@egreg: Juve bought Hernanes!

@Werner Yay!

Hello!! When I use the command \begin{split} inside the command \begin{itemize} I get an error... Why? Can we not use it there?

The command I wrote is the following:

\begin{description}
\item $\begin{split}+ \ \ : \ \  x_1+x_2 &= a_1+b_1\sqrt{t^2-1}+a_2+b_2\sqrt{t^2-1}=(a_1+a_2)+(b_1+b_2)\sqrt{t^2-1} \\ & \mapsto (a_1+a_2, b_1+b_2)\end{split}$
\item $0\ \  : \ \  0=0+0 \sqrt{t^2-1} \mapsto (0,0)$
\end{description}

Have I done something wrong?

Do you have an idea @egreg ?

@MaryStar Isn't split meant for displayed equation, rather than inline? Anyway, as you realized, it's better to ask a proper question :-)

@MaryStar Rather than split you should use aligned

@egreg \label spacing.....

@DavidCarlisle What's the problem?

@egreg nothing really, just the usual, my answer is better than yours:-)

@JosephWright @yo' I've been playing with some speed tests of your code. Josephs version is more than twice as fast ;-)

Though they are still both fast enough that one will not know the difference.

@daleif :-)

@daleif I worry about performance quite a bit: siunitx still needs to be faster, for example (I have plans for v3 but it's a challenge keeping everything unchanged for the user)

@DavidCarlisle Waiting to see what @FrankMittelbach says on the str question :-)

@JosephWright I know. The major speed problem in siunitx isn't that the table formatting? Being able to switch off selected parts of the parsing will help a lot (I think).

@DavidCarlisle Also working on the lowercase business

@daleif To some extent: the problem comes in that people still want various effects. My latest plan has a much faster parser and data store.

@daleif If I'd just stuck to +1.2e3 and the like it would be a lot easier

@JosephWright Well, we still need table notes etc. But just being able to selectably switch off parts of the parsing one does not need... For example, if a user specifies format 2.4, does the parser even need to look for e syntax, or ()'s?

@daleif True

@daleif I'll think about that in the new code

@JosephWright Of course the formatting might be wrong if the user has then added ()s by mistake. But letting users know about the speed cost, might make then a bit more careful (hopefully)

@daleif What I am doing to speed things up is to store the data in a dedicated format rather than a property list. The latter makes sense for units (open ended) but not so much for numbers (rigid format).

@egreg: ^^ :)

@PauloCereda Not sure I'm happy.

@egreg Really? I find him a great player.

@PauloCereda We'll see. ;-)

@egreg Indeed. :) And we can always bring Felipe Melo back. :P

@PauloCereda NO WAY!

@egreg :)

Any biblatex experts around?

@1010011010 I am sorry, I am here just for the TeX pr0n. :)

@PauloCereda Implying biblatex can't be a part of that. ;-)

@1010011010 :)

@1010011010 as many as last time you asked:-)

@DavidCarlisle Philipp Lehman will one day be in this chat room.

Then we will see!

@1010011010 What do you want to know?

@1010011010 Moar TeX pr0n. <3

@JosephWright I want to know about either the bibliography environment or \DeclareFieldFormat. From the documentation I see no reason not to treat these commands as usual, though they seem to behave in a slightly different way than I anticipated it. I have an example here which globally sets a box, which is typeset outside of the bibliography environment afterwards. This doesn't seem to typeset anything. :-( So I'm confused about the inner workings of biblatex (even though the source seems....

well, the source seems to reinforce my thought that biblatex simply doesn't do anything fancy). Judging from the source the package just has a lot of options, and that's about the only reason for the large amount of lines in biblatex1.sty...

@1010011010 The commands in general only work inside the bibliography

@1010011010 There are two versions largely as the BibTeX/Biber split is tricky and the source isn't in .dtx format (which would make code sharing easier)

@daleif Well, I was just using whatever was ready out there :)

@Joseph: ^^ The best bit of this message is that Bruno does not even ask why, but goes directly to the how part. :)

@PauloCereda :D

@yo' <3

@JosephWright Hm, thanks for the answer I guess. I'm having some trouble compiling an example which shows what I mean, but it's on the way (I'll post it as a question anyway).

Have anyone had issues with FF pdfreader vs beamer handout + pgfpages? I have one that changes \sum and \Bigcup and I cannot figure out what is causing it.

If we remove the handout \sum and \Bigcup renders just fine.

