisle your exact MWE.
Nevermind, now it stopped doing that. Strange ....
 
@daleif perhaps I fixed it at a distance:-)
@daleif see geometry is another example while it should be fixed, the fix should't be for it to load luatex85, it just needs to set \pagewidth` instead of \pdfpagewidth. If it (or a geometry package user) load luatex85 just to define those two registers then the trouble is that emulates all the pdftex image inclusion and other stuff and then every other package has to take care to detect luatex before testing for pdftex so that it isn't tricked into thinking it is on pdftex ..
.. and using a sub-obtimal pdftex backend rather than a luatex one.
 
1:05 PM
@DavidCarlisle I blame cosmic rays, as usual.
@DavidCarlisle Well, I tend to favor the pdf backend over the lua one, and only use the lua one if I really need to (it's just so slow...)
 
@daleif ooh ray charles
 
@daleif sure but the slowness is luatex rather than pdftex, not between luatex and luatex with some pdftex emulating macro definitions
 
@DavidCarlisle sure. I just find the slowness annoying enough to not use it in normal projects
 
1:21 PM
@daleif I don't use tex at all but still need to make lualatex work:-)
 
@DavidCarlisle someone should write l3geometry ;-).
 
@UlrikeFischer ooh
Soon we will have L3TeX instead of LaTeX3. :D
 
1:38 PM
@UlrikeFischer well....
 
user image
2
@DavidCarlisle you mean something like this? ^^^
 
Anyone know a context (under pdflatex) where \pdf(h|v)origin would be zero? memoir sets these to 1in if tested to be 0pt. Not quite sure why (not explained in the dtx)
 
 
1 hour later…
2:58 PM
$ grep -l horigin /usr/local/texlive/2018/texmf-dist/tex/latex/*/*.*
/usr/local/texlive/2018/texmf-dist/tex/latex/bxjscls/bxjsarticle.cls
/usr/local/texlive/2018/texmf-dist/tex/latex/bxjscls/bxjsbook.cls
/usr/local/texlive/2018/texmf-dist/tex/latex/bxjscls/bxjscompat.sty
/usr/local/texlive/2018/texmf-dist/tex/latex/bxjscls/bxjsreport.cls
/usr/local/texlive/2018/texmf-dist/tex/latex/bxjscls/bxjsslide.cls
/usr/local/texlive/2018/texmf-dist/tex/latex/bxpapersize/bxpapersize.sty
/usr/local/texlive/2018/texmf-dist/tex/latex/crop/crop.sty
@daleif ^^
@daleif it simplifies the calculation if you know it is set to a fixed amount, but of course if a user has set it setting it to 1in shifts everything by an inch
@daleif it's not unreasonable to set it to 0 (that's why it was added) to avoid the weird tex default origin but if all the existing page size code is assuming that 1in default then forcing it to be 1in also works. Just setting it to 1in if it was 0 is a bit odd
 
 
2 hours later…
4:47 PM
user image
6
\usepackage[pigeon]{stackengine}
4
 
@PauloCereda Are the shadows made by TikZ?
 
5:47 PM
@marmot ooh advanced pigeon shadow
 
@PauloCereda But why are we calling seagulls pigeons?
 
@PauloCereda I have suggested something like that ...
 
6:10 PM
 
@AlanMunn ooh sea pigeons
 
@marmot <3<3<3
 
6:52 PM
@PauloCereda from the society for putting things on top of other things.
 
@UlrikeFischer ooh
@ChristianHupfer ^^
 
 
3 hours later…
9:24 PM
@UlrikeFischer I will try to provide a small example code within the next days (I am away over the weekend).
@UlrikeFischer Thanks for taking an interest in the question!
 
9:57 PM
I have a new palindrome:
user image
3
 
@Kurt Your previous one was better, this one is not a prime number.
 
@marmot LOL
 
10:15 PM
Yes, \addfontfeature dates back to v0.2 or something of fontspec and it's probably not the best interface.

Almost certainly I need a new command like \addfeaturetofamily\rmfamily{...} which does do a redefinition.
 
Yes, because you are a genius. All of you are genius in fact: ChristianHupfer, Mico, Ulrike Fischer, samcarter, barbara beeton... and many others I'm forgetting or simply I don't know them. — Dog_69 10 mins ago
 
@UlrikeFischer I am flat out right now so any assistance with syntax/semantics and adding an issue to the Github tracker would be welcomed :)
 
@egreg It is certainly better to be called a genius by a dog than being called a dog by genius. ;-)
2
 
@marmot :-D
@marmot But everybody can notice a missing name from the list… ;-)
 
@egreg Oh, you are referring to our super-genius, right? ;-)
 
10:35 PM
@marmot Waiting for a witty comment from him.
 
10:45 PM
@egreg he will say dogs are not as reliable as ducks and marmots.
 
@PauloCereda Talking dogs are very smart!
 
@egreg 🐧
 
@PauloCereda 🐶
 
@egreg I disagree with this statement... ;-)
 

« first day (2685 days earlier)      last day (2250 days later) »