@JosephWright In a moment, they are in my laptop. :)

@TeXnician I did!

oh no

@JosephWright, @TeXnician ^^

@PauloCereda yeh! Finally you have some text in your thesis document.

@DavidCarlisle you are mean :)

@PauloCereda according to BBC's flagship natural history series running at present, you are not only mean but part of a massive global disaster and responsible for killing baby whales (see twitter link above)

@DavidCarlisle OH NO

Rubber ducks should not be discarded, they should be kept as a souvenirs. :)

@PauloCereda twitter.com/i/videos/tweet/932344200712085504

@DavidCarlisle ooh I must capture them all

@PauloCereda amazon.com/…

@TorbjørnT. oh no

@DavidCarlisle: what has the BBC ever done for us

@TorbjørnT. scientific proof: ducks are evil

@DavidCarlisle OH NO

@PauloCereda @DavidCarlisle The book is more about how stuff floating around the oceans give clues to the way ocean currents move. (So can unfortunately not be used as an argument for saying that ducks are bad.)

@DavidCarlisle ^^ LISTEN TO HIM

@DavidCarlisle oi

@PauloCereda Damn, VS Code does not read my .bashrc (custom PATH). Now I know that it at least can work.

Now it's official: per our president's request, here's the video: vimeo.com/243630927

@PauloCereda Is there a way in arara to tell it "ignore all following rules". That means that the following runs pdflatex only once?

% arara: pdflatex
% arara: endofrules
% arara: pdflatex

@UlrikeFischer version 3 or 4? :)

@PauloCereda currently arara 3.0a - The cool TeX automation tool (from texlive).

@UlrikeFischer hmmm sadly that version has no way of ignoring rules. The new version has two ways IIRC. Let me see if I can find a way of doing that in version 3 as well.

@PauloCereda But in 4 it would work? No need to spend to much time on it, I was merely curious (I have a very long test file with lots of junk in it and older rules down the file can interfere).

@UlrikeFischer Yes, in 4 it surely would work. We have a halt directive and there's also a command line option for that. Heiko's suggestion.

@PauloCereda Heiko is clever ;-)

@UlrikeFischer I knew I was in trouble with arara when I got an email from him. :)

@UlrikeFischer ctan.org/author

@JosephWright He might be an organization. :)

@UlrikeFischer: halt is the directive to go, but we have this:

 -H,--header               extract directives only in the file header

Which only considers directives in the header. Directives "lost" throughout the source code are not considered.

@PauloCereda How is the "header" defined here?

@UlrikeFischer Good question. :) From the beginning of the file, I consider a header any line that is a comment or an empty line. The header ends at the first non-empty, non-comment, line. :)

@PauloCereda Sounds like a reasonable definition.

@UlrikeFischer I need to check how I implemented it. :)

@JosephWright that masks some hidden maintenance arrangements:-)

[paulo@cambridge target] $ cat test.tex
% arara: pdftex
% This is a comment
% arara: pdftex

% arara: pdftex

Hello.
\bye

% arara: pdftex

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

Processing 'test.tex' (size: 100 bytes, last modified: 11/20/2017
11:07:02), please wait.

(PDFTeX) PDFTeX engine .................................. SUCCESS
(PDFTeX) PDFTeX engine .................................. SUCCESS
(PDFTeX) PDFTeX engine .................................. SUCCESS
(PDFTeX) PDFTeX engine .................................. SUCCESS

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

Processing 'test.tex' (size: 100 bytes, last modified: 11/20/2017
11:07:02), please wait.

(PDFTeX) PDFTeX engine .................................. SUCCESS
(PDFTeX) PDFTeX engine .................................. SUCCESS
(PDFTeX) PDFTeX engine .................................. SUCCESS

Total: 0.68 seconds

@UlrikeFischer ^^ yes, my explanation is correct. :)

@JosephWright Oops, didn't realize that css meant bad syntax. I have something else entirely in mind

@AGoldMan CSS's history is a bad as LaTeX's one :-)

@DavidCarlisle :)

@AGoldMan Not sure it means bad syntax: perhaps I wasn't entirely clear in that I'm thinking about the idea of relationships/named parts, not necessarily all of the detail

@DavidCarlisle: I can write longtable in the lion's shirt. :)

Wow, Arno Pro is really that expensive?!

@PauloCereda Looks pretty standard to me

@JosephWright The price?

@PauloCereda Yes

@JosephWright I found US$ 175

@PauloCereda quite not that much for a font...

@yo' really?

@PauloCereda yes. Fonts are expensive, as only very small number of people need them and even less people are not getting them "somewhere"

@yo' Ah that's a good point.

@PauloCereda and you get a reasonable number of fonts for free or with common software. I don't speak of the quality of these of course. (In general, you still have good fonts for free)

@yo' Got it.