@daleif Do you have a sample file?

@PauloCereda can I send it to you by email?

@daleif Sure thing! :) Do you have my email?

@PauloCereda nope

;-)

Found it, it is caused by \underbrace. It was only used once in the slides, with out it, things renders ok, with it, things go strange in FF. I'll prepare an MWE

@daleif ooh :)

@daleif cereda at users dot sf dot net

hmm, not `\underbrace on its own. MWEs can be hard ;-)

@PauloCereda I'll send an email in 2min

@daleif Thanks. :)

@daleif -- but the spacing of the lower limits is execrable.

added it as a question

@daleif: replied. :)

@PauloCereda thanks for testing

@daleif My pleasure. Something fishy is going on.

@daleif Can I force a document to open in the preview if I forbad this crap super-new FF feature?

(btw, English verbs are crazy: forbid -- forbade/forbad -- forbidden, but bid -- bid -- bid)

@yo' Forbadsen. :)

@PauloCereda well, I've always thought that the past tense is simply forbid ...

Another day, another siunitx version :-)

@JosephWright Yay!

@PauloCereda Minor version: just bug fixes

@PauloCereda I am working on v3 but it's hard

@JosephWright :)

@JosephWright Tell me about versions. :P

@PauloCereda New code is much clearer than the existing version (which is clearer than v1): github.com/josephwright/siunitx/tree/master

@JosephWright: arara 4.0 is stuck since forever because I have a PhD to do! Oh no! :)

@JosephWright <3

@PauloCereda Luckily v1 isn't on GitHub, so you have to work harder to find it!

@JosephWright: arara is the best British tool ever written by a Brazilian. :)

@daleif Can you please check if this is relevant?

github.com/mozilla/pdf.js/issues/6394

@JosephWright You naughty boy. :)

@PauloCereda I have got a LaTeX-related project in mind, but too little time to work on it. And there's yet an unfinished project for the university :-(

@yo' :)

@PauloCereda don't mention latex3....

@PauloCereda also I wonder if there would be a use for such a project...

@PauloCereda Before I used Git

@DavidCarlisle Indeed

@DavidCarlisle You can guess what's coming next

@JosephWright LaTeX4? :-O

@JosephWright xor?

@DavidCarlisle We have a winner :-)

@PauloCereda a not an :-)

@DavidCarlisle I need Frank to reach a decision about engine prefixes first, though :-)

@PauloCereda with a typo? You're like @David :)

@DavidCarlisle Oh no! :)

@JosephWright then I can write xor in lua, and Bruno can use the engine prefixed tex code to write a lua interpreter in pdf and xe tex.

@DavidCarlisle :-)

@PauloCereda <3

@DavidCarlisle: no one will ever notice the fix. :)

One suspects a certain amount of hand correction to the proofs....

@percusse I don't think so, because the same file renders just fine in Chrome (though I'm not sure if my Chrome is the very latest)

@yo' <3

@yo' note sure what you are referring to?

@daleif if I click a PDF file, FF never ever uses the PDF.js thingy (or how's it called), I taught FF to ask me whether I want to download the file or open it in AR

@yo' in settings, see under applications, Portable Document Format, I have two of those for some reason. I'm not sure what trickers the show inside FF feature.

@daleif but I am sure I don't want to modify anything there :-)

@yo' it is a quite easy dropdown interface

@percusse Then again. I can see that I'm not using the very latest.

@Paulo Finally there's a good synth/keyboard that knows what organ stops are! youtube.com/watch?v=jv9JLTMsOCE Now I only need to see if I stay here or move elsewhere. If I move, I'll buy this there almost for sure.

@JosephWright most likely that I'm for a full str module ... but I need a little more reading and backlog clearing first ... and have my last day at HP tomorrow so need to clean out a lot of stuff today .. and that does have a deadline now

@FrankMittelbach Ah

@FrankMittelbach Clearly a priority!

@FrankMittelbach I have other things to think about :-)

@JosephWright kind of ... need to make sure I have everything needed once I have given back my company laptop

@FrankMittelbach buy an external HDD and make an HDD image? :)

@yo' I was going to suggest simply taking the disk out and giving it back with no disk:-)

@yo' that's not the issue. with the laptop I lose access to the company network and anythign I need from there during the next two months, such as email address of payroll etc etc needs to be 100% collected first

@DavidCarlisle if you're sure you can plug it elsewhere :)

@DavidCarlisle That might be, erm, questionable

@FrankMittelbach ah that is an issue.

