« first day (1794 days earlier)      last day (3434 days later) » 

10:01
Oh my, I leave the chatroom for one day and suddenly @egreg and @DavidCarlisle are cooperating in fixing 2e! Wow! :)
10:30
@PauloCereda Not real cooperation.
@egreg ah David broke LaTeX again. :)
And @Joseph was blamed during the process. :)
@PauloCereda My fault :-)
@PauloCereda You wait until Thursday!
@JosephWright oh no
@PauloCereda No! @egreg was wrong and thought it was a bug, I corrected his error
@DavidCarlisle ah :)
10:39
@PauloCereda No @JosephWright broke it, I got a badge for pointing out a fix.
@DavidCarlisle :-)
@DavidCarlisle ooh
@DavidCarlisle Next job I think is to get fontspec added to the CI system :-)
@JosephWright I noticed that tlmgr installed a new luajittex today. Have you ever tried that?
@DavidCarlisle Not to date
@DavidCarlisle Apparently no LaTeX format is pre-built
@DavidCarlisle Format built, testing now
@DavidCarlisle Probably another one to special-case in l3build so it looks at the normal .luatex.tlg files
10:52
@JosephWright good plan
@DavidCarlisle Looks like it passes the tests other than (as expected at present) not using the LuaTeX-specific code, etc.
@DavidCarlisle At present there's not luatexjitlatex.fmt by default, so the set of people who might use this is pretty small :-)
 Check failed with difference files
 - ../build/test/d3dvipdfmx.luajittex.diff
 - ../build/test/d3dvips.luajittex.diff
 - ../build/test/d3pdfmode.luajittex.diff
 - ../build/test/d3xdvipdfmx.luajittex.diff
 - ../build/test/m3box001.luajittex.diff
 - ../build/test/m3box002.luajittex.diff
 - ../build/test/m3box003.luajittex.diff
 - ../build/test/m3box004.luajittex.diff
 - ../build/test/m3char001.luajittex.diff
 - ../build/test/m3coffins001.luajittex.diff
 - ../build/test/m3expl001.luajittex.diff
 - ../build/test/m3expl003.luajittex.diff
@JosephWright yes presumably at some point they'll decide that it's stable enough and quicker and make it the default, hard to know when to use it if both formats are bult, good to have ability to test it occasionally though.
@DavidCarlisle Yes
@DavidCarlisle I'll work out what we need
11:16
@DavidCarlisle Probably assume they share .luatex.tlg files (i.e. no luajitex.tlg cases)
@JosephWright yes if that ever fails, probably time to report a bug to Hans:-)
@DavidCarlisle I have a feeling it will: looks like the 'out by one' error might not apply in quite the same way to the two versions!
@JosephWright good opportunity to raise that again then...
@DavidCarlisle Might need to spend a bit of time later making sure I've got everything set up correctly first: getting the name for the engine correct in all places is not so easy!
11:44
*** ..\build\test\d3dvips.luatex.tlg	2015-09-29 12:43:45.555171300 +0100
--- ../build/test/d3dvips.luajittex.log	2015-09-29 12:43:45.573171300 +0100
***************
*** 23,29 ****
  > \box27=
  \hbox(6.94444+1.0)x86.45851
  .\special{ps:gsave}
! .\special{ps:currentpoint currentpoint translate /savedmatrix matrix currentmat \ETC.}
  .\hbox(6.94444+1.0)x0.0, glue set - 86.45851fil
  ..\hbox(6.94444+1.0)x86.45851
  ...\OT1/cmr/m/n/10 G
--- 23,29 ----
  > \box27=
  \hbox(6.94444+1.0)x86.45851
  .\special{ps:gsave}
@DavidCarlisle ^^^
@JosephWright lovely! You'd think the tex side of things would be the same....
  Check failed with difference files
  - ../build/test/d3dvips.luatex.diff
  - ../build/test/m3msg004.luatex.diff
  - ../build/test/m3show002.luatex.diff
  - ../build/test/m3show003.luatex.diff
  - ../build/test/m3tl013.luatex.diff
  - ../build/test/m3token005.luatex.diff
All due to the line-wrap business
@JosephWright glad I asked about luajit now?? seems like it's creating a lot of work for a format no one uses.
@JosephWright ask on the list? I suppose it's possible that the difference is just due to the jit version being a later build rather than being due to the luajit itself, but I don't really want to start building luatex from dev sources.
@DavidCarlisle Sounds likely, but first let me take a look at the raw log comparison
@DavidCarlisle Ah, tracked it down I think: I needed to add another place where luajittex ~ luatex
12:47
@DavidCarlisle All working now :-)
@JosephWright :-)
@DavidCarlisle Probably don't need to add it to the test set at the moment!
@DavidCarlisle What I do need to worry about is fontspec
@JosephWright yes good if you could make it not load half of ctan at the same time:-)
@DavidCarlisle lm is the big download, but yes I need to work on getting Will to on get it down to expl3, lm and luaotfload
@DavidCarlisle Once the format update goes in we'll remove one dependency
@DavidCarlisle Problem I guess is xunicode/tipa
@JosephWright can't be needed, either of them!
12:55
@DavidCarlisle I meant the fact removal impacts on users
@DavidCarlisle I'll probably put in some pull requests to @WillRobertson soon-ish
@DavidCarlisle For example, Heiko's code is I think now all covered in expl3 itself
@DavidCarlisle Probably in terms of downloads we are not really needing to worry as lm is the biggest and really can't be skipped
@JosephWright probably doesn't matter but is there a usable lm already in the tl200x that comes with the base vm?
