« first day (3173 days earlier)      last day (1758 days later) » 

2:04 AM
\documentclass{article}
\usepackage{expl3}
\begin{document}
\end{document}
! LaTeX Error: File `l3backend-pdfmode.def' not found.
$ tlmgr show l3backend
[...]
revision:    51533
[...]
cat-date:    2019-07-02 14:50:16 +0200
Hm.
@JosephWright ^^^^^
 
 
2 hours later…
4:12 AM
@HenriMenke You've not done a full update ...
 
 
1 hour later…
5:14 AM
$ tlmgr update --self --all
tlmgr: package repositories
	main = mirrors.ctan.org/systems/texlive/tlnet (verified)
	tlcontrib = contrib.texlive.info/current (verified)
	luatex-dev = hmenke.github.io/texlive-luatex-dev (not verified: pubkey missing)
	pgf-development = pgf-tikz.github.io/pgf/tlnet (not verified: unsigned)
tlmgr: saving backups to /usr/local/texlive/2019/tlpkg/backups
@JosephWright ^^^
I should be up to date.
 
C:\texlive\2019\texmf-dist\tex\context>tlmgr search --file l3backend-pdfmode.def
l3backend:
        texmf-dist/tex/latex/l3backend/l3backend-pdfmode.def
@HenriMenke ^^^
@HenriMenke --reinstall-forcibly-removed? Just in case ...
@h
@HenriMenke We sent everything to CTAN ... and it works for example in Travis-CI (so completely automated)
 
@JosephWright I found the problem. Something changed permissions on the ls-R files and they got out of sync.
 
 
1 hour later…
6:45 AM
user image
3
@AlanMunn ^^^^ Do you already have some of these?
 
7:33 AM
When xcolor can input colors in RGB (aka 255 based), shouldn't there be an equivalent CMYK for cmyk 255 based?
 
7:52 AM
@daleif I added RGB to color as some drivers (Y&Y if I recall correctly) used that as the native syntax, can't recall ever being asked for a CMYK version.
 
@DavidCarlisle our design office specifies colors in base 255, is there a tracker where I can add a feature request. CMYK colors for printing is a huge pain in LaTeX, I can never get them to look right.
 
@daleif it would need to be in xcolor rather than color thee days so:
%% ----------------------------------------------------------------
%% Copyright (C) 2003-2016 by Dr. Uwe Kern <xcolor at ukern dot de>
%% ----------------------------------------------------------------
%%
%% Please send error reports and suggestions for
%% improvements to the above email address.
 
Thanks, I'll send an email to Uwe
 
 
2 hours later…
9:39 AM
@HaraldHanche-Olsen oh no
@DavidCarlisle oh no
 
People are caring for me ^^^ ;-)
 
@UlrikeFischer You can read this in many ways ... do you know Stanislaw Lem? ;-)
 
@marmot I know Lem but it's long ago that I last read it. What does it say to this topic?
 
@UlrikeFischer Well, something along the lines that they may want to torture you so a heart attack is too quick...
 
@marmot We seem to be back to a Fish named Wanda ;-)
 
9:53 AM
@UlrikeFischer Grilled or fried?
 
 
1 hour later…
11:07 AM
@marmot No, they're commemorative 50 cent pieces not in general circulation. :)
 
11:19 AM
Do we have any information on verifying CMYK colors in PDF files? Including specifying whatever color specification that CMYK color is to be using.
 
@daleif Perhaps Preflight can extract some data?
 
I have a beach flag design I want to do. It can hit the correct BG color in RGB, but the printer wants CMYK and I cannot fhit or verify the correct color.
@PauloCereda I'm primarily on Linux so no Adobe available (do have a Win10 VM though)
The interesting part is that most of our own printers are CMYK, and none of them hit the correct color if I'm playing with cmyk
@PauloCereda where is that preflight stuff?
 
