« first day (2917 days earlier)      last day (2018 days later) » 

12:33 AM
@egreg succinctly getting to the root of the problem. (Pun intended)
Well, if I try to go past the errors, I get a six meter wide output. No wonder you get overflows. — egreg 3 hours ago
 
 
7 hours later…
7:08 AM
@UlrikeFischer I'll try and look later...
 
@DavidCarlisle fine. I didn't push anything yet. I'm on the flight back today so can't do something now, but in the evening it should. But I really don't think that there is a real problem. The most curious diff was a luatex failure on a xetex test and one probably don't need to care about this ...
 
 
1 hour later…
8:40 AM
Another day, another new Bruno-based expl3 module
 
@JosephWright tic toc tic toc?
 
@DavidCarlisle ;-). I'm now off. I just tried to look a bit at the module declarations of luaotfload and it is a mess ;-(. Every file seems to use another variant.
 
@barbarabeeton interesting, now I know at least one Linux user using Acrobat. Thank you for telling me. I wonder though why you don't use xpdf or something like that for the Linux machines.
 
@JosephWright Just saw your mail. I'm using l3str-convert in tagpdf. Seems to work fine ...
 
8:53 AM
@JosephWright what did the lord of expansion came up with?
 
@UlrikeFischer we simplified (removed most of) the module declaration stuff in luaotfload when we adopted it so older declarations work but don't really do anything
 
9:14 AM
@UlrikeFischer Sure ... was one of my 'targets'
@Skillmon This one isn't expansion-based, but it is clever for benchmarking
 
9:41 AM
Quack!
 
9:54 AM
Apr 15 at 9:38, by David Carlisle
@PauloCereda dinner!
 
@DavidCarlisle oh no
@DavidCarlisle you are mean
 
Jun 29 '17 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)
 
@DavidCarlisle oh
 
@PauloCereda we should publish an RFC for a communications protocol based on this handshake
 
@DavidCarlisle ooh I like this idea
 
10:17 AM
@UlrikeFischer I do not think that there is any value in these module statements so we should replace them with comments like in luaotfload-filelist.lua. We just have to add a line for License information.
 
10:34 AM
@MarcelKrüger @UlrikeFischer current documentation for provides_module says
% \DescribeMacro{provides_module}
% |luatexbase.provides_module(|\meta{info}|)|\\
% This function is used by modules to identify themselves; the |info| should be
% a table containing information about the module. The required field
% |name| must contain the name of the module. It is recommended to provide a
% field |date| in the usual \LaTeX{} format |yyyy/mm/dd|. Optional fields
% |version| (a string) and |description| may be used if present. This
% information will be recorded in the log. Other fields are ignored.
 
@MarcelKrüger @JosephWright I think it would be useful to have some policy like \ProvidesPackage/File for lua files so that not everyone invents his own system.
 
@JosephWright we should probably recommend yyyy-mm-dd format now, but in luotfload it may make sense to drop all the pther fields if you keep this
 
@MarcelKrüger @JosephWright I think it would be useful to have some policy like \ProvidesPackage/File for lua files so that not everyone invents his own system.
 
@UlrikeFischer that's more or less what luatexbase.provides_module is supposed to be
@UlrikeFischer I don't think the indirection via luaotfload.module helps, you can simply do provides_module{name=.., version=...}
 
@DavidCarlisle imho luaotfload uses it in one place but without date/version. And how do you use for sub files?
 
10:40 AM
@UlrikeFischer it doesn't do anything other than put stuff in the log, so if you want to echo in the log I think you can declare (anything) as a module, or if you just think of it as part of the main file, just use a comment i guess.
@UlrikeFischer but that module code isn't extensively used/tested in that form. It would be good if luaotfload used the "standard interface" as described in ltluatex.dtx but if you find it would be convenient to have more features that could be arranged, rather than extending it in luaotfload
 
@DavidCarlisle I simply want the version and date in the files in a systematic way.
@DavidCarlisle and I'm quite fine with any "official" suggestion - it would make easier tagging If one reuse functions across packages.
 
