TeX, LaTeX and Friends

12:21 AM

@wilx -- nobody ever said you couldn't provide incorrect (or even absurd) \hyphenation rules. but you'd better watch for what a (human) editor will call you out for. nonsense like that won't be appreciated very much.

wilx

@barbarabeeton :)

yo'

7:01 AM

@barbarabeeton :)

David Carlisle

@wilx You should make @barbarabeeton happier and change that to \hyphenation{tw-itch} :-)

yo'

7:23 AM

"Hey guys! I have a great idea for a migration!" "Dammit, Kevin."

@PauloCereda ^^ ducks or geese -- who knows the difference? :)

pogo-otter

@yo' geese are pure evil.

yo'

@pogo-otter unless you keep them for foie gras

pogo-otter

@yo' there's a reason why some people keep them as watch geese instead of dogs. (apparently, that's the single point for them - they don't seem to steal and kill chicken.. but who knows.)

@yo' sounds like the caviar-case to me. everybody tells you it's delicious om-nom-stuff but when you try it .. oh well. not quite so good.

yo'

@pogo-otter oh I love it! (one thing is, many various products around the world are called foie gras, but not all of them are actually foie gras)

pogo-otter

7:40 AM

@yo' ugh. just checked what it really is .....sorry, but I won't join the pro-foiegras party anytime soon. :D here, liver's more something you terrorise children with

yo'

@pogo-otter that's commonly the case. (I love it, my brother not so quite, actually the only liver product he eats are pâtés)

Joseph Wright

1

Q: Chinese Characters And Image Insertion In LaTeX 2.09

I am working on my CV and have made the unfortunate decision to add a couple lines of Simplified Mandarin and an image to the document. While normally this is quite easy to do through the use of the CJK and graphicx packages, those are not supported in 2.09 and MP_Res has not been ported over fro...

Oh goodness!

CV classes are bad enough as it is (they really don't work, in my opinion)

@DavidCarlisle Did you build a new LuaTeX which fixes the \penalty issue?

David Carlisle

@JosephWright yes, seems good

@JosephWright but got twice as many failures as the logging for all \discretionary has changed

Joseph Wright

@DavidCarlisle Ah

@DavidCarlisle Can you send me a couple of examples? I'll see what I can work out for l3build (if anything: it's getting tricky to do at all)

@DavidCarlisle BTW, what did you think of the message on the new list?

David Carlisle

@JosephWright still lists itself as 0.98.4 so you just have to go by the svn revision (which I don't think you can see at runtime)

Joseph Wright

7:51 AM

@DavidCarlisle I guess I can ping Akira

David Carlisle

@JosephWright I wondered what Hans would make of it:-) seems about right though,

@JosephWright mail coming up...

@JosephWright although they all look like:

*** ../build/test/tl2e2.luatex.tlg      2016-09-05 22:54:08.461192400 +0200
--- ../build/test/tl2e2.luatex.log      2016-09-05 22:54:08.490200300 +0200
***************
*** 119,124 ****
--- 119,126 ----
  ....\OT1/cmr/m/n/10 i
  ....\OT1/cmr/m/n/10 n
  ....\OT1/cmr/m/n/10 i
+ ....\discretionary (penalty 50)
+ .....< \OT1/cmr/m/n/10 -
  ....\OT1/cmr/m/n/10 t
  ....\OT1/cmr/m/n/10 a
  ....\OT1/cmr/m/n/10 l

Joseph Wright

@DavidCarlisle :-)

David Carlisle

or like

*** 420,428 ****
  ....\OT1/cmr/bx/n/12 p
  ....\OT1/cmr/bx/n/12 r
  ....\OT1/cmr/bx/n/12 o
! ....\discretionary50| replacing
! .....\kern0.375
! .....\OT1/cmr/bx/n/12 -
  ....\OT1/cmr/bx/n/12 d
  ....\OT1/cmr/bx/n/12 u
  ....\OT1/cmr/bx/n/12 c
--- 421,429 ----
  ....\OT1/cmr/bx/n/12 p
  ....\OT1/cmr/bx/n/12 r
  ....\OT1/cmr/bx/n/12 o
! ....\discretionary (penalty 50)
! .....< \OT1/cmr/bx/n/12 -
! .....= \kern0.375
  ....\OT1/cmr/bx/n/12 d
  ....\OT1/cmr/bx/n/12 u
  ....\OT1/cmr/bx/n/12 c
***************

Joseph Wright

@DavidCarlisle I'm not saying I don't understand their position, but I think it's missing the longer-term and wider situation

@DavidCarlisle That I guess might be trappable: provided the lines have a 'marker' at the start and a common form they can be re-written

David Carlisle

@JosephWright yes I'm not sure quite what the format is yet, in my understanding there ought to be three lists (no break/pre/post) plus the penalty but it only seems to log two plus a penalty, I suppose some experimentation needed:-)

@JosephWright I suspect that if you dropped the (penalty 50) the < marker on the - line and the whole = \kern line then you'd get back to the pdftex tlg (not the already saved luatex 0.95 tlg)

yo'

8:06 AM

@DavidCarlisle probably empty discretionary list is not shown?