@daleif It's an Adobe tool, I guess. We could ask @UlrikeFischer, she's the expert. :)
 
@PauloCereda I'm not an expert for preflight - I only use it to check if I broke the pdf ;-).
@daleif It's a tool that you get with adobe pro (but also in software from callas which imho actually write the tool) to check various aspects of pdf files.
 
@UlrikeFischer ooh we have @DavidCarlisle for this purpose too. :)
 
11:33 AM
@UlrikeFischer took a while to fine, unintuitive interface
 
@daleif conversion from rgb to cmyk is a bit difficult - See e.g. tex.stackexchange.com/a/436080/2388
 
@UlrikeFischer The interesring part is that our design department has a specification of the color in CMYK, but 255 based) and using CMYK: Coated FOGRAF27 (ISO 12647-2:2004), the latter I have no idea how to apply. Preflight does seem to give the right /255 two digit colors, it just looks completely wrong on screen.
 
@daleif you should be able to load the profile with pdfx or colorspace. But beside this: cmyk is for printing, so don't worry about the screen.
 
11:49 AM
@UlrikeFischer I just do not want to receive and have to pay for something that looks completely the wrong color. We should at least be able to print it correctly on our own CMYK printers, shouldn't we?
Which of the xpdf options should be used? I'm playing with x-1a that does not seem to be getting a color profile.
Hmm, colorspace seems to work a lot better
lol, but definespotcolor is better at hitting a sort of right color if it is given an RGB to begin with, cmyk devided by 255 not so much
 
12:05 PM
@daleif I would have to study the documentation too. I don't have so much to do with colors normally, I only know how to look in the pdf to check what is there and can make a straight face if someone says "colorprofile" ;-).
 
ooh das Farbprofil
 
@UlrikeFischer colorspace seems to work given the fact that I do have a Pantone specification as well. Evince cannot print it correctly (but that seems to be a known issue), adobe acrobat does hit sort of the correct color on our own CMYK printer.
 
12:56 PM
Hi, I have a short question: I'm working on a LaTeX document, often I make a mistake and the document doesn't compile, when I fix this mistake and compile again the compiler "hangs" at a certain point, telling me that

File ended while scanning use of \@writefile.
<inserted text>
\par
l.9 \begin{document}
?

I can then press enter and the file compiles as normal. The next time I compile this message doesn't appear, it only appears directly after a compilation failed or was aborted.
So I'm wondering: What kind of a situation can make this happen?
 
@s.harp Sometimes deleting the *.aux file (and other intermediate files) helps.
 
@s.harp you are quiting the run in a bad way and corrupting the aux file, if that happens you can delete the aux file, but quiting cleanly is better.
 
ah, yes deleting the aux files after a mistake gives normal compilation
 
@s.harp so if you make a mistake and get a tex error what do you do?
 
usually I close the terminal and fix the error then re-compile (or was this a socratic question?)
 
yo'
1:01 PM
@s.harp no, it wasn't a socratic question. It's better to hit x<enter> rather than to directly close the terminal.
 
@s.harp I guessed as much, closing the terminal aborts the running job with no chance for it to flush its write buffers so if it was writing the aux file and you kill it you end up with a half-written command and mis-matched { in the aux file, so when that is input next time bad things happen
 
ah
entering x compiles a document up to the page of the error, thats interesting
 
@s.harp sometimes the aux file gets corrupted even if you do use x (eg if the error is actually in a section head and tex can not recover while writing the table of contents information) but it is fairly rare, whereas if you kill the program externally from the operating system then any open file might get corrupted in this way.
 
@DavidCarlisle I see you are more modern than me: using \ifincsname and causing issues
 
@JosephWright I was just reading that and trying to remember if it was me or Frank:-)
@JosephWright I guess we have to back it out for now anyway
@JosephWright Frank:-)
 
yo'
1:25 PM
@s.harp another option is that if the next compilation fails, you can try to enter q. Most times it helps and the compilation after this will be ok. If it does not help, then you have to delete the aux files though.
 
