@HaraldHanche-Olsen Yes, noun phrase. Sorry (off at a concert).

@HaraldHanche-Olsen Well there are no 'firm grammatical rules' (until you get to a very deep level of abstraction). There are only individual grammars, with similarities among groups of speakers and social agreement (usually based on the societal power structure) on which groups' grammars are the most highly valued. The rules of these grammars become what you're calling a 'firm grammatical rule'.

@HaraldHanche-Olsen In the case of the "We have that ..." construction, I agree with your supposition that that this is an idiom with quite restricted use that has come out of mathematics and philosophy (as my informal survey of my collection of linguistics PDFs attests). There's no principled reason why it should be ungrammatical for me (or you), it just is.

@AlanMunn tl;dr: it's complicated. (No, just kidding about the tl;dr. Chat is wonderful that way.) And thanks for the feedback.

@barbarabeeton or even being correct

@daleif odd, I didn't think of looking in the manual:-)

@DavidCarlisle on my tablet, looking in the source is a bit difficult,... Plus sometimes the implementation is actually explained in the manual ;-)

Btw not at work today, hence the tablet...

@samcarter OH MY

POOR DUCKY

RIP Hawking!

@samcarter WHAT

@samcarter Oh my! :(

@PauloCereda And of all days on Einsteins Birthday!

@samcarter Indeed!

And Pi day

@PauloCereda can arara change the current directory and switch into a parent or an absolute folder?

@UlrikeFischer In version 3.0, no. In version 4.0, yes. :)

@PauloCereda Hm. Can one disallow it?

@UlrikeFischer It is doable. :)

@UlrikeFischer: let me explain a bit: in v4, paths are all normalized, so we have a canonical reference to them. However, I usually use the relative information by design and leave the other one as "for the last resort".

@UlrikeFischer Also, it's not exactly like that. We can change the current directory on a very specific situation, inside a subshell, so subsequent calls are not done there.

@PauloCereda I asked because of this tex.stackexchange.com/q/420094/2388

@UlrikeFischer Ah understood.

@UlrikeFischer: I will try to write an answer for Carla later on.

@UlrikeFischer an important question! @PauloCereda It should be the other way around, actually, for security reasons

May I ask a question here in chat?

@JohnDorian You can ask, of course

So I have to type this equation: $ \sigma \rho = {}^{\sigma}\!\rho \delta(\sigma, \rho) $. In my view, \rho and \delta look much too close together, it looks much better if I insert a space: $ \sigma \rho = {}^{\sigma}\!\rho \, \delta(\sigma, \rho) $ . Would it be bad style to do so, because it breaks consistency?

(Of course I have a macro for {}^{#1}\!{#2} )

@JohnDorian I don't know what it means but if the prefixed rho is some kind of function, an alternative to adding space by hand would be to use \mathop{{}^{\sigma}\!\rho} so it gets operator spacing (like \log)

@DavidCarlisle Actually \sigma, \rho, {}^{\sigma}\!\rho and \delta(\sigma, \rho) are just four real numbers which are multiplied (well actually they are elements of some group but that doesn't matter here).

@UlrikeFischer, @CarLaTeX: not the best answer so far, but here's some pointers: tex.stackexchange.com/a/420109/3094

@DavidCarlisle And that is also the reason why I want that on the RHS of the equation it becomes more apparent that {}^{\sigma}\!\rho is one thing and \delta(\sigma, \rho) is another thing, that's what the space should be for

Also, if the space is included, the space between \sigma and \rho on the LHS looks about the same as the space between {}^{\sigma}\!\rho and \delta(\sigma, \rho) on the RHS

@JohnDorian in the end TeX is a typesetting system not a computer algebra system. If you need to add space to make things more readable then you need to add space (math journals do that all the time (even if they don't like authors doing it:-)

@DavidCarlisle OK, with your permission I feel comfortable doing so ;-)

ooh vampires

@marmot @PauloCereda @samcarter : What about this?