11:07 AM
@DavidCarlisle, @UlrikeFischer One for the team list? A short example?
@UlrikeFischer We should probably look at what ConTeXt does
 
@JosephWright later, just boarding ...
 
@UlrikeFischer Safe flight
@DavidCarlisle Yup
@UlrikeFischer As @DavidCarlisle says, I think we do have a policy in ltluatex, it's perhaps just not well-known. I'll make sure the L3 files doe it, and perhaps put in a PR for luaotfload
@UlrikeFischer BTW, I was checking on GitHub: if you do decide to go that way, it is possible to transfer repos from one user to another, so your one could become a team one ...
 
11:35 AM
@JosephWright ooh
 
12:30 PM
@Skillmon Just to be a bit contrary: If a rectangle is a polygon all of whose sides are right angles, as the name suggests, a rectangle in hyperbolic geometry must have at least five sides.
 
I have to use begin{align} in my latex file, but it says: '! Package amsmath Error: \begin{align} allowed only in paragraph mode.' How to overcome this?
hmm! found solution!
 
@Silent That seems odd: I can't find that message in amsmath.sty.
 
@HaraldHanche-Olsen but aligned thingy really worked :)
@HaraldHanche-Olsen i am new to latex, can you recommend some online material to get some grasp over it?
 
@Silent So your problem was that you used align in math mode? Then I understand why it went wrong, but not why you got that particular error message.
 
oh
 
12:40 PM
@Silent Perhaps someone else can; I have been using it far too long to be familiar with the current offerings on that front. There is the LaTeX wiki book, but from what I hear, its quality has declined. I don't know if that has improved or not, as I hardly ever look at it myself. Again, perhaps someone else can comment.
 
thanks for that reply:)
 
@Silent If I try to put an align inside an equation*, I get the error message below. But perhaps the error message has been clarified in a later version, and you are running an older version.
! Package amsmath Error: Erroneous nesting of equation structures;
(amsmath)                trying to recover with `aligned'.
 
I am using MikTex
don't really know what it is :)
 
@Silent I get the following in my log file. Not being a windows user myself, I have no idea how to check which miktex version you have nor how to upgrade it. But you could compare what you find in your logfile and see if your amsmath is older. (If you care.)
Package: amsmath 2017/09/02 v2.17a AMS math features
 
12:56 PM
@Silent there are thousands of tutorials on the internet for amsmath (some of them are even correct) but I'd start with the documentation that comes with the package (amsldoc.pdf) which you will have on your system but is texdoc.net/texmf-dist/doc/latex/amsmath/amsldoc.pdf
 
1:22 PM
@JosephWright just managed it in time for the next boarding ;-)
 
 
2 hours later…
cis
2:58 PM
Hello my friends! What is the meaning of the quotation mark ( " ) when I use
`\pgfplotstablename \pgfplotsretval`
and get `"TableName`? (https://tex.stackexchange.com/questions/456576/pgfplotstable-how-to-use-pgfplotstablename)
 
3:10 PM
@cis It is the backslash; if you add \usepackage[T1]{fontenc} you'll see it.
 
@UlrikeFischer @JosephWright @egreg @DavidCarlisle Dear lualatex experts, sorry for bugging you in the chat, but do you by chance have an idea what's going on here? (I am sorry for pinging you like this, but I think that there is a small chance that this is a problem of lualatex and not pgfplots.)
 
@marmot Probably a bug in the LuaTeX driver for pgf
 
@JosephWright And do you happen to know how to resolve it? (Looks a bit like a < vs. <= problem.)
@CarLaTeX When you entered this room with your fast car, I was afraid for one sec that you'd hit me, but then you made a small correction and all is fine. You're a really good driver! ;-)
2
 
cis
@egreg Ahaaaa.... again learned something.
 
@JosephWright @marmot Agreed
 
3:19 PM
@egreg @JosephWright Maybe you can ask @CarLaTeX for help, she's such a careful driver, much more careful than the LuaTeX driver. ;-)
3
 
3:38 PM
@marmot She might need another car to drive luatex, the red one has LaTeX written on its side.
2
 
@samcarter Good point! ;-)
 
@marmot I wonder if she will continue to drive with the convertible in winter - this must be getting cold!
 
