« first day (1906 days earlier)      last day (3028 days later) » 

cfr
cfr
12:16 AM
@AlanMunn Do you have a good answer to tex.stackexchange.com/questions/287601/…? I told the OP not to award the bounty until at least you'd looked at it and seen if you could do better.
 
egreg
egreg
12:41 AM
@cfr Done. :)
 
cfr
cfr
12:59 AM
@egreg I upvoted your answer even though you are pretending to be egreg.
 
 
4 hours later…
Faheem Mitha
Faheem Mitha
4:57 AM
Hi. If I have a line \includegraphics[height=750pt]{foo.jpg} in a file (foo.tex), if I compile from one level up, it fails, even though though foo.tex and foo.jpg are in the same directory.
What's a suitable fix for this?
Oh, I see this has been asked and answered already. Never mind.
 
Alan Munn
Alan Munn
@egreg Now for this @DavidCarlisle should swoop in and write an even better answer. :)
 
Faheem Mitha
Faheem Mitha
5:15 AM
Ah, there are other people here. Exciting.
 
cbward
cbward
Is there a way to customize the QED symbol alignment using thmtools?
 
 
2 hours later…
Paulo Cereda
Paulo Cereda
7:06 AM
@DavidCarlisle You are a genius. <3
 
Joseph Wright
Joseph Wright
@cfr I hope my answer helps, even though it's perhaps not what you want
 
Romain Picot
Romain Picot
@PauloCereda Quack !
 
Paulo Cereda
Paulo Cereda
@RomainPicot Coin coin! <3
 
Romain Picot
Romain Picot
@PauloCereda Psmith is here?
 
Paulo Cereda
Paulo Cereda
@RomainPicot Not now, I am sorry. :) I am in a hurry because today I return to my travel routine. :)
 
Romain Picot
Romain Picot
7:16 AM
@PauloCereda I wanna ask him if I should work or sleep today :/
 
Paulo Cereda
Paulo Cereda
@RomainPicot ooh that's easy: you could try sleeping for a couple of hours, then you get your energies recharged and ready for a good day of work. :)
 
Romain Picot
Romain Picot
@PauloCereda Good idea my favorite duck ! <3
 
Paulo Cereda
Paulo Cereda
@RomainPicot awww <3
Coin coin!
:)
@egreg: Shaky buses all over again. :)
 
Romain Picot
Romain Picot
@PauloCereda Quack ! <3
 
Paulo Cereda
Paulo Cereda
@RomainPicot I almost wrote Psleep instead of Psmith. :P
 
Romain Picot
Romain Picot
7:22 AM
@PauloCereda You can modify it to provide some answer when using choose like: "sleep", "procrastinate" and so on ;-)
 
Paulo Cereda
Paulo Cereda
@RomainPicot ooh I like this plan. :)
 
Romain Picot
Romain Picot
@PauloCereda he will choose procrastinate instead of thesis if @DavidCarlisle ask him ;-)
 
Paulo Cereda
Paulo Cereda
@RomainPicot :)
Friends, time for me to go to SP. Have a great day!
To infinity and beyond! Quack! :)
 
Romain Picot
Romain Picot
@PauloCereda Coin coin !
 
Faheem Mitha
Faheem Mitha
7:55 AM
Apparently David is a genius. Good to know.
 
egreg
egreg
8:41 AM
@PauloCereda Have a good shake.
 
daleif
daleif
9:05 AM
@egreg here is another one from the person with the mathtools vs unicode-math problem yesterday. I've reduced his MWE a little.
\documentclass{minimal}
\usepackage{amsmath}%%% you may comment this line; no difference
\usepackage{mathtools}
\usepackage{unicode-math}
\setmathfont[Extension=.otf]{texgyrepagella-math}
\setmathfont[Extension=.otf,range={"25C3}]{xits-math}
\begin{document}
\[
\mathfrak{P}
\]
\[
\mathtt{X}
\]
\[
\mathfrak{P}
\]
\end{document}
why does the second \mathfrak{P} disappear?
To my knowledge mathtools should not mess with the fonts.
 
egreg
egreg
@daleif Well, it seeems likely a bug in mathtools. For some reason, that results in making \mathfrak trying to use cmtt10
 
daleif
daleif
@egreg well replace the contents of mathtools with
\ProvidesPackage{mathtools}%

\endinput
and you get the same problem
 
egreg
egreg
@daleif Interesting. I'll look into it later.
 
