8:00 AM
Is it just me, or is scrlayer-notecolumn hideously fragile and worse than just using plain \marginpar or any other work-arounds it was supposed to solve?
It is labelled as experimental, but I had it recommended to me a few times on the site, and saw in other questions it looked like people were doing cool stuff with it.
But it seems like I have to fight it every step of the way.
I guess it does have some advantage over \marginpar in that it won't just allow content to flow over the page. my my goodness the level of finicky it is to do anything advanced with it...
ahh it feels good to get that off my chest

@LyndonWhite I suspect most of us stick to \marginpar :)

8:15 AM
What I really want is a margin paragraph that is willing to float upwards to avoid going off the page bottom.
But I think nothing in latex will ever float upwards?

@LyndonWhite it's a bit tricky to float up if they try to stay with the main text paragraph, easier if you just let marginpars filll from the top.
@LyndonWhite the hard part actually is not so much coding it up as coming up with a spec of what you want to do (in all cases:-)
@LyndonWhite er \begin{figure}[t] floats upwards....
@JosephWright most people use \usepackage{a4wide} and have no margins. No I couldn't have said that.

@DavidCarlisle :)

thinking about really what I want:
each margin paragraph given a unhappyness rating (i.e. a loss function).
Which is distance from where it is declared, with an additional factor for how far up, and then an additional factor for each time it is broken across a page.
Then optimize to minimize the unhappyness of all the margin paragraphs.
But that is NP-hard
because of having a constant loss factor for an optional occurrence -- breaking over a page.
Which I'm pretty sure lets you convert it into Mixed Integer Programming.

Hola' to everybody and good morning
@egreg Hi Prof. I would like to ask you a question about compiling a .te file. Specifically, I would like to buy a € 1000 laptop with a 7th generation Intel processor with at least 4GB of RAM and 8GB of motherboard RAM. Is compiling a .tex file, with png or pdf or graphics in TikZ, faster or slower?

8:32 AM
@Sebastiano Sorry, but I'm no hardware expert. But the more RAM you have, the better.

@Sebastiano As @egreg says. When buying computers, I have never bothered to pay for the fastest CPU, which invariably costs a bunch of extra money for a very marginal improvement – especially as seen over the lifetime of the computer: After just a year, that superfast CPU will be just a run of the mill CPU anyhow. But I always max out the RAM. Disk too, as far as I can afford to.

@LyndonWhite NP hard doesn't actually matter. A a sane document doesn't have more than say 10 marginpars on a page, and you want to keep them in order, so the numbers can't get that big.

@Sebastiano Oh, and having a fast disk may be more meaningful than a big one. So go for SSD if available and you can afford it. It will have a big impact on how fast the computer feels in practice. (I only need a big disk because of my photography hobby.)

Fair.
They don't really have to stay in order, e.g. it is preferable to reorder marginnotes than to have one split across page boundries.
But I can't see anytime that reordering would get you something that just pushing the notes up or down to avoid breaking would not get you.

8:50 AM
@DavidCarlisle ooh. :)

9:01 AM
@DavidCarlisle ah I see some discussion of this kinda thing at tex.stackexchange.com/questions/225795/…

Hi :-)

Aloha!

I have a question for which I'm not sure how to phrase it: When I add \printglossaries to my document, it changes something so that the sections look different. Before, if I have multiple sections below each other (nothing in between) they are close together. After adding \printglossaries there is a lot of space.

Any idea where I have to look for the problem?

9:16 AM
@LyndonWhite if you penalise how far a margin note is from its target paragraph then potentially you could reduce the penalty if you kept all the marginnotes in position except moved the last one up to the top where there was space, rather than shift all the notes up to make space for the one at the bottom (depends on your cost function)

@PauloCereda Aloha! Romanza from Mozart's n. 20 playing just now. ;-)

@MartinThoma if you made a test file and posted it as a question on the main site someone could easily answer just by tracing the code, otherwise you rely on someone here having seen that before

I'm currently trying to make a minimal example. The document is quite big and that takes quite a while, but I think I found it: It was not related to the glossary (I'm not sure if that was anther problem or what happened there). A big minipage after the sections seems to cause that behaviour.