@DavidCarlisle that disk is hardware encypted :-) so no way, but it isn't the disc content that is the issue

@yo' ooh

@yo' Ze disk, we must save ze disk! :)

\documentclass{article}

\usepackage{marvosym}

\begin{document}
AC\Lightning DC
\end{document}

@percusse: ^^

@PauloCereda but seriously, if I move to Japan, I'll buy this one there! (And if not, I'll buy it here and finally join some band hopefully!)

@daleif: I love the Doctor Who reference. :)

@PauloCereda Es-tu français? :)

@yo' Ze french, I spracht kleine Français, allons-y!

@PauloCereda oh mein dieu!

@yo' LOLOLOL

A question, with no background, it just came to my mind. What would be the downsides of defining token lists in expl3 as \protected? So that \tl_use:N \l_tmpa_tl was necessary to expand them.

@Manuel what would be good about it? \protected moreorless only affects moving arguments anyway, it does not prevent expansion.

I don't know, I just saw \cs_set:Npx \foo: { \tl_set:Nn \exp_not:N \l_tmpa_t { \token_to_str:N _ } }.

@PauloCereda hehe

@Manuel that looks to me as a wrong approach.

@Manuel I'm not sure of the usefulness of \foo:, as executing them would again set \l_tmpa_tl to contain a category code 12 _.

@PauloCereda Can you please learn this and teach me later?

coffeescript.org

I'm too lazy to learn

But I need it for Bokeh fiddling

@egreg exactly, the right thing to do would be \tl_gset:Nx \c_underline_tl { \token_to_str:N _ } (or maybe \token_to_str:n)

@yo' Nope, wasn't an example, it was a shortcut (that didn't work). The code is different, just looking inside siunitx I see many \exp_not:N \l_whatever_tl.

@percusse Got the general idea, so I can help you, sir. <3

Online mode!

Not now please. I'm out of brain :)

I left it at home

@percusse Ah no worries. :)

I is listening to ACLIGHTININGBOLTDC.

@yo' It wasn't literal, just a (bad) example. The thing is, wouldn't it be “more better” that token lists would be protected, so if one wants expansion just needs \tl_use:N?

@percusse And what's the other status: being "in your brain"? :)

@yo' \tl_const:Nx \c_underscore_other_tl {\token_to_str:N _}

@egreg ah \tl_const I forgot about it!

@Manuel You more often want them to expand!

@yo' ...

yet you still ask difficult questions

@percusse LOLOLOLO

@egreg would you Latin in English such as e.g., i.e. or per se in italics?

@egreg But you have to use the _use:N functions for everything, and just because now with non-toks token lists you don't necessary need them… well I don't think we should drop the \tl_use:N.

@yo' no

@egreg ok

@Manuel It's always been optional for token list variables.

@PauloCereda Now with Mad cow flavor.

Or flavour.

@percusse hahaha

Lolo is a chocolate bar in here. :)

english is taff

@yo' But “per se” or “a fortiori” can be in italics. I was referring to “e.g.” and “i.e.”

@egreg so no for the abbrv. and yes for this more rare stuff? What about et al.?

@yo' It's common usage.

a fortiori (latin): In such a way that even a brick can understand ?

@egreg ok

@percusse Uhm, no!

@percusse in such a way that even @percusse in the brain-less mode can understand? ;)

@yo' What? :P

@percusse “Because we are under a much stronger assumption”.

So why not saying that in english? Too cocky?

@egreg That's where I go. I don't think it wouldn't hurt if it wasn't optional. In which case, apart from the fact that it isn't optional, would be downsides with \…_tl being protected?

@Manuel That they wouldn't expand in an x context, which is usually wanted.

anyway I'll do what I don't have any thinking left in me.

Drive home :)

Tot later

@percusse good idea. I'll be here probably until about 9pm today. Bye!

@egreg That's what \tl_use:N would do in that hypothetical case.

@Manuel that starts to be really complicated :)

@Manuel Problem is \exp_after:wN \foo \l_some_tl would still work, useful in one way but a bit confusing in another

@yo' I'm sorry I explain too bad in english :) I mean, if the rule was use \tl_use:N to retrieve the content's of a token list (so it's not optional, which is actually seen in many places). And then we wouldn't have to \noexpand them in such contexts.