David Carlisle
David Carlisle
9:25 AM
@FaheemMitha I'm disappointed that you only just noticed!
 
daleif
daleif
@egreg here is an even simpler MWE
\documentclass{minimal}
\usepackage{amsmath}%%% you may comment this line; no difference
%\usepackage{mathtools}
\makeatletter
\@fileswithoptions\@pkgextension[]{mathtools}
\makeatother
\usepackage{unicode-math}
\setmathfont[Extension=.otf]{texgyrepagella-math}
\setmathfont[Extension=.otf,range={"25C3}]{xits-math}
\begin{document}
\[
\mathfrak{P}
\mathtt{X}
\mathfrak{P}
\]

\end{document}
running it with xelatex
interestingly, it works with lualatex
 
David Carlisle
David Carlisle
@daleif or just 
\documentclass{article}

\expandafter\def\csname ver@mathtools.sty\endcsname{2016/01/01}

\usepackage{unicode-math}
\setmathfont[Extension=.otf]{texgyrepagella-math}
\setmathfont[Extension=.otf,range={"25C3}]{xits-math}
\begin{document}
\[
\mathfrak{P}
\]
\[
\mathtt{X}
\]
\[
\mathfrak{P}
\]
\end{document}
 
daleif
daleif
@DavidCarlisle yes, I'm a bit curious as to why it only seems to be xelatex that is affected
 
David Carlisle
David Carlisle
9:42 AM
@daleif oh I didn't try luatex. anyway needs more debugging than I can do at work, later or just get @JosephWright to ping Will :-)
 
Joseph Wright
Joseph Wright
@DavidCarlisle Email fixed
 
David Carlisle
David Carlisle
@JosephWright morning! (so were the backlog queued or were they really bouncing them as it appeared?)
 
Joseph Wright
Joseph Wright
@DavidCarlisle Do seem to have bounced :-(
 
Christian Hupfer
Christian Hupfer
10:03 AM
Good maen
 
Danu
Danu
10:31 AM
Hi all
I've encountered the challenge of drawing a torus knot in TikZ.
This is basically a line that winds around the torus in a nontrivial way (like going around the hole once while wrapping around the donut shape several times).
Is there any way to do this without drawing the lines by hand?
Reference:
Torus knot
In knot theory, a torus knot is a special kind of knot that lies on the surface of an unknotted torus in R3. Similarly, a torus link is a link which lies on the surface of a torus in the same way. Each torus knot is specified by a pair of coprime integers p and q. A torus link arises if p and q are not coprime (in which case the number of components is gcd(p, q)). A torus knot is trivial (equivalent to the unknot) if and only if either p or q is equal to 1 or −1. The simplest nontrivial example is the (2,3)-torus knot, also known as the trefoil knot. == Geometrical representationEdit == A torus...
 
Joseph Wright
Joseph Wright
@DavidCarlisle Do we want luatex85 or just the alias stuff to go to CTAN?
 
David Carlisle
David Carlisle
10:50 AM
@JosephWright both, I thought we said, that's what I made the script zip up yesterday, I was originally thinking luatex85 might have more stuff (eg suppress loading of ifpdf etc) but currently seem to be getting by without doing that if all the aliases are there, might still be needed though once testing in the wild starts. But anyway need to synchronise with latex tools release (since shellesc is there now)
 
Henri Menke
Henri Menke
I just clicked the link to the TeX.SX newsletter. There are unsubscribe links in this web preview. I can only recommend trying them!
 
David Carlisle
David Carlisle
@JosephWright luatex.def could go to ctan now, although it fits better if graphics.cfg updated and that I think still goes via mail to texlive list, so I suppose I should start there:-)
@JosephWright do you think we could get it not to say warning (pdf backend): forcing decimal digits to 4 all the time (I assume you get that as well?)
 
egreg
egreg
11:11 AM
@daleif The suspicious code is in unicode-math-xetex.sty, lines 2040–2091. If I remove them, the fraktur P reappears.
 
David Carlisle
David Carlisle
@egreg good hunting, I noticed some differences re \g__um_empty_fam in the log but gave up at that point:-)
 
Danu
Danu
I asked my question on the main site:
0
Q: Is there an efficient way to draw torus knots and related ''complicated lines'' in TikZ?