@daleif -- You can get at least an idea if the color is correct if output printed on three different cmyk printers looks identical. My guess is that color printers at most organizations aren't regularly calibrated, so one isolated example may not be a good test. Maybe the design department is more careful about calibration, so maybe you could persuade them to do a test print when you think you've gotten it right.
 
yo'
@barbarabeeton in CMYK, many printers try to do corrections to meet the standard. The question is how successful they are; and this highly depends on the brand, series, price, ...
 
@yo' -- I guess I'm really familiar only with the "final" step, at the print shop. There, proper calibration is essential. The "local" printers at AMS were maintained regularly, but color calibration either wasn't careful, or wasn't included, because a color image didn't necessarily come out the same on two different machines. (There was a calibrated proof printer to be used if it mattered.)
 
yo'
@barbarabeeton yeah, of course. I just wanna say that the printer does weird stuff to convert CMYK to the actual inputs to the inks/lasers. Calibration is about it knowing how to convert the data. So I think we're in agreement here :)
 
1:43 PM
@DavidCarlisle @JosephWright got my mail?
 
@UlrikeFischer from earlier in the day? yes. will reply in a bit
 
2:39 PM
Today marks a new day in the 26-year history of Red Hat. IBM has
finalized its acquisition of Red Hat, which will operate as a distinct
unit within IBM.
 
@PauloCereda So I guess now they'll be called Blue Hat.
 
@AlanMunn ooh I see what you did there. :)
 
@AlanMunn Makes sense, no one wants to give them away.
 
@AlanMunn in the early beginning, Fedora's logo was a blue fedora. It makes sense now:
RH and Fedora were my favourite. :)
 
3:10 PM
@FaheemMitha What else should people use, C?
 
@KhaledHosny Weird PASCAL-in-some-strange-wrapper, of course
 
@JosephWright Of course, silly me :)
 
@JosephWright Fortran
 
@DavidCarlisle Do you write Fortran? you would be the second person I know who does Fortran.
 
@KhaledHosny Real Programmers use FORTRAN.
@KhaledHosny Note the capitalisation. From the days of FORTAN IV, when men were men and ducks were ducks.
 
3:15 PM
I'm really confused by the latest LuaTeX development and not sure what should I do. We certainly don't want 5 subtly different LuaTeX engines in TeX Live, right?
 
@KhaledHosny I don't but those people around this office who do are the ones who keep my family fed and sheltered:-)
 
@HaraldHanche-Olsen You should include line number in the name too :)
 
@DavidCarlisle Your house is built from punch cards?
 
@DavidCarlisle Whatever puts food on the table
 
@HaraldHanche-Olsen I weirded out all the engineers I worked with by writing my Fortran code in lowercase.
 
3:18 PM
@AlanMunn You are mean.
 
@KhaledHosny yes we (latex team) would like to help get a better situation but not clear what to say (or who to say it to)
 
@DavidCarlisle (that would be 6 if we count "lmtx", so that one is probably not "subtly different).
@DavidCarlisle I don't know what to do either.
 
@DavidCarlisle TeXhax is interesting today
@KhaledHosny Can't even build the plain format with lmtx, pending feedback from Hans, so I wouldn't worry about that one
 
@JosephWright I see.
 
@JosephWright for a strange definition of interesting
 
3:32 PM
@KhaledHosny I did try, moderately but not extremely hard (the files must be there ...)
 
3:49 PM
@UlrikeFischer Regarding our little exchange in the comments to this answer of yours: It appears that it has been fixed (whether by Will or not I do not know):
\documentclass{book}
\usepackage{unicode-math}
\setmathfont{Latin Modern Math}
\message{A: \the\fontdimen5\textfont2}
\setmathfont{STIX Two Math}[range={"2987-"2988}]
\message{B: \the\fontdimen5\textfont2}
\setmathfont{STIX Two Math}
\message{C: \the\fontdimen5\textfont2}
\begin{document}
\end{document}
A: 4.31046pt B: 4.31046pt C: 4.7305pt
 
@KhaledHosny There are other languages besides C and C++.
@KhaledHosny 5 subtly different LuaTeX engines? Do tell.
 