meanwhile dynamic memory allocation in matlab mex files is calling....
13:21
I'm currently playing with a font that has no math setup, so I'm attempting to make one, by using a math font that sort of works, and then replacing a-z-A-z0-9()[]{}.,;: and perhaps +-= etc. I'm currently looking at [] and {}, as I know the slot for (), anyone have a good advise where to look up the slot numbers for large symbols? I'm currently using this to replace the ): \DeclareMathDelimiter{)}{\mathclose}{xxx}{)}{largesymbols}{"01}`
@daleif -- sadly, there's no "natural" place for the large symbols except in the pua, if access is by unicode value. (and there never will be.) are these opentype fonts, or some other format?
@daleif You can't really replace the large symbols from your base math fonts with ones from a text font as you need the font metrics to string them together in size sequence and specify when to start building stretchable ones out of repeatable sections
@barbarabeeton weird legacy metric format known as tfm, you may have heard of it
@DavidCarlisle -- uh, yeah. but @daleif still needs to know where to look for the larger symbols or the pieces in the font(s). we know why dek made the extension font a separate thing.
@DavidCarlisle hmm, might get away with just replacing letters and numbers and I already have this.
It is either this, or someone in my organisation need to learn how to typeset math in InDesign. For now we are looking into making the few articles she needs for her report, in LaTeX, and then importing the PDF into indesign.
13:36
@DavidCarlisle -- the text-size parentheses and square brackets for computer modern do reside in the base text font, cmr, unless i'm sorely mistaken.
@barbarabeeton yes but I think replacing just those then using a different font for larger is probably worse than using the other font set for all sizes. Probably not too noticeable with ( but if \{ and \bigl\{ come from non-matching fonts it might be more jarring than if they both come from same font set, different to the set used for letters, perhaps...
13:54
@DavidCarlisle -- the braces for cm do come from a different font (cmsy). i agree that all sizes of each delimiter should come from a single basic design.
@barbarabeeton yes that's what I meant by set cm* but would you want normal size ( from Times and larger sizes from cmex?
@DavidCarlisle -- no; all sizes of each delimiter should come from a single basic design. and getting the delimiter sizing to work properly is a tricky business, as we're discovering from testing the stix fonts. (that still doesn't help @daleif find the larger shapes in whatever fonts he's using.)
yo'
yo'
@DavidCarlisle It is noticable with () very well for some fonts! Much more than with {}! For instance compare CM and Euler. I ended up defining something like \def\lrg#1{\text{$\vcenter{\hbox{\larger$#1$}}$}} (I don't remember exactly) to get a scaled parenthesis by \mathopen{\lrg(} so that the one from CM is not used.
@barbarabeeton, we'll later on when I get some of the data to be used.
14:15
@JosephWright / @DavidCarlisle — I think you guys might be getting confused! fontspec basically has no dependencies besides euenc and xunicode. euenc is trivial and xunicode is a can of worms. I have no idea what to do about it; I think it needs a massive re-think about how we're supporting encodings in L3. I'm considering forking xunicode.sty and stripping out everything that's not vanilla LaTeX &c and see what's left.
@JosephWright / @DavidCarlisle — unicode-math has a few more dependencies but as of the last update no longer loads any of Heiko's packages. The biggest dependency there not covered by expl3 would be Martin Scharrer's filehook package.
@WillRobertson: Happy wedding!
14:50
It is so quiet today.
People are hiding!
Or David will blame them! :)
@WillRobertson vvv
\documentclass{article}
\usepackage{fontspec}
\begin{document}
\stop
produces
(/home/davidc/texmf/tex/latex/base/fontenc.sty
(/usr/local/texlive/2015/texmf-dist/tex/latex/euenc/eu2enc.def)
(/usr/local/texlive/2015/texmf-dist/tex/latex/euenc/eu2lmr.fd)(load luc: /usr/lo
cal/texlive/2015/texmf-var/luatex-cache/generic/fonts/otf/lmroman10-regular.luc)
) (/usr/local/texlive/2015/texmf-dist/tex/xelatex/xunicode/xunicode.sty
(/usr/local/texlive/2015/texmf-dist/tex/latex/tipa/t3enc.def(load luc: /usr/loca
l/texlive/2015/texmf-var/luatex-cache/generic/fonts/otf/lmromanslant10-regular.l
@WillRobertson which looks like an awful lot of dependencies somewhere:-)
@WillRobertson I suppose mostly via xunicode
15:18
@PauloCereda: Bug fixing again assoccnt :-(
@ChristianHupfer just assign any bad behaviour to "user error"
@DavidCarlisle Yes, the DavidCarlislean Way -- the user is to blame ;-)
@ChristianHupfer it's correct more often than not!
@DavidCarlisle: :D Yes, but I am still looking why my package and calc do not work together. I am sure I've done some stupid thing. I will probably have to rewrite from scratch and using expl3 then
Can anybody take over this question: I am no expert in floatrow/caption/subcaption stuff -- the OP changed the question somehow ;-)
0
Q: Using label in multicols