DanuIn the process of typing up a set of lecture notes for a course, I found myself in need of a nice picture of a torus knot, which basically comes down to drawing a very complicated line. With some trickery, I was able to produce the simplest such knot, which wraps around both circles only once. Th...

tikz-pgf
(after at least being able to produce the 1,1 torus knot on my own :3)
 
daleif
daleif
11:26 AM
@egreg nice catch
 
egreg
egreg
@daleif I guess the patch is trying to cope with \cramped.
 
daleif
daleif
@egreg hmm, if I outcomment those lines, I get an too many }'s error
but I can also just end the file there and indeed the frak P comes back
 
egreg
egreg
@daleif There's a strange nesting (probably wrong) there: you may have also commented out lines 2038 and 2039. I said 2040–2091. ;-)
@daleif If you look at the corresponding part in unicode-math-luatex.sty, you'll see the mysterious nesting.
 
daleif
daleif
@egreg yes, \AtEndOfPackageFile * { mathtools } inside other \AtEndOfPackageFile * { mathtools }, that is not particularly readable.
 
Joseph Wright
Joseph Wright
@egreg Do you plan to answer tex.stackexchange.com/questions/288347/…? (The change needed is trivial ...)
 
egreg
egreg
11:40 AM
@JosephWright I'm reading the numprint manual
 
Joseph Wright
Joseph Wright
@egreg s/n/N in column header
 
egreg
egreg
@JosephWright Yes. Go ahead.
 
Joseph Wright
Joseph Wright
@egreg Oddly enough, I knew pretty much exactly what to look for ...
@egreg OK
 
egreg
egreg
@JosephWright Tabular numbers should be used, too.
 
Joseph Wright
Joseph Wright
@egreg I'm not sure what the OP wants in that sense
@DavidCarlisle Sounds good
@DavidCarlisle For luatex.def?
 
David Carlisle
David Carlisle
11:47 AM
@JosephWright no I get from 
\documentclass{article}
\begin{document}
x
\end{document}
terminal output of 
No file digit.aux.

warning  (pdf backend): forcing decimal digits to 4
[1{/usr/local/texlive/2015/texmf-var/fonts/map/pdftex/updmap/pdftex.map}]
(./digit.aux))
(see the transcript file for additional information)
 
Joseph Wright
Joseph Wright
@DavidCarlisle Ah
@DavidCarlisle Same for 
x
\bye
@DavidCarlisle One for Hans
 
David Carlisle
David Carlisle
I tried changing your dat file where it says it wont need changing 
% Low-level settings unlikely every to need to change
compresslevel    = 9
decimaldigits    = 4
pkresolution     = 600
horigin          = 1 true in
vorigin          = 1 true in
as I thought maybe if it started at 4 it wouldn't warn, but still get same
@JosephWright unrelated typo in the above I just saw ^^ every->ever
 
Joseph Wright
Joseph Wright
@DavidCarlisle Ah 
\pdfvariable decimaldigits=4 %
x
\bye
@DavidCarlisle Did you rebuild the format?
 
David Carlisle
David Carlisle
@JosephWright I thought so, at least fmtutil witted on a bit, I'll check...
 
Joseph Wright
Joseph Wright
@DavidCarlisle If you use 4 or more in my demo all is well, if you use 3 you get the warning
@DavidCarlisle I guess suggest a change to this setting? Makes PDFs a bit bigger but would be consistent and avoid a warning
 
David Carlisle
David Carlisle
11:53 AM
@JosephWright I think I saw somewhere in the manual or sources about using 4 digits
 
Joseph Wright
Joseph Wright
@DavidCarlisle Fixed
 
David Carlisle
David Carlisle
@JosephWright rebuilding formats again..
 