Moving frame of reference perhaps better suited for looping animation?

\documentclass{article}
\usepackage{animate}
\usepackage{tikzducks}
\usetikzlibrary{shapes.callouts,calc}
\newcommand{\DuckSong}[1]{\ifcase#1
All
\or
my
\or
little
\or
ducklings,
\or
Swimming
\or
in
\or
the
\or
lake,
\or
Heads
\or
dunk
\or
in
\or
the
\or
water,
\or
As
\or
little
\or
tails
\or
do
\or
shake.
\or
All
\or
my
\or
little
\or
ducklings,
\or
Swimming
\or
in
\or
the
\or
lake,
\or
Heads
\or
dunk
\or
in
\or
the
\or
water,
\or
As
\or
little
\or
tails
\or
do
\or
shake.
\fi
}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

@AlexG I think we should hire you for "tikzducks production" ;-). Now it need some sound youtube.com/watch?v=NwcNy1qvxPk

@UlrikeFischer it's all very impressive but it was my duck that got the tick in the "tex mascots" question.

@DavidCarlisle Or the pizza ;-)

@AlexG ooh

tex.stackexchange.com/help/badges/40/tumbleweed this is the saddest place on TeX.SE^^

@AlexG Very nice, but I would still put the wave before the duck (i.e., nearer the viewer). Swimming ducks swim in the water, after all, not above it. The next challenge could be to make the text scroll through the speech bubble. One might actually be able to read it that way.

Next update will address both issues. Currently in a meeting :-)

@AlexG Looks great! But could you perhpaps give the callout node a common text width like text width=1.5cm such that it does not breathe too much?

Happy Pi day!

@PauloCereda oh!

@PauloCereda maisondesvinsdulanguedoc.com/cuvee/3-14-2014.html

@yo' ooh pi wine

@PauloCereda well, check the winemaker's name on the vignette :)

@yo' ooh Python. :)

@PauloCereda only in weird countries that put the month before the day, other countries celebrate that on the third day of the 14th month.

@PauloCereda lol

@DavidCarlisle oh

@egreg Weird... Should I change my question so that it asks about how this behaviour of \tikzcdset arises?

@JohnDorian Maybe complete it with the example. As you see from the picture, the middle part of the stem is too thick and the tail is higher than needed.

@JohnDorian Maybe some TikZ expert can sort out the matter.

Anything bad if I use l3intarray in duckuments? I plan to implement RC4 for XeTeX.

@Skillmon intarray is all internal code

Is there a breakable \thinspace?

@Skillmon At present, we've not made this public: the use cases, etc., are a bit complex

@JosephWright I know, that's why I ask.

@PauloCereda \hspace{.16667em}

@DavidCarlisle Thank you!

@marmot @samcarter @UlrikeFischer @PauloCereda : Doubling the number of frames; changing call-outs every forth frame thanks to some modulo computation:

@JosephWright Anything else I can use? I need a way to easily use an array in which I can swap values.

@Skillmon Normally array data is handled using a property list, though @loopspace has written an array module as part of his hobby code

\documentclass{article}
\usepackage{animate}
\usepackage{tikzducks}
\usetikzlibrary{shapes.callouts,calc}

\newcommand{\DuckSong}[1]{\ifcase#1
All
\or
my
\or
little
\or
ducklings,
\or
Swimming
\or
in
\or
the
\or
lake,
\or
Heads
\or
dunk
\or
in
\or
the
\or
water,
\or
As
\or
little
\or
tails
\or
do
\or
shake.
\fi
}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\usepackage[active,tightpage]{preview}
\makeatletter
\def\@anim@@newframe{\@ifstar\@anim@newframe\@anim@newframe}
\def\@anim@newframe{\end{preview}\begin{preview}}

@marmot @samcarter @UlrikeFischer @PauloCereda : Doubling the number of frames; changing call-outs every forth frame thanks to some modulo computation:

@AlexG Wow! This is getting better and better!