David Carlisle

@yo' probably (although that leaves open the question if the usually-empty field is not empty does it remember to log itself:-)

yo'

@DavidCarlisle you'll have to find out ;-)

David Carlisle

@yo' (@JosephWright)

...\tenrm a
...\discretionary (penalty 50)
....< \tenrm o
....< \tenrm n
....< \tenrm e
....> \tenrm t
....> \kern-0.27779
....> \tenrm w
....> \kern-0.27779
....> \tenrm o
....= \tenrm t
....= \tenrm h
....= \tenrm r
....= \tenrm e
....= \tenrm e
...\tenrm b

from

a\discretionary{one}{two}{three}b

yo'

@DavidCarlisle interesting that the kerning is symmetric around w; wouldn't have expected that :)

David Carlisle

@yo' always expect the unexpected

Joseph Wright

8:53 AM

@DavidCarlisle OK, so if we spot 'replacing' we need to do a shuffle, the < can simply be removed, ...

Johannes_B

@DavidCarlisle I removed my comment asking for the extra stuff. But the new comment by the OP ... what?

David Carlisle

9:04 AM

@JosephWright yep

@JosephWright to make it magically look like the pdftex version which is:

...\discretionary replacing 5
....\tenrm o
....\tenrm n
....\tenrm e
...|\tenrm t
...|\kern-0.27779
...|\tenrm w
...|\kern-0.27779
...|\tenrm o
...\tenrm t
...\tenrm h
...\tenrm r
...\tenrm e
...\tenrm e
...\tenrm b

@Johannes_B I left another comment

@JosephWright as you probably saw on list, just built luatex again, the interactive no-file use seems fixed.

Joseph Wright

10:16 AM

@DavidCarlisle Probably will wait to see what Hans goes with once he's had a bit of time to think about this (may yet change again!)

Paulo Cereda

@JosephWright: why the discretionary thingy has changed? I thought the measurements were supposed to be consistent between implementations.

Joseph Wright

@PauloCereda LuaTeX is specifically not back-compatible: Hans is pretty clear that logging really can change drastically

David Carlisle

@PauloCereda consistent between implementations of TeX but luatex hyphenates using a different algorithm, in different places, and reports them differently in the log even when they are in the same place.

Paulo Cereda

@JosephWright, @DavidCarlisle: Ah I see, thanks.

David Carlisle

@PauloCereda and luatex .98 hyphenates in different places, and uses a completely different log reporting to luatex 0.95 (or a luatex 0.98 built yesterday)

Paulo Cereda

10:27 AM

@DavidCarlisle Got it.

David Carlisle

@PauloCereda meanwhile users expect to change from pdflatex mythesis to lualatex mythesis and have it all work the same way... (users who have a thesis.tex, that is)

Paulo Cereda

@DavidCarlisle oooh ~~I am safe~~ you are so mean!

Joseph Wright

@DavidCarlisle Really not feasible, as you know (though getting this message out is hard)

@DavidCarlisle BTW, did you see @barbarabeeton's recent answer about Unicode Maths fonts missing a couple of fontdimens?

David Carlisle

@JosephWright getting @PauloCereda to have a thesis.tex is not feasible? seems that way.

Joseph Wright

@PauloCereda Pick one engine and stick to it

Paulo Cereda

10:30 AM

@JosephWright Sir yes sir. :)

David Carlisle

@JosephWright no, although I saw some possibly related back discussion, link?

Joseph Wright

3

A: Wrong parentheses size in \binom with xelatex and unicode-math in displaystyle

The unicode-math and stix/xits fonts are natively OpenType fonts. Setting of math is accomplished by means of parameters provided by the OTF math table. The OpenType mechanism was a creation of Microsoft. The math table, although it is based largely on the mechanism used by TeX, as described in...

David Carlisle

@JosephWright ah thanks yes I think I saw that in the stix testing forum... (one or two places where the opentype math table could do with some extra slots Fred Wang has some discussion about square root positioning)

@egreg unfair advantage you have, to know about \ped

David Carlisle

11:24 AM

@barbarabeeton :

0

A: Wrong parentheses size in \binom with xelatex and unicode-math in displaystyle

If you wamnt to use \left/\right I don't think you need an array and \mathchoice just use \atop or \over without the delimiters, then wrap in left/right pair. \documentclass[a4paper,oneside]{memoir} \usepackage{amsmath,unicode-math,xltxtra} % Those are needed to use same fonts as I do. % Pa...

daleif

Just wondering here. Related to the problems with \binom, the steps in which say \sqrt grow in a particular math font, is that also governed by features of the font? Here only related to normal TeX fonts (with pdflatex).