@JosephWright \foo:V \l_some_tl or \exp_args:NV \foo \l_some_tl isn't that confusing.

@Manuel also, note that all usages of \exp_not:V are moreorless "inside" \exp_not:N or \exp_not:n, i.e., in places were some other things have to be \noexpanded.

@Manuel True, I'm thinking low-level stuff (where you don't want braces)

@Manuel Probably this might have been worth exploring if we'd started expl3 when e-TeX was finalised, but as with many things the ideas go back a long way and at some stage you have to make decisions

@JosephWright Okey. It's just that I asked myself if there weren't downsides with \protected token lists, and couldn't think of this things.

@JosephWright True, it was just curiosity.

@Manuel I've wondered if one would use registers at all if you were starting today: with \numexpr, etc., you could do everything with macros and avoid any need for \<thing>_new:N (i.e. soft typing). There are then performance questions, termination issues and primitive cases to worry about, but I suspect in principle it's doable.

@JosephWright unless you decided to implement tls as \toks, no?

@yo' True

@yo' Even with a modern TeX system the total registers available might be an issue: you can have a lot more macros than toks

@JosephWright yeah, I know. Also, you probably don't gain any performance

@yo' Like I say, I've wondered about various things in this area

@JosephWright Yes, I thought about that some time ago. I think I used some counters as macros at some point \def\increment#1{\edef#1{\the\numexpr#1+1\relax}} or something like that.

@JosephWright do you mean like, setting \dim_set:Nn -> \def #1 {\the\dimexpr#2\relax}?

@Manuel Yup, we had that as 'fake counters' at one point

@yo' Yes

@JosephWright oh my goodness :)

can it work with glues?

@yo' Issues are for example that #1 here is then not terminated, which is a problem at the low level but would also be a pain for speed (re-parsing)

@yo' \glueexpr

@JosephWright ah there is \glueexpr (one learns new things every day :-) )

@yo' Like I say, I can see some advantages

@yo' Soft typing might be fun :-)

@JosephWright What does “terminated” mean there? I'm not familiar with the terminology :)

@JosephWright like the ability to define 10000000 glue variables?

#1 is "bare" there, so if someone does \dim_set:Nn { \hello{world} blabla} {10pt}, strange things happen and the error is difficult to understand to

@Manuel For example if \foo is a count then \ifnum #1>\foo doesn't keep looking for a number, but if \foo is a macro then TeX keeps going

@JosephWright Got it. I hope \ifnum{#1}{\foo}{True}{False} would work hehe

@Manuel Like I say, one can speculate for a long time on these things. @FrankMittelbach and @DavidCarlisle can I am sure tell you lots of other good/interesting ideas that have been explored/mentioned/imagined over time.

@Manuel The big issue for me is delivery: we have to make some decisions and go forward even if we therefore cut off interesting other things

@Manuel Perhaps I should knock up a set of data structures using just macros, for a bit of fun [and a set that are all protected :-)]

@JosephWright I'm just exploring things myself “for fun”. I don't mean as serious suggestions, and as you say you already thought of everything. It's just that I'm getting at those points myself so I ask for opinions :)

@Manuel well, skepticism is a way to understanding :-)

@Manuel I guess I'd favour (slightly) the current set up even if starting today as it's normally \exp_not:V that applies in an expansion context when using tl data. That would be true whether they are protected or not. Certainly there is no big technical reason either way in my mind: it's primarily historical (expl3 pre-dates LaTeX2e and so e-TeX!)

@Manuel not really possible in TeX is it?

@DavidCarlisle Huh?

@JosephWright tex being a macro language means macros expand without being prefixed by \tl_use. \protected would affect expansion contexts but not use "in the wild" I don't see any way of having a macro that by default doesn't expand.

@DavidCarlisle Ah, yes

@DavidCarlisle I don't mean that, it's just that inside in \edef they wouldn't expand.

@DavidCarlisle Probably the 'always use toks' suggestion would be needed then

@Manuel yes but that means you can not make tl_use mandatory

@JosephWright yes but then you have no macro arguments

@DavidCarlisle My 'don't use registers' plan is a bit easier :-)

@DavidCarlisle \tl_new:N is not “mandatory” either.

@DavidCarlisle Token lists don't need arguments

@Manuel Not exactly, that is true

@DavidCarlisle well, you can, by defining \tl_set:Nn -> \def #1 {\tl_badly_used: { #2 } } and \tl_use:N = \@secondoftwo

@JosephWright Because you check for existence?