@AlexG However if I watch this for too long I get seasick!

@sam Only refining other's work. Seasick? MeToo.

@marmot: tex.stackexchange.com/a/420151/3094 :)

@PauloCereda Very nice!

@PauloCereda @samcarter Another riddle: what is this : \tikz{\duck*\duck*} ?

@AlexG Really great! You should send this to Martin Scharrer!

@marmot How?

@AlexG bitbucket.org/martin_scharrer/mwe/issues/3/sample-movie-files

@marmot thanks!

@AlexG ooh!

@AlexG -- trivial suggestion ... now that the size of the speech bubble is persistent, i think moving ducky a bit to the right would give a more balanced appearance. (it is delightful!)

@AlexG I AM HYPNOTIZED

@DavidCarlisle: did you enjoy my amazing duck-based counting system?

@PauloCereda no

@barbarabeeton: s/3\.1/4.0/g

@DavidCarlisle oh

@AlexG -- i'm snowbound today, and can't even try to reproduce this. (but i'll be happy to look at a picture.)

Duck goes 2.5D (and a bit to the right):

OH MY

The code:

\documentclass{article}
\usepackage{animate}
\usepackage{tikzducks}
\usetikzlibrary{shapes.callouts,calc}

\newcommand{\DuckSong}[1]{\ifcase#1
All
\or
my
\or
little
\or
ducklings,
\or
Swimming
\or
in
\or
the
\or
lake,
\or
Heads
\or
dunk
\or
in
\or
the
\or
water,
\or
As
\or
little
\or
tails
\or
do
\or
shake.
\fi
}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\usepackage[active,tightpage]{preview}
\makeatletter
\def\@anim@@newframe{\@ifstar\@anim@newframe\@anim@newframe}
\def\@anim@newframe{\end{preview}\begin{preview}}

@DavidCarlisle oh no

@AlexG \begin{advertizing} it may become 2.8D if you use these waves \end{advertizing} ;-)

@PauloCereda @samcarter Answer to the riddle : Starducks (a duck-owned coffee company)

@samcarter sooo good ;-)

@HaraldHanche-Olsen You are right. Some more waves may even be added.

(see above)

@AlexG Oy. Now we're talking!

@AlexG -- absolutely wonderful!!!

@AlexG Maybe a little paper boat could sail in the other direction in the background?

\draw (0,0) -- (2,0) -- (2.4,1) -- (1,0.6) -- (-0.4,1) -- (0,0) (0.6,0.71) -- (1,1.2) -- (1.4,0.71);

@sam I have even thought about a shark-fin passing by

@sam the boat is gorgeous! I will add it.

@AlexG Oh, I like the shark idea!

@AlexG So sad that the academy awards (Oscars) have just been awarded, so you need one year...

Ha, ha. I am quite patient.

By the way, it is rather you, @marmot, who deserves it.

@AlexG Certainly not me. I was just using your code and @samcarter's nice ducks.

@marmot The nice ducks which where inspired by a questions by @CarLaTeX and the fascination for ducks of @PauloCereda and created with the help of so many kind user like @UlrikeFischer and @egreg and many others ... Oh my, the Oscar speech will be quite long

^ consider supporting =)

@EmilioPisanty More worrisome is the fact that we may lose our beautiful site theme altogether. @JosephWright Do you have any inside knowledge of this?

@AlanMunn I'm afraid not

@AlanMunn well, it's a matter of degree, no?

@EmilioPisanty I'm not sure what you mean. We worked hard initially on the design of the site, and it looks very nice. If every site is generically similar, I don't think that's a good thing, and simply keeping site specific buttons won't really help that.