David Carlisle

@daleif yes the tfm file encodes the square root pretty much as it encodes a stretching bracket, as a linked list of fixed size getting larger then finally as a top, straight-extender and bottom. The over bar is a drawn rule not taken from the font (of a width to match the height of the character used for the left)

@JosephWright speaking of harfbuzz, see this?

frederic-wang.fr/opentype-math-in-harfbuzz.html

Joseph Wright

11:40 AM

@DavidCarlisle I'll read a bit later

pogo-otter

@DavidCarlisle so, my thesis.tex has still a managable size. you'd advise to change to either pdflatex or lualatex early on in the process?

yo'

@pogo-otter the point is: will you need the functionality of lualatex? If no, just stick to pdflatex.

wilx

There should be like a database of this package should be included before this and and this and this, or the other should be included after this and this. Then we could like create a preprocessing tool that would read the LaTeX source and figure out the appropriate ordering based on the known constraints.

pogo-otter

@yo' yeah. i'm trying to figure that out tex.stackexchange.com/questions/126206/… answers to previously read posts seemed to hint that lualatex would take less compiling time or might be easier for larger documents. But I'll read up.

yo'

@pogo-otter I've seen thousands of pages produced by pdflatex; and lualatex is very problematic with microtype, which I hated

pogo-otter

11:48 AM

@yo' see christians first comment on the linked topic

wilx

@yo' Problematic?

yo'

@wilx I'm not sure I have the time to explain that; you need some careful setup to make luatex even start compiling in reasonable time if you use microtype

wilx

@yo' OK. I have always found LuaLaTeX to be slower than XeLaTeX.

yo'

@wilx never used xelatex myself, so i dunno

Paulo Cereda

[paulo@cambridge documento] $ java -jar arara.jar thesis.tex
  __ _ _ __ __ _ _ __ __ _
 / _` | '__/ _` | '__/ _` |
| (_| | | | (_| | | | (_| |
 \__,_|_|  \__,_|_|  \__,_|

Processing 'thesis.tex' (size: 11 KB, last modified: 09/06/2016
08:51:35), please wait.

(PDFLaTeX) PDFLaTeX engine .............................. SUCCESS
(PDFLaTeX) PDFLaTeX engine .............................. SUCCESS
(BibTeX) The BibTeX reference management software ....... SUCCESS
(PDFLaTeX) PDFLaTeX engine .............................. SUCCESS

@pogo-otter ^^ less than 6 secs doesn't seem too much, at least for my thesis. :)

@DavidCarlisle: ^^

yo'

11:54 AM

@PauloCereda that's because as we all know, your thesis miss any character...

pogo-otter

@PauloCereda 11KB is also quite economical. But you already said that you only had one figure :D

yo'

@pogo-otter do you use tikz?

Paulo Cereda

@pogo-otter Have you ever heard of \include and \input? :)

pogo-otter

@PauloCereda will you use ducks in the thesis? (i hid a delorean in my masters presentation...)

Paulo Cereda

@pogo-otter Of course! :)

pogo-otter

11:56 AM

@yo' yes, which is why i am concerned if i can avoid future trouble in compiling time

@PauloCereda ah, damn. what about the pdf, then?

yo'

@pogo-otter ever heard of externalization?

pogo-otter

@yo' Do you mean to only include some subfiles in the compile process?

Paulo Cereda

[paulo@cambridge documento] $ pdfinfo thesis.pdf
Creator:        TeX
Producer:       pdfTeX-1.40.17
CreationDate:   Tue Sep  6 08:53:20 2016
ModDate:        Tue Sep  6 08:53:20 2016
Tagged:         no
UserProperties: no
Suspects:       no
Form:           none
JavaScript:     no
Pages:          67
Encrypted:      no
Page size:      595.276 x 841.89 pts (A4)
Page rot:       0
File size:      240556 bytes
Optimized:      no
PDF version:    1.5

So far... :)

yo'

@pogo-otter no, I mean tikz externalization; a tikz figure is only recompiled when it's modified; otherwise it's included as a PDF

pogo-otter

@PauloCereda ducks or didn't happen.