@egreg ooh that's great! :)

@MartinThoma space around sections is (typically) stretch so it will stretch or shrink depending what else is around so the space changing is just business as usual, probably.
Ohh it's a birthday

9:23 AM
@DavidCarlisle After removing \usepackage{parskip} the problem seems to be gone.

9:46 AM
@egreg \ensuremath :-)

@DavidCarlisle Twice!

@egreg even better

@DavidCarlisle In Rome they would say “quanno ce vo' ce vo'” (I don't think you can Google translate it).

@egreg On the same lines, Carla mentioned that I used Arrivederci Roma as soundtrack. Oopsie. :)

@DavidCarlisle Perhaps, “when it takes, it takes”.

9:55 AM
@egreg I shouldn't make google work on its birthday (Non dovrei fare il lavoro di google per il suo compleanno)

@DavidCarlisle Use Bing. :D

@PauloCereda Conheço todos esses idiomas sem ajuda automática

@DavidCarlisle good point. Do you still think it is not problematically NP-hard?

@LyndonWhite presumably not in reply to my portuguese? :-) Oh it probably is, but as I say that isn't necessarily an issue as the numbers are not that big. See Frank's alice code on optimising figure placement across a whole document, that must be harder, it slows things down but not impossibly so

@LyndonWhite asymptotically NP hard, but as David mentioned, numbers are kept at a doable range. :)
@DavidCarlisle ooh :D

10:03 AM
@DavidCarlisle which code? (nothing is coming up from google on that)

10:30 AM
@LyndonWhite yep sorry (I was going to give a link but got called from my desk, that's the thing about day jobs:-)
@PauloCereda you wait: there will be a biologist using A0 paper with 1000 marginal notes and 10000 footnotes,

@DavidCarlisle ooh or a linguist telling us the eskimos have 40 words for snow.

@PauloCereda oddly enough over the years I have had more "feature" reports about footnotes from biologists than anyone else:-)

@DavidCarlisle uh-oh. :) Do we have any biologists in the house? :)

1 hour later…
11:47 AM
High Sierra anyone?

12:02 PM
@DavidCarlisle By the way, in another answer I removed \ensuremath from a definition.

@egreg trying to atone for earlier sins?

@DavidCarlisle No, the removal was earlier.

@egreg that's worse then: you are falling in to sin:-)

1 hour later…
1:31 PM
@DavidCarlisle Re. your comment on tex.meta.stackexchange.com/questions/7420/…: I'm not sure what you mean about LuaTeX. Yes, the code is today different from that in TeX90 but it's been created by modification not by reimplementation from output only.

2. Unusually short abstracts, intentionally abridged in a rare display of academic humour. (12/46)
@AlanMunn, @JosephWright ^^ :)
The fact of the first author being JK (an internet slang for joking) makes everything more amusing.

@JosephWright originally created by modification, true but to a different language and then parts of it extensively hand re-coded since then (other parts less so) but I chose the wording (hopefully) to be true without going into details it is a different, and separately maintained code base in its present form.

I should blame TeX and friends for making me extremely paranoid on kerning, alignment and whatnot. I am beginning to think that's one strong reason my thesis is taking so long... :)Paulo Cereda 16 secs ago

@PauloCereda I have seen a PhD thesis in algebra with that very same abstract.

@HaraldHanche-Olsen Really? That's amusing!

1:43 PM
@HaraldHanche-Olsen that abstract would be longer than @PauloCereda's entire thesis!

@DavidCarlisle oi

@PauloCereda Knowledge is a dangerous thing. I recall getting so confused by “its” vs “it’s” I had to relearn the difference. With the result that, every time I see the wrong one now, I nearly jump out of my skin.
@DavidCarlisle Well, I have no doubt the thesis is full of clever TeX macros. Whether anything is actually typeset, is a different question. Lorem ipsum to the rescue …

@HaraldHanche-Olsen uh-oh. I suffer with kerning syndrome.
@HaraldHanche-Olsen or kantlipsum. :)

@PauloCereda Wait, wait … is that a thing? Hang on while I texdocit … ooh, @egreg to the rescue! Who would have thought?