@Skillmon -- corporate policy, i guess. that decision is 'way beyond my pay grade. but it's also true that it appears to be expected that most subscribers to ams journals are using acrobat, and who knows what platforms they're operating on. similar decisions apply to choice of browsers being tested as well.
 
4:00 PM
@barbarabeeton anything you do internally of course is reasonable but most linux users would have great difficulty getting hold of acrobat now I think. It's unsupported and would get flagged as a security vulnerability if you managed to find an old copy and tried to install it. So expecting any non windows users to have acrobat isn't an expectation that can be met.
 
@samcarter You seem never to have lived in Southern California. ;-)
 
@marmot lol
 
@DavidCarlisle -- what you say is absolutely true. what i was responding to, though, was @Skillmon's statement "I know not one Linux User who uses Adobe." it's true that the breed is nearly extinct. never said, nor meant to imply, anything different.
 
@barbarabeeton yes well I wasn't really complaining to you just to the ams in general at such a policy (not that I expect you to pass it on:-)
 
@DavidCarlisle -- even if i did try to pass it on, nobody would pay attention.
 
4:18 PM
@barbarabeeton no doubt true:-)
 
Back in germany - and they lost our luggage in munich again ;-(
 
@marmot -- i've never lived there, but i've visited often enough to know that it's not my natural habitat. (one of my nephews does live there, and two more live in the bay area. they all miss seasons. my father lived/worked in china lake for years, and moved to santa monica after he retired. it was lovely to have fresh avocados on the tree in the next-door neighbor's yard in january, but that's not enough for me to give up new england.)
 
@barbarabeeton or even better than new england, would be the original england?
 
@barbarabeeton I've lived for 2 years in Hamburg, and I have accumulated more than enough dark and gray days for one lifetime... ;-)
 
@barbarabeeton in England, we even have capital letters and lower case 'u's.
 
4:30 PM
@DavidCarlisle -- but not the riotous reds of the autumn trees. (which we actually probably won't have this year in providence, owing to all the rotten weather earlier in the year. in fact, many trees are dying for lack of proper conditions.) and it's probably still easier to go to quebec than it will be soon for you to cross the channel. what a mess!
@marmot -- à chacun son goût. as dorothy said, after returning from oz, "there's no place like home."
 
4:45 PM
@marmot well the problem seems to be with the combination of smooth, soft clip and that the begin of the soft clip domain is the first value. If I use soft clip = {domain=0.3001:0.6} the output is fine again. So I would suspect a rounding problem or as you said < instead of <= somewhere.
 
4:56 PM
@marmot just saw that you already wrote this in your answer;-)
 
5:33 PM
@CarLaTeX lul
 
Now GitHub is officially Microsoft-owned.
 
5:51 PM
@barbarabeeton Does that mean you have no burrow? ;-)
 
@marmot -- oh, gordon and i have (what i think is) a very nice burrow, high above maximum high tide (even with rising sea level), just down the street from the local library, and within easy walking distance of a japanese "pub food" restaurant. (for weekend brunch, one of their options is okanomiyaki. yum!)
 
@barbarabeeton Sounds like a home to me ;-)
 
@DavidCarlisle oh
 
@marmot -- well, isn't a burrow a home?
 
6:02 PM
hmm burritos
 
@barbarabeeton Yes, but above you seemed to say that "there's no place like home."
 
@marmot -- home is the place you can always go back to and be welcome.
 
6:19 PM
@barbarabeeton ... and there is a big supply of honey liquor. ;-)
 
@manooooh learn unknown latex?
 
6:30 PM
@CarLaTeX what is unknown latex?
 
6:44 PM
@marmot -- hmmm. rhum agricole is pretty good too.
 
@barbarabeeton Yes, but being a marmot, I prefer honey liquor....
 
@marmot -- totally understandable. one of my co-workers is also a beekeeper. i shall ask him if he's familiar with honey liquor.
 