3

Paulo Cereda

11:58 AM

@pogo-otter <3

pogo-otter

@yo' no, can you give me a link to start reading on that? Seems extremely interesting. Compiling time is no joke atm. :D

yo'

@pogo-otter been there done that (the time price of tikz)

pogo-otter

@yo' (but it's so pretty!)

yo'

the basic idea is that you do \usetikzlibrary{external}\tikzexternalize in the preamble. If it works out of the box, you're fine; if not, you've got to debug it, there are issues in that sometimes.

Torbjørn T.

@pogo-otter Chapter 50 in pgfmanual.pdf.

Ulrike Fischer

12:01 PM

@pogo-otter lualatex and xelatex are slower than pdflatex. Normally even if I intend to use lualatex at the end I setup documents so that I can run pdflatex during the drafts.

David Carlisle

@pogo-otter change to those from what?

barbara beeton

@DavidCarlisle -- hmmmmph. troublemaker.

David Carlisle

@pogo-otter since when I started to use tex 15 minutes per page was common, I find it very hard to be sympathetic to people complaining about the length of time it takes to process tex these days.

@barbarabeeton you must have pinged the wrong david

pogo-otter

@DavidCarlisle funny, a few days ago, a collegue told me about the war over the limited computation time of tape card programs in the data offices

@TorbjørnT. @yo' thanks, I'll try it out.

@DavidCarlisle sorry, that was unclear - i was refering to the question if it is helpful to change to LuaLatex. As is discussed with the others. But i'll try externalising the tikz figures now.

David Carlisle

@pogo-otter yes but change from what, which are you using now, pdftex or xetex?

wilx

12:12 PM

@PauloCereda What is the "Page rot" indicator?

Paulo Cereda

@wilx No idea, maybe if there are rotated pages?

David Carlisle

@wilx I'd like to say it's the number of pages that have been written for long enough to show ageing, but it's the number of rotated pages.

Paulo Cereda

@DavidCarlisle That's evil. :)

wilx

@DavidCarlisle Heh. I honestly did not think of rotated pages. It did evoke the rotting meaning for me. :D

David Carlisle

@pogo-otter impossible to say in general, if you are writing in Arabic or Chinese or something you should use xetex or luatex and opentype fonts, because while those kind of scripts have been made to work via 8bit fonts on (pdf)tex it's just massively painful. If you are writing in the latin script and have a lot of math, then probably pdftex is the engine of choice, unless you want to risk your thesis being typeset by a math typesetting system that's still in development and has some rough edges

egreg

12:24 PM

@DavidCarlisle Aren't you a linguist? :P

mirabilos

Morning

How can I make the content of a \multirow cell justify?

They’re left-aligned, unlike all other cells (environment is longtabu)

The answers to tex.stackexchange.com/q/315256/43807 don’t help because they make them all left-aligned ☹

yo'

@mirabilos ummmm not use multirow?

mirabilos

hmm

good point, but then I’d need to get the equivalent of a \cline through the middle of another cell

I basically have this, in ASCII art:

| foo  |  see below  |  bar baz bla |
+------+             |              |
|   bla bla bla bla  |  bla bla bla |

column 1 exists only on the first line, column 2 is deferred to the second line, column 3 has the long explanatory text which I currently use multirow for

if there’s another / a better way to split up the first two columns, I’m all ears

well eyes, but you get my point

use case, if someone wonders: long table of config items (key, default/example value, description), and all but two fit their default/example values into the narrow second column

the other two cannot be wrapped either (long LDAP search filter expressions)

yo'

12:44 PM

sorry lunch time

mirabilos

enjoy

pogo-otter

@DavidCarlisle Ah, thank you. That was helping to put the advantages of lua into perspective. My thesis will include a fair amount of math and literally no arabic and chinese. (though, some Sufi thoughts at the beginning of chapters would certainly improve the quality of the thesis... readability ... maybe not so much.)

barbara beeton

@DavidCarlisle -- yes, i know that. but as you know, \atop and \over were "disallowed" in amsmath by (among others) frank mittelbach, since they're primitives that latex was/is trying to discourage. that, at best, is a very mixed message.

pogo-otter

“Reflection is the lamp of the heart. If it departs, the heart will have no light.”
- quote for the fundamental chapter on wave propagation.

David Carlisle

@barbarabeeton yes sure disallowed in the document but obviously allowed in the coding as that's how amsmath genfrac is implemented. Are there any real examples where the implementation just using \atop and not \atopwithdelims makes a poor setting?

pogo-otter

12:50 PM

“Worry is itself an illness, since worry is an accusation against Divine Wisdom, a criticism of Divine Mercy.” - chapter on error propagation.

“I know you're tired but come, this is the way.” - conclusions.

David Carlisle

@egreg Kot bi morali vedeti, sem mojster vseh jezikih.