@Manuel With checking active, yes (Bruno is working on 'strict mode' code, slower but with more checking)

@Manuel because you can check for existence, by going to strict mode (or how's it called)

@JosephWright @yo' Okey, it was just to put David on what I was trying to say.

Why spend time doing editing work, when you can spend time making an Emacs macro that can to 10% of the work ;-)

@yo' well, given that I'd been using Emacs for 10 years before I started making macros.....

@daleif :)

@JosephWright @DavidCarlisle Have you got any ideas on this, please? ell.stackexchange.com/questions/65923/…

@yo' car allowance

@yo' I suspect there may be a different term in US English

@JosephWright ok, thanks

@yo' The answer is --> youtube.com/watch?v=RnoDb0bMQuk

@percusse answer to what?

@JosephWright you're about to get mail :(

@DavidCarlisle Yes

@DavidCarlisle What does reldemac do that's so tricky?

@JosephWright it has series of footnotes for different types of footnotey thing, quick eye over the code I think by default it has 10 of them but duplicates for minipages as latex footnotes do the mpfoot... ones don't need to be real inserts but it probably simplifies the code if they are. So that's 20 inserts and more if the user declares a new footnote series

don't use footnotes in your books Don

@DavidCarlisle Ah

@JosephWright I was thinking while writing the mail so not tried it yet that given that the new \newinsert takes from the float list I could define \reserveinserts to add that number of "classic" insert registers to the float list where later \newinsert will find them, would need a few checks but should only be a line or two of code.

@DavidCarlisle Strings seem popular

@DavidCarlisle Not a bad plan, at least as you describe it (I leave such 2e complexity to you!)

Writing \label instead of \ref can give surprising results :)

@yo' Really? ;-)

@egreg one is especially surprised why is Figure 1-1 referred as Figure 1-6 ;)

@yo' \label after \caption!

@egreg that's not the problem; the problem was that Section 1-6 contained a sentence with: ... as we can see in Figure~\label{fig:whatever}

@yo' I see, but that's a case of \label in bad position with respect to its caption, isn't it? :P

@egreg no, the figure itself is coded properly. Well, yes, somehow it is "label in a bad position", but more it's "using label instead of ref"

@yo' Just joking!

Hi @YiannisLazarides!

Hi! Guys

@YiannisLazarides Long time no see! How are you?

Everything ok, just been too busy...

@YiannisLazarides Ciao!

Hi egreg, how is Italy, still hot?

@YiannisLazarides Very.

@egreg any rain recently? It's just started here I think (and I'm still at work)

@yo' Possible thunderstorms tomorrow. Probability of rain in the afternoon 30%

@yo' No rain in the last week.

@egreg no rain here either for about a week, and for the last two days, temp. up to 37 again :-( Hopefully it was the last time this year.

@DavidCarlisle \outer bites :-)

@DavidCarlisle We did have a comment in expl3 about \reserveinserts being \outer (as that came up with a dodgy redefinition of \RequirePackage at one point)

@JosephWright in my code, :(

@JosephWright looking forward to discussing allocation changes again on team list

@DavidCarlisle :-)

@yo' Trying to fight the hot temperature with a glass of Pilsner Urquell. :)

@egreg ooh :)

@PauloCereda And air conditioning, to be honest. ;-)

@egreg You little rascal. :)

@PauloCereda -- rascal, definitely. not so inclined to apply "little".

@PauloCereda old is the more usual (and in this case appropriate) prefix to "rascal"

@barbarabeeton oh sorry. :)

@DavidCarlisle ooh

@PauloCereda -- you need to apologize to @egreg, not to me.

@egreg: sorry. :)

@barbarabeeton Done. :)

@PauloCereda -- how come you've got two different avatars showing up on your comments?

@barbarabeeton Really? My current one is a duck. Maybe it's some caching going on.

@PauloCereda you should get one of these, new coin minted today:

@PauloCereda -- probably caching. the "large" one is the "bunny duck", and the smaller one is the duck with a blue-with-white-trim hat. (i've never seen you as anything but a duck.)

@DavidCarlisle ooooh long live the queen. :)

@DavidCarlisle -- i do like the hat! (rather mad hatter style, but lacking the price tag.)

@PauloCereda she passes queen victoria as longest ever monarch this week.

@barbarabeeton the price tag is round the other side, behind that feather.