@FaheemMitha Yes, but a lot of 'core' code is written in C or C++
 
I'm hoping for one LuaTeX engine. That seems like quite enough.
@JosephWright Core of what?
 
@FaheemMitha luatex lmtx luajittex luahbtex harftex, ....
 
@FaheemMitha Operating systems, libraries, etc., as in 'serious code'
 
3:58 PM
@JosephWright Oh. Yes, that's true. Unfortunately.
@DavidCarlisle Eek.
@KhaledHosny You're integrating Harfbuzz with LuaTeX, right?
 
@FaheemMitha that accounts for the last two in my list:-)
 
@DavidCarlisle Two different integrations?
 
@FaheemMitha There are, but which would be suitable for a library like HarfBuzz?
@DavidCarlisle you missed luahbjittex
 
@FaheemMitha long story, end of which is not yet written (hopefully)
 
@KhaledHosny Dunno. Languages that people actually seem positive about include Common Lisp, Haskell, and O'Caml. I'm sure there are others...
 
4:01 PM
@KhaledHosny we could include sile as well for good measure
 
Whether you would be able to do what you want in them is another question, of course.
 
@FaheemMitha I can't see any of those working at the level you need here
 
@DavidCarlisle That is neither a TeX engine nor included in TeX Live :)
 
For legacy reasons we seem to be stuck with C and C++ for the forseeable future.
 
@KhaledHosny i know:-)
 
4:02 PM
@DavidCarlisle For performance reasons, or other reasons?
 
@FaheemMitha performance and other, but mostly performance
 
@FaheemMitha that is the only question, you pick the language good for the job, not the language then the job
 
@DavidCarlisle Hmm.
@KhaledHosny Since I have only a sketchy idea what the job is, so I can't really comment.
But it's depressing when the answer to a question is C++. I wrote it for most of a decade, and learned to hate it.
Though I doubt I would have liked Fortran much better.
It would be nice if people would try more actively to get away from it (C++).
@KhaledHosny Anyway, whatever you are doing, it sounds like a big job.
 
@DavidCarlisle As far as I know, one can write quite efficient code in Common Lisp and Haskell, but it sometimes requires deep knowledge of the language. There are plenty of traps for the unwary, especially in Haskell.
 
Haskell seems to be growing in popularity.
@HaraldHanche-Olsen Do you know either of both of them?
 
4:10 PM
@HaraldHanche-Olsen people writing code in assembler as C is not enough control over register use, I can't see suggesting they use haskell would go down well.
 
SBCL is fairly fast. Though IMO the garbage collector is an issue.
 
@FaheemMitha I have written a bit of Common Lisp, but nothing large, and most of it long ago. I've been meaning to learn Haskell for several years now, but life intervenes.
 
@JosephWright I read the thread about \cr. It's curious where somebody can arrive in order not to admit having read too much in a control sequence name, showing utter ignorance about the history of TeX, when saying that it was probably developed on a Macintosh, because it used <CR> as record terminator. ROTFL
 
@DavidCarlisle Also since C API seems to be the API exposed by all systems in use today, any low level library needs to expose a C API to be usable by higher level ones.
 
@HaraldHanche-Olsen Haskell definitely sounds interesting.
 
4:11 PM
@egreg you only saw half the thread (phil and I got the directors cut with extra interesting scenes)
 
I'd like to learn CL better. I learned a little bit, but not really enough to do much.
 
@DavidCarlisle I can imagine!
 
@DavidCarlisle Probably not, but there is a foreign function interface that can be used to call out to code written in C, assembly or whatever.
 
@KhaledHosny certainly. Our "Fortran library" is mostly used by non fortran programmers via it C header interface
 
I imagine it is a very effective language if you know what you are doing. But knowing takes some doing. And the implementations definitely need much to be desired, particularly from a performance standpoint.
Certainly one cannot fault C and C++ from a performance and memory usage standpoint. And they offer all the control one could desire.
 