@HaraldHanche-Olsen Surely not Kant. :)

1:53 PM
@DavidCarlisle At least, an empty thesis contains no errors.

@HaraldHanche-Olsen ^^ I wrote a paper on generation of chant melodies (unpublished still) and decided to be fancy about drop capitals. :)

@PauloCereda Combine it with kantlipsum, and you've got a Kant chant!

@HaraldHanche-Olsen ooh I like this

@PauloCereda (Faint echo of Dr. Seuss there.)

2:08 PM
@HaraldHanche-Olsen Nice package, isn't it? I was recently in a bookshop and saw a copy of Kant's “Critique of judgment”: it's written exactly like kantlipsum. :-)

Philosophy (and it must not be supposed that this is
true) is a representation of the never-ending regress in the series of
empirical conditions; however, the Antinomies have nothing to do with,
in the study of philosophy, the discipline of practical reason.
Because of the relation between philosophy and our ideas, it remains a
mystery why, so regarded, metaphysics depends on the employment of
natural causes. The pure employment of the Antinomies, in particular,
is a body of demonstrated science, and all of it must be known a

@PauloCereda @JosephWright I'm not sure if I've posted this before: realclearscience.com/blog/2015/12/…
2
@DavidCarlisle I Kant even.
@PauloCereda You're just listening to the wrong people. Some of us are not obsessed with this. We use TeX for its other useful properties.

2:26 PM
@AlanMunn :)
@AlanMunn I was exagerating, but I became aware of more things since I started using TeX. :)
@AlanMunn Fantastic

2:54 PM
@AlanMunn That was a funny one, thanks for sharing! :)

@PauloCereda Pushed the first try for random ducks to github, see page 11 github.com/samcarter8/tikzducks/blob/master/documentation.pdf

@samcarter ooooh
@samcarter: I WUB YOU

@PauloCereda <3
@PauloCereda Next step is to do something with the colours.

@samcarter <3
@samcarter rainbow duck
Installing High Sierra now, let's see. :)

3:08 PM
@samcarter close enough. :)

@PauloCereda google never disappoints in such things :)

@samcarter nope. :)
@samcarter: I will use the random ducks in my info boxes for arara 4.0

@PauloCereda Yeah! I'm really looking forward to 4.0 - the possibility to use conditionals is so great!

3:28 PM
@samcarter me too. :)
@DavidCarlisle: do you know this project from Apache? commons.apache.org/proper/commons-jelly/index.html

@samcarter @PauloCereda It seems to me that arara is becoming a new Emacs: nice and sexy operating system :)

@PauloCereda we used to have a jelly based website running at w3c for the mathml test suite

@yo' ooh :) That would be arTeXmis, my next project. :)

I guess it still is

@DavidCarlisle interesting

3:34 PM
@PauloCereda I always found it frustrating, just like vi suffers from not being emacs, jelly suffers from not being xslt

@DavidCarlisle you are mean :)

Jun 29 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)

@DavidCarlisle oh

@PauloCereda somehow the air of surprise seems faked

@DavidCarlisle oh no

3:43 PM
@yo' just easier to exit :)
4

@samcarter ooh

@Skillmon Please see image above :) Your package is really fantastic! Thank you for putting it on ctan!

4:10 PM
High Sierra still installing.

@PauloCereda Exciting! Like watching paint dry, except you don't know if it's gonna blow up in your face.

@HaraldHanche-Olsen Very accurate. :)

@samcarter Are there safety glasses for tikzducks? @PauloCereda could use a pair.

@samcarter @Skillmon Which is the name of your package?

4:26 PM
@HaraldHanche-Olsen Unfortunately not, only normal ones and sunglasses.