David Carlisle

1:01 PM

@barbarabeeton in particular since it's clearly broken as it is now I was wondering of we should ship amsmath.sty using the version of \genfrac that I suggest in my answer if xetex (or perhaps also luatex) is detected.

Joseph Wright

@DavidCarlisle Sounds good to me

yo'

@mirabilos I think your table is not quite a table; rather a particularly formatted list. But I still don't understand what you're trying to achieve, so I don't know...

@DavidCarlisle I'd say if unicode-math is detected, no? Because you can use the good old TeX fonts (like lmodern, mathpazo, eulervm) and there it works... Or am I wrong?

Joseph Wright

@yo' You know that's a bad idea, yes?

David Carlisle

@yo' oh perhaps yes (in which case Will could just put it in unicode-math and we don't need to touch amsmath)

yo'

@JosephWright my thesis does exactly this

Joseph Wright

1:05 PM

@DavidCarlisle Doesn't help plain or anyone loading by hand or ...

@yo' Hyphenation will be wrong :-(

yo'

@JosephWright hyphenation in math?!

David Carlisle

@JosephWright is "I don't care" a reasonable response?

mirabilos

@yo' “a particularly formatted list” is a fitting description, I think

David Carlisle

@JosephWright given crazy syntax of \atopwithdelims I wouldn't want to try to catch that at the macro level. So anyone using plain/primitive syntax is on their own there, they just need documentation to use \atop instead and add the delimiters.

Joseph Wright

@yo' Are you loading an OpenType text font? If so, hyphenation is OK but of course you've now got a font mismatch

yo'

1:08 PM

@mirabilos yeah, then you want some list environment, and not a table ;-) I don't say it's easy to setup, but it would be more correct, and maybe avoid your problems

Joseph Wright

@DavidCarlisle Of course

@DavidCarlisle Fair enough

@DavidCarlisle Great irony here is of course this makes no odds to you or me :-)

yo'

@JosephWright That's exactly what I do, yes. I wasn't willing to consider unicode-math reliable, so I opted out of it; but I needed way too many math alphabets for which lualatex seemed better; also one of the fonts wasn't available as TeX font.

David Carlisle

@JosephWright plain version:

\def\atopwithdelimiters#1#2{\message{Sorry I ate your delimters, if you need them put them back with \left and \right}\atop}

Joseph Wright

@DavidCarlisle :)

yo'

@DavidCarlisle something I was thinking about :-) I only wasn't sure if it's expandable...

David Carlisle

1:11 PM

@yo' \protected\def.... then it's safe enough.

daleif

1:27 PM

@DavidCarlisle that might explain the problems with the font I'm using. I would have preferred a more gradual growth, currently it just in very distinct jumps

David Carlisle

@daleif you can play some games without chnaging the font eg here I skip the 4 fixed size ones and jump straight to the expanding one, it allows smoother size changes but in cm at least the minimum size of the version made out of the expansion pieces is a bit big

\documentclass{article}
\DeclareMathRadical{\sqrtsignb}{largesymbols}{"74}{largesymbols}{"74}
\begin{document}

\[\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{X}}}}}}}}}}\]

\let\sqrtsign\sqrtsignb
\[\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{\sqrt{X}}}}}}}}}}\]

\end{document}

daleif

1:56 PM

@DavidCarlisle the problem is the big gap between normal and number two. I guess there is not much to do about it, blame it on the fonts. Thanks

mirabilos

2:12 PM

@yo' maybe, but that’s what I have right now, and it looks correct-ish, save for the text justification

% Just before "text" is expanded, the \multirowsetup macro is expanded to
% set up any special environment.  Initially, \multirowsetup contains just
% \raggedright.  It can be redefined with \renewcommand.

and indeed…
@@ -72,2 +72,3 @@
\usepackage{multirow}
+\renewcommand{\multirowsetup}{}
\usepackage{wrapfig}

voilà, works

egreg

2:32 PM

@DavidCarlisle I see you practiced your Slovenian during holidays

David Carlisle

@egreg ja

Paulo Cereda

@DavidCarlisle já in Portuguese is already.

David Carlisle

@egreg I would have practiced my Italian, but as we were only in Italy for 10 minutes it didn't seem worth it.

@PauloCereda Napačna vrsta naglas

Paulo Cereda

@DavidCarlisle Buongiorno, mi piace anatre!

@DavidCarlisle LOLOLOLOL

David Carlisle

@PauloCereda especially with orange sauce

Paulo Cereda

2:36 PM

@DavidCarlisle you are mean. :)

egreg

@PauloCereda Slovenian borrowed "yes" from German, contrary to other Slavic languages which have "da"; however, the Czech word for "yes" is "ano".

@PauloCereda "Mi piacciono le anatre" (the subject is plural): the verb "piacere" in Italian has a very peculiar usage.