Mate > honey liquor
 
@barbarabeeton He must be. Otherwise there is no point in keeping bees. ;-)
 
7:06 PM
@manooooh Fernet-branca >> mate
 
@manooooh honey liquor + marmot = drunk marmot ;-)
 
HAHAHA
@PauloCereda :OOOO
 
7:28 PM
@manooooh I had to find a word with initial u (lul)
 
@CarLaTeX let us lerax ;-)
 
@marmot let us larmot :)
 
lul (?)
@marmot I am trying to graph x^3 - y^2 - y = 0 using \addplot[red,thick] {x*x*x-y*y-y}; but Overleaf gives me error
It says: `! Package pgfplots Error: Sorry, you can't use 'y' in this context. PGFPlots ex
pected to sample a line, not a mesh. Please use the [mesh] option combined with
[samples y>0] and [domain y!=0:0] to indicate a twodimensional input domain.`
 
@manooooh You wrote it
 
So I did \addplot[mesh, samples y>0, domain y!=0:0,red,thick] {x*x*x-y*y-y}; but now 4 errors appears D:
@CarLaTeX hahaha, good wordplay
I do not wanna to make a mesh but an implicit function
 
8:16 PM
You might to want to use contourplot for that:
\documentclass[tikz,border=3.14mm]{standalone}
\usepackage{pgfplots}
\pgfplotsset{compat=1.16}
\begin{document}
\begin{tikzpicture}
\begin{axis}[view={0}{90}]
\addplot3 [
contour gnuplot
] {x*x*x-y*y-y};
\end{axis}
\end{tikzpicture}
\end{document}
compile e.g. with `pdflatex -shell-escape` and you need to have gnuplot installed. This is described in section 4.6.8 Contour Plots of the pgfplots manual.
 
@marmot thanks, I read something about gnuplots, but I would like not to use pdflatex -shell-escape method. I was making a question in the site
 
@manooooh In this case you can solve the equation by hand: x=pow(y*y+y,1/3). All you need to do is a parametric plot then.
 
@marmot using y=sin((x) r) and x=pow(sin((x) r))*sin((x) r))+sin((x) r),1/3)?
 
\documentclass[tikz,border=3.14mm]{standalone}
\usepackage{pgfplots}
\pgfplotsset{compat=1.16}
\begin{document}
\begin{tikzpicture}
\begin{axis}
\addplot[samples at={-3,-2.9,...,-1,-0.95,...,1,1.1,...,3}] ({pow(x*x+x,1/3)},x);
\end{axis}
\end{tikzpicture}
\end{document}
 
@marmot do you want me to post a question? No problem for me, this kind of problems should not be solved in chat
@marmot wait, that is not the function I think
It should be
x^3 - y^2 - y = 0
0
Q: Graph x^3 - y^2 - y = 0 using tikzpicture environment [Overleaf]

manoooohPlease consider this MWE: \documentclass{article} \usepackage[utf8]{inputenc} \usepackage{pgfplots} \pgfplotsset{compat=1.8} \pgfplotsset{soldot/.style={color=black,only marks,mark=*}} \pgfplotsset{holdot/.style={color=red,fill=white,very thick,only marks,mark=*}} \begin{document} \begin{cent...

@marmot ^^^^^^ answer there :)
 
8:34 PM
@manooooh Mate brasileiro > mate argentino :)
 
@AlanMunn lol! What is the difference, @PauloCereda?
 
@manooooh Argentinian mate is more bitter and drier, Brazilian mate is greener and less bitter.
 