Joseph Wright
Joseph Wright
@DavidCarlisle Works here
 
David Carlisle
David Carlisle
@JosephWright and here, maybe I rebuild before editing (I've done both a few times:-)
@JosephWright manual of course says you can use pdf.setdecimaldigits without saying that if you set it less than 4 it complains and sets it back:-) Should we flag that as a change on texlive list? (will affect pdftex mostly I suppose)
 
Joseph Wright
Joseph Wright
@DavidCarlisle Yes
 
David Carlisle
David Carlisle
12:03 PM
@JosephWright or we could ask luatex to allow 3 dp as before unless luatex has specific reason for needing 4 (which seems odd)
 
Joseph Wright
Joseph Wright
@DavidCarlisle Certainly can ask
 
David Carlisle
David Carlisle
@JosephWright I'll send mail....
 
yo'
yo'
12:14 PM
@cfr Sorry if I wasn't clear about the \theta, that's somehow what I meant. A font need not include both variants, and if it does not, you can't provide them, so it's clearly not a bug. It's still a feature request, however, an unsatisfiable one.
 
cfr
cfr
12:47 PM
@yo' Oh, yes. I see. I thought you meant a feature request of the kind which might reasonably be satisfied. Really, it would need to be a feature request to the developer/maintainer of the font, I think, rather than the packager(s). At least I don't have anything to do with creating fonts or adding glyphs or whatever. I only know about packaging them, converting them and so on.
 
yo'
yo'
@cfr I got it. Sorry once more for causing confusion.
 
cfr
cfr
1:22 PM
Do the hyperlinks in interface3.pdf's index work for other people? For me, they all point to page 1, but I have no issue with hyperlinks in other PDF documents (including TeX documentation like the PGF manual).
@yo' No problem. I'm just beginning to wish I'd never created that package because I keep getting people saying there are bugs in it when the behaviour is a result of existence failures on the part of font glyphs. However, I think basically the font was never really intended for maths and so maybe the bug report should really be that the package exists at all ;).
@yo' I guess you might think the package should use upright Greek but that does not seem to be the norm.... (And it would be a PITA to implement.)
 
yo'
yo'
@cfr I don't think anything :)
 
Faheem Mitha
Faheem Mitha
@DavidCarlisle I'm well known for being slow on the uptake. But I usually get there in the end.
Hello, people.
 
cfr
cfr
1:46 PM
@yo' Sounds wonderful ;).
 
Torbjørn T.
Torbjørn T.
@cfr Seems to work fine here.
 
yo'
yo'
@cfr :)
 
 
2 hours later…
David Carlisle
David Carlisle
3:33 PM
@FaheemMitha :-)
0
Q: Why "p" in pmatrix?

ColasI am using the command \begin{pmatrix} [...] \end{pmatrix}. I wonder why this command is not just called \begin{matrix} [...] \end{matrix} and why a "p" before "matrix".

matrices
@PauloCereda answer: the p is silent as in pshrimp ? ^^^
 
cfr
cfr
@TorbjørnT. Thanks. That's very odd.
Does anybody have any idea why all the hyperlinks in interface3.pdf point to page 1 for me but not, apparently, for other people and not for other documentation on my machine? To be clear: the hyperlinks from the ToC work fine. Only the hyperlinks from the index fail.
 
David Carlisle
David Carlisle
@cfr welcome to my world:-)
 
Manuel
Manuel
3:48 PM
Hello chat. I have a suggestion / doubt / (I have no idea what it is) / desire, etc.
One for l3keys for expl3 coders :) 
\keys_define:nn { foo }
 {
  bar .code:n = { [bar={#1}]} ,
  unknown .code:n = {[unknown={\l_keys_key_tl}]}
 }
\keys_set:nn { foo }
 {
  bar = It ~ \textit{prints} ~ ok. ,
  It ~ \textit{doesn't} ~ print ~ ok.
 }
The problem there is that the version of unknown get's \detokenized. Then check if exists, which is not the case, and then we cannot recover the original “code”. Any arguments pro this or against this behaviour?
My solution is to say \tl_if_in:nnTF {#1} { = } { \keys_set:nn { foo } {#1} } { [unknown={#1}] }. Any solution more “straightforward”?
Basically I'm designing a more robust theorem approach, and I would like \begin{theorem}[Whatever] and \begin{theorem}[label=thm:whatever,name=Whatever] and even \begin{theorem}[label=thm:whatever, Whatever]to work as expected.
 