4:13 PM
I wouldn't want to the person to tell Karl some software in TeX Live needs Rust/Haskel/Lisp compiler, he was very upset when HarfBuzz stopped building with gcc 4.8.
 
@HaraldHanche-Olsen yes but writing a few million lines of fortran then calling it from a shell haskell function isn't really "using haskell" (it's useful though and that's how we expose teh functionality to pythin or vba for excel etc)
 
@KhaledHosny Is that a requirement?
 
@KhaledHosny I can tease him. :)
 
@DavidCarlisle But of course, for writing libraries like your colleagues are doing, squeezing every bit of performance out of the code is worth it.
 
@KhaledHosny I convinced him to accept my Java binaries. :)
 
4:14 PM
@DavidCarlisle Fortran and new "systems programming" languages like Rust provide C API for this very reason.
 
@DavidCarlisle I'm ignoring the jit versions but count the experimental versions and have formats with and without harf code ...
 
I take it Harfbuzz requires very fast code, then.
 
@PauloCereda Because you built them yourself, now go tell him he needs a java compiler now :)
 
@DavidCarlisle But probably only a few thousand lines of those millions need to be optimised really hard.
 
@KhaledHosny that's a good point. :D
 
4:15 PM
@HaraldHanche-Olsen well that's true but the trick is to know which bits that is:-)
 
@FaheemMitha For software included in TeX Live, yes.
 
@KhaledHosny Ok
 
@DavidCarlisle Mm, like advertisements.
 
@HaraldHanche-Olsen Optimization is a fantastic research area when studying compiler techniques, for a wacky definition of fantastic. :) Disclaimer: I like it a lot.
 
@HaraldHanche-Olsen What is your general feeling about Common Lisp? I guess neither of us has used it a lot.
 
4:21 PM
@FaheemMitha It does, opening a medium size document it LibreOffice typically calls HarfBuzz a millions (or may be billions, it was a while since I last measured this) times. Just asking the compiler to optimize HarfBuzz gave measurable performance difference.
 
@KhaledHosny Which optimization level did you use?
 
@FaheemMitha I find it very pleasant to work with. It's easy to mix various programming styles, and debugging is so much easier since exceptions are handled without rewinding the stack first. Often, one can recover from errors and move on.
 
@KhaledHosny Wow
@HaraldHanche-Olsen Yes, the integrated debugger is very nice. Though I haven't really learned to use it properly. Much nicer than dumping the user on the floor and leaving him/her to figure out what went wrong. It has other nice features too. I particularly like the handling of scoping, which is typically a nightmare.
And SLIME is super cool, though that's technically not part of the language.
But I suppose if I wanted something to run as fast as possible, it's not what I would be using.
 
@FaheemMitha yay for alpha reduction. :)
 
Though at least in scientific computatation, using sensible algorithm and data structures matters far more than the language you use.
E.g. Mercurial beats the crap out of at least one version control system written in C++.
@PauloCereda alpha reduction?
 
4:26 PM
@FaheemMitha lambda calculus thingy.
 
@PauloCereda Oh
Sorry, I'm CS-illiterate.
 
@FaheemMitha no worries, it's one of those functional thingies. :)
 
@PauloCereda And if you don't understand it, you're obviously dysfunctional. :D
 
@HaraldHanche-Olsen LOL
 
4:28 PM
@PauloCereda I am mean.
Apr 9 '18 at 14:40, by Paulo Cereda
@HaraldHanche-Olsen: you are not mean :)
oh.
 
@HaraldHanche-Olsen that's an eta reduction, do not worry. :)
 
@HaraldHanche-Olsen It's also great if you are refactoring. Since anything between matches parens is a valid piece of syntax.
 
@PauloCereda LibreOffice seems to use either -O2 or -Os, but it wasn't passing it to HarfBuzz build initially
 
@FaheemMitha Editing Lisp code in emacs can be amazingly effective for that reason. No wonder, that's where emacs came from.
 
@HaraldHanche-Olsen Yes, it's a good experience in general.
If one tries to refactor C++, it's basically an exercise in being flogged by the compiler.
 