@AlanMunn it's not all or nothing, there's still plenty of theming in the new scheme

cf. the image above

the core things from the tex.se theme that look in danger are the Questions | Tags | Users | ... bar and the buttons

at least from what I can see on a first pass

@EmilioPisanty We are I think in the 'Cadillac' group (relatively long-standing, the chap who did the design was very keen to get it right)

@EmilioPisanty What I don't get at all is how the 'left navigation' is helpful: it uses up part of the window for ... ???

@JosephWright I guess when they say 'Cadillac' I tend to think things like photo.stackexchange.com

?

@EmilioPisanty Perhaps, but there was a lot of effort put in at the end of the beta ...

@JosephWright I don't doubt it

@EmilioPisanty :)

but mostly I don't see that many design elements where the tex.se theme is involved that are not included in the scheme in my previous image

beyond the navigation bar and the buttons

It may well be me not seeing stuff which is indeed there, though, which is why I'm asking

Stack Exchange Inc. got into the phase when it's unable to do the side-step and look at their ideas "from outside". They started taking fragile parts of their products for granted and believing they won't break. The only possible outcome is that they will break.

@yo' how'd you mean exactly?

@EmilioPisanty well, they take for instance the fact that sites have communities for granted, ignoring the needs of the communities. This will make the communities weaker, to the point they won't possibly feel as communities anymore.

@yo' fair enough.

("won't possibly" or "possibly won't"?)

@EmilioPisanty the latter, if you consider them different

On the other hand, you could argue that the change now is about making some fragile parts of their products less fragile.

The current unification effort is only one such step; important features considering the chat were never implemented, with the idea that the chat is a marginal product.

@EmilioPisanty I would say "less troublesome to maintain". Also, remember that the important steps were already implemented, e.g. unification of widths of the elements

@yo' I don't see the difference =P

@EmilioPisanty I do. Maintenance = money. Community = users.

@yo' there's product fragility that doesn't come from community considerations

@EmilioPisanty sure

also, SE grew huge and starts to be unmaintainable and unmanageable; this is quite possibly only a natural evolution of things

Rip Stephen Hawking, one of the persons that caused me studying physics...

@ChristianHupfer oh no

Could somebody correct the spelling of the libertine package in the accepted answer here: tex.stackexchange.com/questions/10941/free-ornaments-font

I can't do it if I don't add at least 6 characters

@JohnDorian That will bump the question to the front page, I think. I am not sure what people think of doing that for such a minor problem.

@HaraldHanche-Olsen Right now it's on the front page anyway^^

@JohnDorian Ah, hadn't noticed that. All right, I edited the answer.

@EmilioPisanty -- still thinking about this one. but the proposal to change to a single font, if implemented, would be disastrous, so i've added tex to the list of sites that need an extra font (monospace) and also added a comment.

@PauloCereda This is life...

@barbarabeeton happy to help

@EmilioPisanty -- thanks!

@ChristianHupfer ... he died on pi day! RIP

@CarLaTeX I know ... :-( And it's the birthday of Albert Einstein as well

@ChristianHupfer Oooh

@ChristianHupfer One week ago I chatted a bit with Kip Thorne, who had an open bet with Hawking...

@marmot what was the bet?

@naphaneal On the features of gravitational waves (of course), essentially if there are some imprints from so-called naked singularities. For details please ask Kip.

well, Stephen now knows, if he won or lost the bet. Kip still has to figure it out

@naphaneal Why should he now know that? As far as I know, he didn't believe in the afterlife.

@Skillmon perhaps he had a moment of clarity just before he died, but has no way to tell others. something like that could be possible.

@marmot Ah, yes, the Playboy Subscription bet?

@ChristianHupfer No, I think that was more serious....

This gets me thinking about when the great fluid dynamicist von Kármán died and went to heaven. There he met God, who promised he could answer any question. So von Kármán wanted to know how turbulence worked. God got out paper and pen, and scribbled a bunch of equations. After studying these for a while, von Kármán shook his head and said, “no, that's still wrong”.