@PauloCereda :)

@yo' in other news, did you enjoy my lion? :)

@PauloCereda oh yes, quite :)

@yo' I might charge for people taking photos. :)

@JosephWright Oh, so we're on the same page then. That's all I took out of CSS

@PauloCereda apart from finishing your thesis, what needs to happen before arara 4 gets released?

@StrongBad only that event (I can write the manual). :)

@StrongBad I apologize in advance, writing a manual is difficult... :(

@PauloCereda harder than a thesis?

@StrongBad actually, for me it is. :) English is probably the biggest barrier for me mostly because I need to write a concise, coherent text in a language that I am not used that much.

@PauloCereda why is it in java? why not perl?

java's slow

@PauloCereda you could write it in Portuguese and let @DavidCarlisle translate it with google. He is good at that.

@AGoldMan I went with the language I was most proficient at that time. I didn't know enough Perl to write decent code back then (and I still don't dare write Perl code). Writing software is easy, writing good software is extremely difficult.

hear that

@AGoldMan It depends on several things. Recent JVM's have very good performance.

@PauloCereda really? on my ubuntu box it's pretty slow

at least comparing pandoc's speed to docx2tex

though that's not a good benchmark

@StrongBad Also, I need to think of how I should organize the subjects. There's a lot to talk about in the new version.

@AGoldMan You might take a closer look at possible bindings the OS adds to the applications as a means to track down errors.

@PauloCereda could I get more info on that?

@AGoldMan I don't know much of it, sadly (it never really bothered me). I think entries like ABRT and application "firewalls" (like SELinux or AppArmor) are probably a good start.

@AGoldMan: on my Mac, for example, the performance is very good. :)

It might also have impact on the JVM vendor, configuration, etc.

@PauloCereda You use a Mac? but apple is evil. Just ask Adam

Is there a way to trap a write node in a box that makes it to an output page?

or will it migrate through all the boxes?

@StrongBad Não apenas bom, absolutamente brilhante.

@AGoldMan ?

@PauloCereda ^^

i.imgur.com/0kvtMLE.gif I found a walking bird. :)

@DavidCarlisle I'm working on an approach to making continuation lines

it's common in some books that when a note is cut off to another page, the first word on the next page is placed underneath the column of notes

@AGoldMan but what do you mean by a node? a tex node, tikz? ...

a tex node

like a \write node as written to the page

I'm fairly certain they're called nodes

@AGoldMan yes they are, but so are other things:)

Basically no, in classic tex you can not dissect a box to see what is in it, (luatex changes the rules)

@DavidCarlisle granted. but is there any way to trap one from migrating

@DavidCarlisle you can by unboxing the paragraph and setting it again with a 1sp \hsize

i got that from you

@AGoldMan write nodes don't migrate inserts and vadjust nodes migrate, but write nodes just stay where they are

@DavidCarlisle oh, so there's no way to stop one from writing then

@AGoldMan you can only take a box apart if it just contains boxes penalty and glue, if t has a \write note you can't see it (there is no \lastwrite as there is \lastbox)

@DavidCarlisle I'm not trying to get the \write, I'm getting it as connected to the first word of the box

@AGoldMan no you either need to not \write or throw away the box or close the output stream so the write just goes to the log

@DavidCarlisle eh, none of those are good options, b/c it's from my indexing commands

@AGoldMan you are doomed:-)

@AGoldMan welcome to my world

@DavidCarlisle I could stick an 0pt hskip before the node, so it gets lost when I do my final reflow..

or even a penalty

@AGoldMan basically you need to make sure the \write is in the correct place to start with

@AGoldMan what's the issue that you are trying to solve?

@DavidCarlisle sometimes my continuation line thing picks up a write node, and then you get two entries in the index

@AGoldMan is the code available?

@AGoldMan since you probably have control over the input you should be able to ensure that you never put a write node in a bad place, and also that you never add a write node to your box at all, if instead of \write\tmpout{foo} you do \setbox0\hbox{\write\tmpout{foo}}\dp0=7sp\box0` then if you dissect the outer list and come across a box of depth 7sp you can remove it with \lastbox and choose to put it back or not, depending.

@DavidCarlisle the code's on my github, let me just push this version

@DavidCarlisle github.com/rabbiveesh/dfn-para.git, the continue branch

@DavidCarlisle the macro that does the work is \dfn@continueline, towards the bottom of the file

@DavidCarlisle I don't really have control over the input, it's for footnotes

You can see the behavior if you overload the page with text, it'll cut the footnote block instead of overprinting

which is better than other footnote packages I've used

@AGoldMan beware missing % at ends of lines, you look like you are adding lots of spaces

@UlrikeFischer Just saw on Italian TV a great goal of Mönchengladbach team!

@DavidCarlisle Most of the stuff happens in vertical mode, I use this regularly

but warning taken

I just pushed a test file that illustrates the issue

if you look at the index

@DavidCarlisle I hope you're reading the file in vim

@AGoldMan no a much better editor (firefox)

@DavidCarlisle thanks for taking a look at it

@AGoldMan oh so you want to duplicate a word at the break?