@barbarabeeton Really? I used to use my own photo as avatar, and recently I had changed it to yet another photo of mine, but peer pressure made me come back to ducks. :)

@DavidCarlisle Awesome! :)

@barbarabeeton: ^^ I was using this avatar. :)

@PauloCereda @barbarabeeton we made him change back as it was too disconcerting and clearly not a duck.

@PauloCereda -- i guess i just wasn't around for those few hours.

@DavidCarlisle <3

@PauloCereda Is there a good way to do this but using arara?

@AdamLiter There is. :) I will include a code argument to all engines, so one can do this in main.tex:

% arara: pdflatex: { code: '\def\toname{Foo}\input{main}' }
\documentclass[addrfield]{scrlttr2}

\begin{document}
\begin{letter}{\toname}
\opening{Dear \toname, }
 A nice letter.
\end{letter}
\end{document}

I am not sure if there's a better way of doing that.

@AdamLiter: I could also include a jobname argument as well!

And it could be a list!

@PauloCereda But what about the for loop from the command line? I guess that's more what I was asking about. Say that I wanted to call arara from inside of a for loop on the command line and pass the index of the for loop to arara as the jobname. Is there a way of doing that?

@AdamLiter Hm I see now. Let me think for a while.

@AdamLiter: I think arara is not ready for getting external information. I'd implement this loop inside arara and not outside.

@PauloCereda Don't worry too much about it. I can do what I'm trying to do by just calling xelatex instead of arara, but I'm just used to using arara to build my documents now. :)

Say, % arara: pdflatex: { jobname: [ foo, bar ] }

@AdamLiter ooh <3

@PauloCereda So is batch generation of documents possible inside arara, then? Would what you just suggested produce two PDF files, foo.pdf, and bar.pdf?

@AdamLiter Not in 3.0, but surely possible in 4.0. Hold on, I'll write a sample for you to check. :)

@PauloCereda And, if so, is there a way to associate say, \def\mymacro{foo} to the PDF with the jobname of foo and \def\mymacro{bar} to the PDF with the jobname of bar?

@AdamLiter Hold on hold on hold on, I'm a slow duck. :)

@PauloCereda ^^^

@AdamLiter awww

% arara: pdflatex: { jobname: [ letter1, letter2 ] }
\documentclass[addrfield]{scrlttr2}
\input \jobname.adr

\begin{document}
\begin{letter}{\toname}
\opening{Dear \toname, }
 A nice letter.
\end{letter}
\end{document}

[paulo@cambridge arara] $ java -jar arara-4.0-jar-with-dependencies.jar main.tex
  __ _ _ __ __ _ _ __ __ _
 / _` | '__/ _` | '__/ _` |
| (_| | | | (_| | | | (_| |
 \__,_|_|  \__,_|_|  \__,_|

Processing 'main.tex' (size: 220 bytes, last modified: 09/01/2015
20:07:27), please wait.

(PDFLaTeX) PDFLaTeX engine .............................. SUCCESS
(PDFLaTeX) PDFLaTeX engine .............................. SUCCESS

Total: 0.48 seconds
[paulo@cambridge arara] $

[paulo@cambridge arara] $ ls letter*
letter1.adr  letter1.log  letter2.adr  letter2.log
letter1.aux  letter1.pdf  letter2.aux  letter2.pdf

@AdamLiter: ^^

@PauloCereda Awesome! :p This will be part of 4.0, then?

@AdamLiter If I don't end up crazier, yes. :)

@PauloCereda lol, how's the manual going?

@AdamLiter Terribly slow. :( I am stuck with my PhD thingy for the moment.

But I will work on it soon.

@PauloCereda Ahh, I heard those thingies take up quite a bit of time. :p

@AdamLiter They surely do.

The idea here is to improve the engine rules, and that's about it. Nothing major to the tool.

So it will happen. :)

@AdamLiter: You will be my tester. :)

@PauloCereda Yes, I remember you also wanted me to test the knitr rule, too, but I can't remember where our email correspondence ended.

@PauloCereda From what I recall, though, it seemed like it was working?

@AdamLiter Yes, definitely. :)

arara grew up to be an awesome tool.

This will be the most epic version ever.

@PauloCereda I'm excited for it. :)

@AdamLiter Me too. <3

@PauloCereda Ah, sorry to be a pain, but I just realized that your solution might not completely cover my current use case. I also want different output directories for each jobname ... :p