4:31 PM
@KhaledHosny ah interesting!
@FaheemMitha you'd be surprised about the things compilers would do to optmize a code. :)
 
@PauloCereda make it go faster and wronger
 
@DavidCarlisle exactly! Speciallly when in the red zone.
@DavidCarlisle that would be the wrongest
English is wonderful
 
I always found that the data structures mattered the most. C really likes continguous blocks of memory, and vectors.
 
@PauloCereda we have automated suites setting optimisation levels separately for each routine for each compiler
 
Keep everything in a slab of memory, and do the bookkeeping oneself. It's not pretty, but it's fast.
Of course, I was doing numerical work. No idea what a font library would need.
 
4:35 PM
@DavidCarlisle really? Very nice!
 
@PauloCereda it does mean a request for "can you compile with the foobar compiler on a widget4 chip architecture" means it takes three months not just type make and come back in the morning....
 
@DavidCarlisle :)
@DavidCarlisle Do you gentlemen use gcc or llvm?
 
@PauloCereda actually no :) for once
 
@DavidCarlisle ah it was a sad reality then
 
@PauloCereda we use everything intel, microsoft gcc, llvm, several commercial fortran compilers, our fortran compiler, ...
 
4:39 PM
@DavidCarlisle ah I see
 
 
2 hours later…
7:04 PM
@JosephWright -- I just accidentally flagged a posting on the main site thinking I was connected to meta. Apologies for the noise.
 
@StefanKottwitz Hi, Stefan can you see a strange comment?
@Sebastiano That you assume the person to be a man speaks of your limited mind — wasa 12 mins ago
@barbarabeeton Hi, please can you help me to translate correctly this comment?
Is it a bad comment?
 
@Sebastiano -- It's not really polite, so I will just comment. The figure in the car need not be a man; it's inappropriate to assume that it is. (Actually, it looks to me to be a drawing of a crash-test-dummy.)
 
7:30 PM
@barbarabeeton It's also inappropriate to profit of “political correctness” for insulting people.
 
@egreg -- True. I said it wasn't polite.
 
@barbarabeeton @egreg I have done a compliment for the :-( drawing. I wrote man but it could be a woman. In the Bible the woman is called "uoma". I am not a sexist. But am I the one seeking insults?
 
8:15 PM
@Sebastiano Some people are oversensitive about “political correctness” and pretend not to understand that in a site like this the language barrier can be high. A truly “politically correct” person would have said “please, mind that it could also represent a woman”.
 
8:26 PM
@egreg also apart from language barriers, not everyone can be a brilliant artist and unambiguously depict men and women.
5
A: Multiple captions under a single figure

David CarlisleYou can have as many \caption as you need in a figure \documentclass{article} \usepackage{graphicx} \begin{document} \listoffigures \begin{figure}[thp] \begin{minipage}{.5\textwidth} \centering \includegraphics{man} \caption{man} \end{minipage}% \begin{minipage}{.5\textwidth} \centering \in...

 
@DavidCarlisle Unambiguous? The one on the right is a long-haired man wearing a kilt.
 
@Sebastiano This is how I'd write your piece of text.
\documentclass[a4paper,12pt]{article}
\usepackage[top=2.5cm,bottom=2.2cm,
            left=3.2cm,right=1.5cm,headsep=10pt,
            a4paper]{geometry}
\usepackage{amsmath,amssymb,mathrsfs}
\usepackage{newtxtext,newtxmath}

\newcommand{\diff}{\mathop{}\!d}

\newcommand{\tder}[2]{\frac{\diff #1}{\diff #2}}
\newcommand{\pder}[2]{\frac{\partial #1}{\partial #2}}

\newcommand{\iv}[3][c]{%
  \csname iv@#1\endcsname{#2}{#3}%
}
\makeatletter
\newcommand{\iv@gen}[4]{#1#3,#4#2}
\newcommand{\iv@c}{\iv@gen{[}{]}} \let\iv@cc\iv@c
 
@AlanMunn it is however clearly labelled so even picky linguists with Scottish roots will not be confused.
@egreg all those @ and not a _: to be seen, disappointing...
 