@naphaneal Yes, possible, but given the amount of time and effort he put into convincing others of his believes I don't think that's very likely.

So anyway, I happen to agree with Hawking about the afterlife, but it still is fun imagining him arguing with God about the big bang and insisting that God had it wrong. It would be just like him, from what I have heard.

@HaraldHanche-Olsen Just like God?

No, just like Hawking. (Of course, some might think he is God.)

@marmot More serious than a One Year Subscription to Playboy? ;-)

@HaraldHanche-Olsen yes, that indeed is amusing.

@ChristianHupfer Playduck nothing to see here

@HaraldHanche-Olsen There's a similar joke regarding a meeting between Jesus and Cardinal Ratzinger. :)

@PauloCereda Who is Jesus? ;-)

@ChristianHupfer The Nazareth bloke. :)

@PauloCereda Brian?

@ChristianHupfer not the Messiah. :)

@PauloCereda Romanes eunt domus ;-)

@PauloCereda great, now I have to create another duck-themed package. Though I don't exactly know what to implement with that title...

@Skillmon ooh

@DavidCarlisle -- have you got time to look at a hyperref linking problem?

@ChristianHupfer Sorry, I misinterpreted your question. You were talking about what the winner will get, and my comment concerns the topic of the bet.

@barbarabeeton probably:-)

@PauloCereda Why are the rabbits helping out the ducks anyway?

@Skillmon because ducks and rabbits are cuddly?

@PauloCereda @DavidCarlisle does have a recipe for a Rabbit-Duck with Orange Sauce, I am pretty sure ;-)

@ChristianHupfer gourmettraveller.com.au/recipes/browse-all/…

@marmot Yes, because I could not remember the topic of the bet. Strangely the eventual profit was stored in my mind ...

@DavidCarlisle As I wrote: I am pretty sure...

@DavidCarlisle you are mean

@DavidCarlisle No oranges.

@UlrikeFischer this has orange but no duck

@PauloCereda Reminds me of the Lemmings of the B.D.A sketch, were Mr. Floppy (the rabbit) is shot ;-)

@DavidCarlisle so now we have rabbit and duck, rabbit and orange, duck and orange, but no rabbit and duck and orange. Sounds a bit like a riddle ;-)

@DavidCarlisle There is no way to cook the the rose killer rabbit! Ordinary rabbit however tastes great :)

OH NO

@DavidCarlisle My faith that everything exists somewhere has been reaffirmend ;-)

@barbarabeeton hyperref?

@DavidCarlisle -- yes, sigh. have just sent.

@barbarabeeton arrived, will look

@ChristianHupfer I see. So you seem to focus on the really important things in life ;-)

^^^ palindrome. sadly, i was too slow to capture 65536.

^^^ i don't know what's happening here, but this just appeared. somebody's tinkering in the background?

@Skillmon I added your package here: tex.meta.stackexchange.com/a/1722/101651. If you like, you could change he prefers to remain nameless into your real name.

@CarLaTeX Thanks. My name is contained in the package documentation and the short description on ctan you link to... But I'll the answer nonetheless :)

@Skillmon I didn't write it there because I didn't know if you wanted to link your real name to your nickname :)

@barbarabeeton must have been our voting squad de-up-voting two your answers :-)

@CarLaTeX it is by providing the link (but not machine parsable). Maybe now google will find more stuff to show if I google myself the next time :)

@Skillmon :)

@yo' -- well, i'd believe that, except that in just a minute or two the score reverted, and there's no report of downvotes or loss of votes. a puzzlement. (but anyhow, it's a nice number to trap. maybe in another ten years i'll get to 131072. (or maybe just to 70k.)

@Skillmon ctan.org/pkg/duckuments the link to the pdf is broken (at least on the oxford ctan mirror)

@DavidCarlisle I got it with the update to TL2018/pretest

@egreg yes but the ctan link is broken

@DavidCarlisle I get it correctly

@DavidCarlisle That's just the English mirror, the Italian one is working well. Never trust the English.