Paulo Cereda

@egreg ooh and ano is year for us. :)

@egreg oooh

egreg

@PauloCereda It has a very different meaning in Italian

Paulo Cereda

@egreg I think I know what it is. :)

David Carlisle

@egreg I knew that of course but was too kind to correct him

yo'

2:39 PM

@egreg and Slovak áno or hej

@egreg has been confusing the world since the existence of Latin I think :-)

Paulo Cereda

@yo' ¡Ay chihuahua!

egreg

@yo' Which sounds like Italian "ah no", which is a strong form for negation,

yo'

@egreg see, we can't even agree on what ano is, how can we agree on anything else?!

Paulo Cereda

@yo' 2016! :)

mirabilos

2:55 PM

$ jmdict -j ano
彼の (あの, かの)
1) тот (предмет удалённый от всех участников беседы)
that (over there)
あのう, あの
1) ah ...
also ...
nun ...
errr ...
well
say
2 match(es) found.

the Japanese guys have their own ano, not derived from annum…

egreg

@yo' Change the termination and you'll know what your word for "yes" means in Italian.

Alan Munn

4:28 PM

17

Q: ABCjs integration now available

Update: ABCjs is now enabled on the main site as well. If there are any problems, please report them the usual way - in a new post tagged bug. Thanks, and enjoy! Y'all asked for ABCjs integration earlier, and I'm happy to announce that we have something ready to test out here. First of all, n...

discussion featured music-notation abcjs

Really cool!

@egreg Along with a bunch of other verbs like it. One of the linguistically fun parts of Italian.

Paulo Cereda

@AlanMunn cool!

egreg

@AlanMunn I expect greater interest in the abc package

Paulo Cereda

4:44 PM

@egreg I wonder who maintains it. :)

Alan Munn

@egreg I had no idea you maintained this.

egreg

@AlanMunn Without knowing anything about ABC and Mup

Alan Munn

@egreg :)

Paulo Cereda

Hey friends, I would like to know more about NTS, which was discontinued. Was Philip Taylor in charge?

egreg

5:00 PM

@PauloCereda Maybe you also want to know about ExTeX extex.org

Paulo Cereda

@egreg Huuum interesting. Looks like both initiatives were powered by Dante?

David Carlisle

@PauloCereda don't ask:-)

egreg

@PauloCereda I think so. ExTeX was quite interesting, as it allowed to easily add new features.

Paulo Cereda

@DavidCarlisle I'm lost. :)

@egreg Cool!

David Carlisle

@PauloCereda being fluent in German, I was at the Dante meeting that launched NTS :-)

egreg

5:06 PM

@DavidCarlisle I seem to remember a TUGboat article on the subject.

Paulo Cereda

@DavidCarlisle So you are the one to ask about it. :)

David Carlisle

home time...

Alan Munn

@PauloCereda Aber du musst ihn auf Deutsch fragen.

Paulo Cereda

@AlanMunn I cannot underztand ze text. :)

Alan Munn

@PauloCereda ఇక్కడ బహుభాషితాన్ని రాష్ట్ర భయంకరమైన ఉంది.

Paulo Cereda

5:15 PM

@AlanMunn o.O

Alan Munn

@PauloCereda You need to get out more. :)

Paulo Cereda

@AlanMunn That breaks my plan of originally not going out. :)

Alan Munn

@PauloCereda Oh right, I forgot you have a thesis to write.

Paulo Cereda

@AlanMunn Oh my, I forgot too!

yo'

> The requested URL /development/feature-requests.html was not found on this server.

David Carlisle

5:37 PM

@AlanMunn in der an der Dante Sitzung verwendete Sprache fragen

Silent

7:27 PM

I have a question about left-aligning different font sizes (ex: super big font, skip line, super small font). I notice the small space before the first letter on a line is bigger if the font is bigger. This makes it look like the two different font-sizes arent actually left-aligned. is there anyway to correct this other than using hspace to line them up?

David Carlisle

@Silent \parindent is normally a fixed length not affected by a font size change so it isn't clear what you are seeing. as always a working example helps (probably as a question on the question site:-)

Silent

here's an example: http://imgur.com/6D9lH2c
code: http://pastebin.com/EX0PvYvZ

im new to latex and finding it a bit hard to search for what im looking for (not sure what the right keywords are to solve my problem), and dont want to ask a question thats already solved but im misrepresenting

yo'

@Silent ah this; well, that's the font property, that whitespace is part of the letter

Silent

right

so i need to hspace it?