@DavidCarlisle But then your brilliant artist claim is unsupported. You can't have it both ways. :)
 
@AlanMunn I can.
 
8:36 PM
@egreg Thank you for defending me with all your heart. @DavidCarlisle wasa is a user who offended me David.
 
@DavidCarlisle Well true, because the picky linguist can point out that your statement is in fact true: not everyone can be a brilliant artist.
 
@Sebastiano yes although your comment under your question was out of line, if it gets flagged you can't complain.
 
@egreg If you could please insert your answer in my question I will offer you a pizza with pineapple :-) I will pay the pizza also for @DavidCarlisle
 
@DavidCarlisle Left is my three year old nephew representation of his dad. He's a better painter than you.
 
@Sebastiano I didn't defend you I just commented that it wasn't a duplicate (and the duplicate vote was retracted) so you may want to delete your comment
@egreg my drawing is better, so there.
 
8:40 PM
@DavidCarlisle No, it wasn't flagged right away. I did not complain, I just wanted to call your attention that it was an unfair user with me.
 
@Sebastiano We know and we acknowledged it. Please, move on.
 
@Sebastiano I had seen the previous question (and the discussion of it here) but i repeat that you should delete your comment.
 
@DavidCarlisle Done!
 
@Sebastiano so now that comment thread ends with a comment that I am sure all agree with.
 
@egreg I wanted put your code as community wiki but I have my name and this i don't no. Can you add another answer, please with your code? That way it's in my question and I don't run the risk of losing it.
\documentclass[a4paper,12pt]{article}
\usepackage[top=2.5cm,bottom=2.2cm,
            left=3.2cm,right=1.5cm,headsep=10pt,
            a4paper]{geometry}
\usepackage{amsmath,amssymb,mathrsfs}
\usepackage{newtxtext,newtxmath}

\newcommand{\diff}{\mathop{}\!d}

\newcommand{\tder}[2]{\frac{\diff #1}{\diff #2}}
\newcommand{\pder}[2]{\frac{\partial #1}{\partial #2}}

\newcommand{\iv}[3][c]{%
  \csname iv@#1\endcsname{#2}{#3}%
}
\makeatletter
\newcommand{\iv@gen}[4]{#1#3,#4#2}
\newcommand{\iv@c}{\iv@gen{[}{]}} \let\iv@cc\iv@c
 
8:55 PM
@Sebastiano That's not so valuable. For instance, you cannot resize the brackets.
 
9:08 PM
@egreg Everything that is culture, knowledge and knowledge is important to me. It is not only necessary to know what is good but also to know how to recognize the validity of what apparently could be not good.
 
 
2 hours later…
10:55 PM
Good evening. Suppose two tables, one very long, more than one page; the other, shorter, it fits in single page. I'd like to put them side by side, aligned by top, that is, both have header aligned at page 1.
I tried supertabular bu the shorter one (on the right) moves to second page.
Any reference to do that?
 
11:42 PM
@Sigur Is this in a two column document? Or are you using multicol? So you want the rest of the second table to appear on the next page? And then what will fill in the rest of the first page?
 
@Sigur Something like this?
\documentclass{article}
\usepackage{longtable,booktabs}
\usepackage{lipsum}

\begin{document}

\lipsum[1][1-4]

\vspace{\LTpre}
\noindent\begin{minipage}[t][0pt][t]{\textwidth}
\raggedleft
\vspace{0pt}
\begin{tabular}{ccc}
\toprule
AA & BB & CC \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
1 & 2 & 3 \\
\bottomrule
\end{tabular}
\end{minipage}

\vspace{-\LTpre}

\begin{longtable}[l]{cccc}
\toprule
AAAAA & BBBBB & CCCCC & DDDDD \\
 

« first day (3173 days earlier)      last day (1758 days later) »