David Carlisle
David Carlisle
@Manuel Joseph's code but it seems suspect that you expect typesetting code to work as a key name (even an undefined one) apart from internal code considerations (you have at some point to make an internal csname corresponding to the key) it would just seem a very surprising interface to me to mix keys and document text in that way.
 
daleif
daleif
4:24 PM
@cfr FOr me the hyperlinks in the index also all points to page i
 
David Carlisle
David Carlisle
@daleif @cfr @daleif I think that came up before I'll try blaming @JosephWright who may remember something
 
Joseph Wright
Joseph Wright
5:00 PM
@DavidCarlisle Yes, this is Will's job
 
David Carlisle
David Carlisle
5:10 PM
@JosephWright going in circles on luatex :-)
 
Joseph Wright
Joseph Wright
@Manuel Hmm, whilst part of the reason for handling unknown keys is to allow some 'fall back keyless' operations, the use case I had in mind was colours, as seen in pgf, where you take something without a key at all as a lookup.
@Manuel I'll think about the semantics here
@DavidCarlisle Indeed: I was thinking of e.g. [color = blue] being the same as [blue] or [c] being the same as [alignment = center] (the latter say in some extended version of \parbox)
@DavidCarlisle Yes
@DavidCarlisle From what Hans says, it might be best to just go for 4 d.p.: makes PDFs bigger but really should be OK beyond that
 
David Carlisle
David Carlisle
@JosephWright sure although since there's more or less no calculation in the pdf hard to believe anything really needs to be positioned to 4dp accuracy but can't harm, but in that case the luatex behaviour is weird and undocumented in allowing lua to be set to 2 :-)
 
Joseph Wright
Joseph Wright
@DavidCarlisle Indeed: I guess 3 was chosen for a reason
 
David Carlisle
David Carlisle
@JosephWright and why does context have it as 6, it would be interesting to see if they have a test file that shows the need.
 
Joseph Wright
Joseph Wright
@DavidCarlisle Do you think that likely?
 
David Carlisle
David Carlisle
5:18 PM
@JosephWright I decided not to speculate:-)
 
cfr
cfr
@JosephWright Any chance somebody might mention the problem to the relevant Will?
 
Joseph Wright
Joseph Wright
@cfr Oh, they have, but I will do again
 
cfr
cfr
@JosephWright Thanks. If I could make myself remember the bug when I'm in the index, it wouldn't be so bad ;).
 
Manuel
Manuel
5:31 PM
@JosephWright I don't really know how everything is handled, but, in any case, I think it wouldn't be a lot of problem to have, apart from \l_keys_key_str and \l_keys_value_tl, may be \l_keys_raw_key_tl or \l_keys_whole_keyvalue_tl (this last is a bit of a stretch).
If not this way, what's the correct way of handling everything? Now just thinking some \keys_set_known:nnN and then use the resulting _tl.
 
egreg
egreg
5:48 PM
Nice typo in the Changelog for TeXStudio: - fix: treat minted as verboten environment
 
 
1 hour later…
Alenanno
Alenanno
7:03 PM
Hi all
 
Joseph Wright
Joseph Wright
@Alenanno Yo!
 
Alenanno
Alenanno
@JosephWright hey!
@JosephWright I think this question is a duplicate of this Text is english, citation style is in another language, since apparently Turkish is not supported. I tested with other languages and it works.
I was going to answer, but the answer would be the same as Mico's.
@JosephWright What do you think?
Hello @StefanKottwitz !
I wonder if the tag works if I add the exclamation mark...
I guess not. :D
Hello @DavidCarlisle
 
Torbjørn T.
Torbjørn T.
7:23 PM
@Alenanno! Did it?
 
Alenanno
Alenanno
@TorbjørnT. Ah, it worked.
 
Torbjørn T.
Torbjørn T.
@Alenanno Punctuation in general shouldn't affect pings I think.
 
Alenanno
Alenanno
@TorbjørnT. Good to know. :D Do you think the questions I linked above are dupes? They feel like that to me.
 
Joseph Wright
Joseph Wright
@Alenanno Hmm, probably yes
 
Alenanno
Alenanno
@JosephWright The solution is literally the same. Unless there is a natbib/biblatex style that uses Turkish (didn't find it).
I'm off to dinner. See you guys later!
 
David Carlisle
David Carlisle
7:27 PM
@Alenanno hi
 
Henri Menke
Henri Menke
7:39 PM
TrumpScript: Making Python great again
 
 
4 hours later…
David Carlisle
David Carlisle
11:57 PM
@JosephWright arrrrrrrrrrrrrrrrrrrrrrrrrg
 
egreg
egreg
@DavidCarlisle This seems to have something to do with LuaTeX. It's the number of r's, maybe.
 

« first day (1906 days earlier)      last day (3028 days later) »