David Carlisle

@Silent off site links are not popular here but you have dozens of missing % at ends of lines so your macros are adding unwanted white space (which being word spaces is font dependent)

yo'

7:32 PM

@Silent first, try adding \usepackage{microtype}

David Carlisle

@yo' no first add 1000 % :-)

yo'

@DavidCarlisle yeah. Or as @egreg recommends, go expl3 yourself :-)

Silent

hm, so always add % at the end of the line?

David Carlisle

@Silent also your use of \raggedright is completely wrong, it does not take an argument in {} but as yiou have not shown an actual test document (from \documentclass to \end{document} hard to guess the intention of the code and correct it.

@Silent an end of line is a space, if you typeset:

yo'

@Silent Remember that }<endline> is moreorless equal to }<space> and this space produces an interword space, which may cause your troubles. Changing it to }%<endline> hides the endline inside a comment, solving this issue. If you then still have this problem, add microtype package as I suggest. If you still have this problem, opt in for \hspace*{-0.1ex} or alike.

Silent

7:35 PM

this is part of the cls file, im trying to modify a template to understand it better

David Carlisle

 one two
three four

Silent

true

David Carlisle

@Silent then it typesets as one two three four, so don't add a newline unless you want a space, or comment it out.

@Silent but the \\[] is in the wrong place and you probably want a \par before the vspace, and as I say the {} after \raggedright need to be removed. the second \raggedright can also be removed as it is still in the scope of the first so that text will be raggedright anyway.

@Silent :

204

Q: What is the use of percent signs (%) at the end of lines?

I see that the code in many packages and examples contains percent signs % at the end of (many) lines. What are they used for? Do they affect the parsing of those lines?

macros symbols sourcecode syntax comments

Silent

thanks, trying to wrap my head around this atm

percusse

7:52 PM

I was wondering today about having some conditional chain that got too long. Do we have a question about making TeX behave like other languages', say in Python.

I was wondering about if nonstop mode can be hacked to behave as such instead of itself trying to insert random chars, it does something given by the user

    try:
        some code
    except someting bad happened:
        do stuff

David Carlisle

tex.stackexchange.com/search?q=exception+handling

@percusse none of ^^^ really address that:-)

@percusse luatex perhaps otherwise no chance really. macro expansion and exception handling don't go well together.

percusse

@DavidCarlisle I've checked it but its context is different.

David Carlisle

@percusse yes

percusse

@DavidCarlisle ah ok. So no L3 magic then

David Carlisle

@percusse don't tell anyone but L3 is just a macro layer to hide your macros, so it doesn't really help if the problem is inherent in macro processing languages:-)

percusse

7:58 PM

@DavidCarlisle well should I tickle Bruno ? :-)

David Carlisle

@percusse ah well he'd invoke the tex is turing complete trump card and re-implement the entire macro processor in teX (actually he's done that already) but with a different error handler. doable, but a macro processor written in tex will be orders of magnitude slower than one written in c (or even pascal:-)

percusse

@DavidCarlisle By the time I understand the whole business I'll forget my original problem probably. Good that I'm out of PhD otherwise it would seem like a good idea

yo'

@percusse :-)

egreg

@percusse You might write a manual “How to finish your PhD thesis”; I'm sure you'll have at least one interested customer.

yo'

@egreg shhhhh he's probably just trying to find his keyboard so that he can type in the thesis...

percusse

8:04 PM

@egreg He just needs some duck tape here and there between articles

egreg

@percusse He has to finish writing his articles, then: I smell an infinite loop.

percusse

@egreg Well it is a CS thesis, he should know better :)

David Carlisle

Nobody ping @PauloCereda or he'll see this conversation.

Paulo Cereda

@egreg Where can I buy this material? It's for a friend...

percusse

@PauloCereda accepting the problem is halfway to <strike>oblivion</strike> graduation

egreg

8:10 PM

@PauloCereda :) Meanwhile, at the Venice film festival VVVVV

Paulo Cereda

@egreg OH MY

egreg

@PauloCereda After the ducks, on the red carpet appeared Rocco Siffredi

percusse

@egreg appear is a very peculiar choice of wording :D

egreg

@percusse He was wearing a tuxedo jacket (and trousers, too)

percusse

@egreg I must resist all the jokes that are rushing to my mind.

egreg

8:23 PM

@percusse He presented a documentary on his private and “artistic” life.

Paulo Cereda

@egreg o.O

yo'

8:40 PM

Speaking of ducks: met this last week while shopping:

yo'

8:56 PM

I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes. I won't get disappointed by stupid downvotes.