ThePredatorTo produce 4 figures (2 on top followed by 2 at the bottom) in a two-column article, I am currently using: \usepackage{graphicx} \usepackage{floatrow} \floatsetup[table]{capposition=top} %caption on top of table \begin{figure*} \begin{multicols}{2} \label{fig:image1} \includegraphics[w...

15:57
@WillRobertson Most are from xunicode I think
16:30
@DavidCarlisle I figured out my error on the font finding code you pointed me too. I was accidentally running TL 2014 instead of 2015.
16:43
@PauloCereda Caches do work on Travis :-)
@JosephWright Holy Tantau!
yo'
yo'
17:11
I'll be teaching 1st-year students tomorrow for the first time after a long while. Should I be a good cop or a bad cop? :)
@yo' I'll start on Thursday. Definitely good cop, becoming a bad one at the final test. ;-)
yo'
yo'
@egreg that's naughty! :D
@yo' My last test was a special offer. But I can't stand if somebody tries to cheat.
yo'
yo'
I think I'll be a good cop. I've always preferred a relaxed air during the teaching, and I won't change this. I'll probably ask them to "tutoyer" to me.
@yo' What course?
yo'
yo'
17:17
@egreg Calculus 1
I would have loved to teach Discrete Mathematics, but it wasn't available unfortunately.
@yo' \int\arctan (x+\sqrt{x})\cos x\,dx: just an easy one. ;-)
yo'
yo'
@egreg that's Calculus 2 here I think :-)
Calculus 1 is logics, functions, sequences, limits, derivatives, analysis of functions
@yo' I can think to some weird limit.
yo'
yo'
@egreg We'll surely have some in the teachers instructions, too :-)
@yo' I agree with Lang that the right order is "derivatives, limits, analysis of functions, sequences”.
yo'
yo'
17:25
@egreg ah, well, no. Here it's the other way around. I think it's really irrelevant. Each has some good points and some bad points.
35 minutes until the simulation is finished. This means I might be at home before 9pm. Yay! :)
18:02
@JosephWright My TeX.SE blog post got 25 views today :-)
@JosephWright Some hundreds per day on LaTeX-Community.org though. I thought, TeX.SE readers may miss the occasion of getting LaTeX books cheeper if they simply don't see it. No mention of the blog post in the right sidebar (where featured and hot meta posts are)
@JosephWright An idea, where TeX.SE book authors and users (who get discounts) may benefit: a meta post, where current promotions are listed as answer, and deleted after finished. Can be made featured when one promotion is active, so a bit visible.
yo'
yo'
@StefanKottwitz It's a mistake of the engine IMHO that the blog post isn't promoted in the CB
@JosephWright Like when I had my book completely for free for 1 day, but many people missed it. Now my books are offered for 50% and 35% less, for limited time, I 'd like people to save it while supporting LaTeX books at the same time
yo'
yo'
@Stefan I think you could make a meta post for each offer, along with the blog post.
@yo' One for each offer is also a good idea! I first thought of a general one, but it's right. Can also be removed once it is over (or closed, for historic reasons)
For the hive brain to work on:
0
Q: Enforcing or encouraging use of the most up-to-date version of a class

Alan MunnBackground I maintain a thesis class for my university. The class changes more frequently that I would like based on the whims of the university thesis office who clearly have little else to occupy their time. The net result of this is that the most common support request I receive comes from p...