@HaraldHanche-Olsen ooh safety goggles

4:59 PM
@egreg, @AlanMunn, @JosephWright, @HaraldHanche-Olsen ˆˆ :)

@samcarter Thank you, it's great!

@PauloCereda Ooh, on a six year old iMac even!

@PauloCereda Oooh

5:16 PM
@HaraldHanche-Olsen my iMac. :)

5:29 PM
@PauloCereda Striking differences, I guess

@egreg Nothing noticed so far.

5:45 PM
@PauloCereda Much more responsive, surely

6:23 PM
@egreg Probably less responsive initially, while Spotlight does its indexing.

6:36 PM
@egreg do you know anything about space tokens?

6:54 PM
@JosephWright @egreg's hiding so which bit of the TeXBook tells me that the first of these is like the second, not the third....
X\vbox{\halign{ #\hfil\cr1\cr2\cr}}X
X\vbox{\halign{#\hfil\cr1\cr2\cr}}X
X\vbox{\halign{\space#\hfil\cr1\cr2\cr}}X

\bye
tex.web makes it explicit @ Spaces are eliminated from the beginning of a template.

7:21 PM
@samcarter Thanks. I'm not so sure about that (the code might be really bad). But I just had to put it together seeing that one ASCII-duck when it was posted here :)

Do I need to use a tikz library in order to set text within an ellipse shape?

@DavidCarlisle :)

@Werner You need \usetikzlibrary{shapes}

Here's an example:
\documentclass{article}

\usepackage{tikz}

\begin{document}

\begin{tikzpicture}
\draw (2,2)
node[shape = ellipse, line width = 3pt, draw=blue]
{Something};
\end{tikzpicture}

\end{document}
The output is a rectangle.
@Skillmon ...looking into it.

@Werner you can leave the rest as is. Just that library and it works. Unknown shapes are replaced by rectangle, that's why your output is a rectangle.

7:34 PM
@Skillmon: Well wouldn't you know it! shapes adds shapes to tikz.

@Werner :)
@Werner Sometimes they make things really cryptic :)

@Skillmon :)
Thanks.

@DavidCarlisle Engaged with a cricket match

8:11 PM
Goal

8:59 PM
@egreg I don't think you have the rules of cricket quite right

9:16 PM
@DavidCarlisle Not sure if it's in the TeXbook, but the behavior is consistent with spaces being ignored after &
@DavidCarlisle Don't they score goals in cricket? Then I've been deceived.

@egreg funnily enough when I looked in tex.web for \halign details, first thing I found was ...
default case that applies if no matching label is found. Thus, we shall use
constructions like
\vbox{\halign{\ignorespaces#\hfil\cr
@egreg looked like "somebody" had forgotten this feature:-)

@DavidCarlisle Very possible; maybe ignoring space at the beginning was added on second thought.

@egreg I started on this search of course as you had omitted % from ends of lines (even though tex ignores the space, think of that lost byte of tex memory:-)

@DavidCarlisle Did I? Oh, well…

@egreg the xetex macron under accent code:-)

9:22 PM
@DavidCarlisle Anyway I'm looking in errorlog.tex and have found no entry for it

Oh no it's start of the template not start of the inserted cell text.

@DavidCarlisle Yes, spaces are ignored only after &, but not after the u-part.
@DavidCarlisle Spaces are ignored also during the scan for \omit in the first cell of a row.

9:45 PM
Hi, is there some users of reftex for citation right now?
I forgot a shortcut and I don't find anything in the manual
In reftex, you can use C-c ] to insert a \cite command. There is another shortcut which add something before C-c ] to use when you want to cite but also to add a page or else (\cite[45][key]).
Do someone know this one?

@ppr I don't have it installed but C-h b will show you all the key bindings in the current mode

@DavidCarlisle I've got it: C-u! C-u C-c c which prompt for optional argument

@ppr oh that's the generic prefix keybinding across emacs

@DavidCarlisle ok I didn't know. It prompt for all kind of optional argument?

10:01 PM
@ppr yes eg C-f = move forward one character C-u 20 C-f = move 20 characters