@DavidCarlisle What I said. 😛

@CarLaTeX the UK mirror clearly has more effective spam filtering

@DavidCarlisle You're the best comedian I know

@egreg Seriously? As a German I have to say: Never trust the Italian, they always change sides (lucky in the big cases, I guess)!

@DavidCarlisle Then go and report that to the guys at CTAN :P

@Skillmon a mirror is a mirror, it will no doubt catch up in a bit (how long has the package been on the main ctan host?)

@DavidCarlisle v0.2 for 4 and a half hours. It'll catch up.

@Skillmon what's the problem with xetex? no random numbers?

@DavidCarlisle No \int_rand:nn, but v0.2 works on XeTeX by implementing a pseudo-random number generator :)

@DavidCarlisle Want to code-review?

@Skillmon When the British vote for Brexit, my uncle (84 years old) said: "It's since December 1943 that I didn't trust the English" (they bombed Milan at that time)

@Skillmon is it the same one? (I started to look at how hard it would be to implement the one in pdftex/luatex (from metafont) but have no time at present:-)

@CarLaTeX sadly the british press were saying the same about more or less every country in europe except britain, which is why we are where we are

@DavidCarlisle I didn't take a look at what is done in \int_rand:nn in pdfTeX and LuaTeX. I just implemented RC4 with a few limitations.

@Skillmon For XeLaTeX you can use \input{random}, see an example at tex.stackexchange.com/a/20411/4427

@DavidCarlisle It's very sad for the EU future

@egreg but where's the fun in that?

@CarLaTeX We marmots have left the UK long before the Brexit ;-)

@egreg And it took me only 98 lines of code or so to implement RC4 with expl3.

@CarLaTeX È un peccato, gli inglesi sono così bravi in lingue, ma ce ne stiamo andando

@egreg Is that expandable?

@JosephWright no

@DavidCarlisle We did have an expandable random number generator done in macros ...

It was of course bonkers

@JosephWright and where is it?

@JosephWright there really isn't that much code in luatex's arithemetic.w it's probably reasonably feasable to do in expl3, but expandable would be a bit of a pain. perhaps we should ask again on xetex list

@Skillmon In the Git history ;) Overall, the feeling was to push for engine updates to get primitive support: realistically more sensible

@DavidCarlisle Indeed: you do it by setting \csnames to \relax ...

@DavidCarlisle Yup: we really could do with it

@Skillmon If you really want I'll dig it out

@Skillmon Probably as an interim I'd set up the equivalent FPU calcs, do an x-type expansion into a variable and use that in a non-expandable context

@JosephWright non expandable wil probably do most of the real use cases

@DavidCarlisle Indeed: we can cover that happily. But we can't provide that inside the FPU ...

@JosephWright I have a working PRNG which is non-expandable and producing values in the range 1 to 256 (which is enough atm). So there is no need for an interim. No need to dick out the old code. I can wait till there is a stable version in l3. And if I somehow think there should be an range and/or speed enhancement I can still use random.tex as a last resort.

@Skillmon It's not expl3 that's the issue here, it's the engine support. We don't write XeTeX!

@JosephWright Yes I know that. But I guess there will eventually be a solution, whether in XeTeX or in l3 is currently not important to me. And it is possible in l3 if the XeTeX team is somehow reluctant :)

@Skillmon We did ask, but things got a bit tricky when people looked at the generator code itself. We really want the same code across engines, so the aim is to get the pdfTeX code into XeTeX ...

@JosephWright and if that's not becoming true? Any plans for that case?

@Skillmon Not at present, largely because it's not a major priority. Bruno and I did discuss the (very slow) macro-based approach (we both had the same idea on how to do it): it will likely depend on the team having a need to do this in XeTeX.

@JosephWright I could as well use the Doom engine approach of randomness :) They hard-coded an array with random values...