18:10
@yo' I just addressed Joseph as moderator, so we use the featured tag in a way we agree
and I don't want to use it just for myself :-) but for any book a) our members publish and b) our users can have discounts
@yo' I will wait for another day to see if it happens that the blog post may appear :-) otherwise I assume they are pushed there by hand by a community manager or so, who check all new blog posts, and one did not desire it for that
I add some more examples to the website before I meta post :-)
yo'
yo'
@StefanKottwitz maybe ask in the mod chat whether this should work?
@yo' yes, good thought, at some time, maybe when I notice at the next blog post again
yo'
yo'
@StefanKottwitz ok :)
damn, 230 kB of TikZ code!
@yo' automatically converted a small Inkscape diagram?
yo'
yo'
@StefanKottwitz no, some tiling generated in sage :)
@StefanKottwitz ^^ not this one, but similar to it.
18:23
@yo' Save some space by converting binary numbers to hex :-)
yo'
yo'
@StefanKottwitz oh this is not binary! (I mean, it is, in the sense of being over the alphabet \{0,1\}, but not in the sense of being base-2)
moreover, there are negative digits (denoted by an overline)
sure, but binary encoded :-)
yo'
yo'
@StefanKottwitz ternary-encoded
@yo' interesting shift of the minus sign :-)
yo'
yo'
@StefanKottwitz quite common in my community. You have negative digits everywhere. In the computer, I represent them as letters, so my digits are \{..., c, b, a, 0, 1, 2, 3, ....\}.
18:26
@yo' Interesting! It's the same on certain cruise ships, depending on the yard. a, b, c are lower decks, 1 - 11 upper decks
It destroys the plain sorting in my network things :-o
yo'
yo'
@StefanKottwitz lol
@PauloCereda Why ducks? What's wrong with them? Why is it popping up in everywhere ?
@percusse Bwawawawawawa it has begun! No idea. :)
@PauloCereda Other than the following I don't care
@percusse OH NO
yo'
yo'
18:36
@percusse This is a true maglev duck :D
@yo' Looks like a hiphop song went bad
pimped the wrong thing probably
18:46
LOLWUT
:P
yo'
yo'
19:11
abilities you learn at scouts: Write a 1.5-page project report in 30 minutes.
@PauloCereda Imagine it the other way round.
yo'
yo'
@StefanKottwitz With the variety of dog breeds, isn't there one that looks like a duck? :)
yo'
yo'
@Johannes_B ?
@yo' Seems like someone messed up the treshold for late answers?
19:25
Is there a problem????? :D
yo'
yo'
497 to me :)
Someone wanna some badges? :D
1
Q: What happened to the late answer queue?

wythagorasThe late answer queue just went to 1.2k on Mathematics. On other sites the same happened. Why did this happen? According to a comment on this post, there are no old late answers left over since begin 2013. Did the definition of late answer somehow change?

4
A: Can we raise the bar for reputation for late answers to enter the review queue?

Jon EricsonQueue growth Review queues lose effectiveness if posts aren't regularly (and accurately) cleared. A queue such as close votes on Stack Overflow that never seems to get to 0 fails to provide time-sensitive feedback that aids learning. So if we raise the bar for late answers which increases the nu...

@yo' it's 458 by now. Who is doing the reviews?
@yo' Should i join the review squad?
yo'
yo'
@Johannes_B I am doing some. I skip anything from this year, since I can see some value in doing something about them (though I don't know what ATM). But for things that are years old -- does it make sense to do anything? IMHO no.
@yo' Probably not, no. I'll have a look at the Qs :-)
@Johannes_B I am doing some but I've already the Silver badge there ... and in the end you're limited to 20 reviews per queue and day anyway
19:35
@ChristianHupfer didn't know that the number is limited :-)
yo'
yo'
@ChristianHupfer good. It means that after 6 more I can finally get my ass out of my work :D
yo'
yo'
@Johannes_B I skipped a lot of them.
@yo' The answer there is not answering the question.
@Johannes_B I flagged it (and many others so far ... some aren't answers)
@Johannes_B What's this?
@ChristianHupfer Seems our good Steven has a second unregistered account :-)
@Johannes_B Or got a lot of downvotes :-P
I have 210 late answers in my review queue!
Is this useful?
\DeclareMathOperator\ShiftLeft{\texttt{<<}}% shift left
@egreg I had 503 :D
Ok, 20 reviews done and skipped a lot more
20:01
@ChristianHupfer Time to get a gold badge!
@egreg Gold badge? For what? Late Answer Reviews? That will take a while, unless there's a daily 500 Late Answers queue :D
@ChristianHupfer :)
are you also seeing this?
ok papiro mode on
3
@percusse :-D
20:09
@percusse See half an hour back in chat :-)
party poopers
21:19
@egreg Indeed: what's going on?
@JosephWright I guess some rule changes about what's considered a late answer.
@egreg Yup, just checking in the mod chat now
So if you see a sudden spike in that review queue, it should be temporary.
@JosephWright See also the starred meta questions.

« first day (1794 days earlier)      last day (3434 days later) »