@AlanMunn I did not notice that. But here we add sugar to sweeten it more
 
@manooooh Done. ;-)
 
@manooooh Yes, a Brazilian would never add sugar. But @PauloCereda is not a gaúcho, so he has strange tastes in mate. :)
 
8:38 PM
@manooooh In the above, I did not care about the signs, sorry. Too much honey liquor. ;-)
 
@marmot hahaha
@marmot you make me put the colors and change the style of the axes again... anyway you like me :)
 
@manooooh e.g. follow the discussion starting here: chat.stackexchange.com/transcript/41?m=3928583#3928583 (You need to click on to the next day too.)
 
ducks and marmots are friends...
\documentclass[tikz,border=3.14mm]{standalone}
\usepackage{pgfplots}
\pgfplotsset{compat=1.16}
\begin{document}
\begin{tikzpicture}
\begin{axis}[axis lines=middle]
\addplot[red,domain=-1.1:0.6,samples=101]
({sign(x*x+x)*pow(abs(x*x+x),1/3)},x);
\end{axis}
\end{tikzpicture}
\end{document}
@manooooh red will make the plot red and axis lines=middle will put the axis lines in the middle.
 
@AlanMunn wow, that was a long time ago. Probably Paulo is getting less sensitive to taste :P. That is a huge book! <3
@marmot yes yes, I never caught the difference between axis lines=middle and axis lines=center
 
@manooooh Yes, it is confusing. Also why it is xlabel and not x label, say...
@manooooh Need to decouple now...
 
8:45 PM
lul!
@marmot wait please!!
@marmot I also want to plot the line y=0
 
@manooooh \addplot[blue,domain=-1:1]{0};
 
So I did \addplot{0}; but 1) there appears black ugly squares, 2) the graph is stretch
 
@manooooh Yes, @DavidCarlisle has produced more than bugs undocumented features in his career.
 
Oh, I forgot the domain
@marmot thank lul!
@AlanMunn :)
 
@AlanMunn hmm
 
8:48 PM
@DavidCarlisle hmm
 
@DavidCarlisle ooh
 
9:00 PM
@DavidCarlisle could you run l3build check on graphics-def? (I uploaded a new build lua)? I'm wondering if the failures I get are in part windows specific (the object numbers are shifted by one in the pdftex and luatex test).
 
@UlrikeFischer looking..
Check failed with difference files
- ./build/test/dvips001.xetex.diff
- ./build/test/luatex001.luatex.diff
- ./build/test/pdftex001.pdftex.diff
- ./build/test/xetex001.luatex.diff
- ./build/test/zz.luatex.diff
- ./build/test/zz.xetex.diff
@UlrikeFischer (ignore the zz ones)
 
@DavidCarlisle Do you get id shifts in the pdftex and luatex tests too?
 
+ Invalid UTF-8 byte or sequence at line 11 replaced by U+FFFD. then ! .......\kern 321.2 (no space after kern now) ah this one:
! <clip-1-2.jpg, id=27, 146.34676pt x 128.8815pt> new code gives 26 not 27
 
@DavidCarlisle yes this one. It seems to be a change from texlive 2017 to 2018.
 
@UlrikeFischer why would we get one less object in new code, I guess just check it in anyway, not a change we can really control
 
9:16 PM
@DavidCarlisle I'm not even sure if there is really one object less. It could be also a change in the ordering. I will recreate the four tlg and upload, I think too that is not something that we can control, I only wasn't sure if there was a general change or only on windows.
 
9:29 PM
i've just received some new fun reading material -- an advance copy of the ieee annals of the history of computing, the first of two issues on desktop publishing. (reason for early acquisition: it contains the first part of a two-part article on tex by dave walden, karl berry and me.) it will be difficult for me to avoid reading it this weekend, and instead work on tugboat. some material is (or will be) online at history.computer.org/annals/dtp -- enjoy.
 
@barbarabeeton !!!!!!!!
@barbarabeeton are you the one with the red heels?
There is only woman, omg!
BTW congrats for the material :) :) <3
 
@manooooh -- no, i wasn't actually there -- an "absentee" author. unless i'm mistaken (which certainly is possible), dave walden is the third person from the right. and don knuth is in the back row, below (and slightly to the right of) the third red logo from the right at the top of the wall.
 
cis
10:23 PM
Hey my friend from the U.S.!

I am sorry, I have a question:
Is it correct to say "an inline table" (because "a table")?
 
11:10 PM
@cis Choose “a” or “an” depending on the following word, not on the noun it refers to.
 

« first day (2917 days earlier)      last day (2018 days later) »