David Carlisle

@yo' I got one today as well:-) (hopefully I'll get rep cap to wipe it out but still:-)

yo'

@DavidCarlisle the downvoter should have waited until 23:59:59 for people like you :-p

David Carlisle

@yo' i'd have been more disappointed if i'd been on 333323 and I was pushed off 333333 :-)

yo'

@DavidCarlisle ah that's why you need the rep cap -- to keep yourself 3 mod 5 :-)

David Carlisle

@yo' yes of course, then a couple of downvotes would be handy once I pass 400k, to get back in sync

yo'

9:02 PM

@DavidCarlisle bad you can't offer a 99 bounty :-)

you should be capped now :-) (but don't tell green squares)

David Carlisle

@yo' ah just got +12 so back to *3 :-)

StrongBad

@DavidCarlisle so are you looking for 3 upvotes?

David Carlisle

@StrongBad no thanks, rep capped today so makes no difference:-)

StrongBad

@DavidCarlisle hopefully someone gets a screen shot for you if it happens while you are asleep.

David Carlisle

@StrongBad 3333333? I got that the other day:-)

yo'

9:08 PM

@DavidCarlisle seriously? :D

StrongBad

@DavidCarlisle doesn't 338,833 count also?

yo'

@StrongBad will be tomorrow morning I think unless he gets two accept votes today

David Carlisle

@StrongBad too easy:-)

egreg

9:22 PM

@yo' MS has strange ideas about downvoting.

yo'

@egreg well, that's why I made the comment in the first place; it's nothing to disappoint about and I simply wanted to count to ten :-)

yo'

9:36 PM

@DavidCarlisle Please, don't you have an idea whether there's a unicode symbol for "external link"?

Paulo Cereda

@yo' U+1F517? :)

yo'

@PauloCereda doesn't quite help in distinguishing internal and external links

Paulo Cereda

@yo' Oh.

David Carlisle

@yo' hmm:

9 matches for "<descrip.*link" in buffer: unicode.xml
    539:         <description unicode="1.1">DATA LINK ESCAPE</description>
  49956:         <description unicode="1.1">SYMBOL FOR DATA LINK ESCAPE</description>
  69872:         <description unicode="1.1">IDEOGRAPHIC ANNOTATION LINKING MARK</description>
 137917:         <description unicode="8.0">SIGNWRITING EYE BLINK SINGLE</description>
 137921:         <description unicode="8.0">SIGNWRITING EYE BLINK MULTIPLE</description>
 142926:         <description unicode="6.0">CLINKING BEER MUGS</description>

yo'

@PauloCereda what I have in mind is something like nos 6--8 here:

barbara beeton

9:42 PM

@yo' -- but i like the second one -- "let's get out of here!"

yo'

@barbarabeeton I don't think it quite delivers the message "this link brings you out of my website"

Paulo Cereda

@barbarabeeton "Free pizza next door"

yo'

I think a NE harpoon will do until I have a better solution: edu.tomh.eu/2016/09/seminar-z-matematiky-informace

btw, if you ever wondered, this is not me!!!

David Carlisle

@yo' wikipedia seems to use this applied via css to extenal links

ah data link vis svg, too log :-)

yo'

@DavidCarlisle it's not that easy to apply custom CSS to things in WordPress...

David Carlisle

9:53 PM

@yo' wpbeginner.com/plugins/…

says google :-)

yo'

@DavidCarlisle am lazy ... says I :-)

@DavidCarlisle it works ... somehow ... will have to look into it

cfr

@JosephWright Ir may take me a really long time. I think this question is basic, though. Does it make a difference when you use __ rather than _ in names? And how do you decide where to put the extras? I thought they marked internal (don't use?) but I'm not sure how to make sense of that here.

yo'

10:11 PM

@cfr I think I can answer that, at least to an extent

cfr

@yo' Yes ? (I only asked JW because it is his code.)

yo'

In most packages, whose interface is not in expl3, everything is internal. The point is, you have your public interface, let's say that in graphicx.sty it would be \includegraphics[foo-bar]{file} and this is only a wrapper around something like \__graphicx_includegraphics:nn {foo-bar}{file}

@cfr which package? (I thought you speak about your own code and deciding what's internal and what's not)

cfr

@yo' Neither ;)

6

A: Avoid @ hackery: How can I preserve spaces in a format specification when processing the specification with expl3 and PGF/TikZ?

There are really two parts to this question, the specifics of the use case and the more general 'how to iterate with spaces' one. I'll tackle the two separately. For the specific case here of replacing numerical part 'holders' with the actual values I would not use a loop at all. There's no pa...

The code at the bottom is what I was looking at.

@yo' For example

\cs_new:Npn \__chronos_print:w #1 - #2 - #3 - #4 \q_stop
  {
    \bool_if:nTF
      {
        \tl_if_blank_p:n {#1} ||
        \tl_if_blank_p:n {#2} ||
        \tl_if_blank_p:n {#3}
      }
      { \ERROR }
      { \chronos_print:nnn {#1} {#2} {#3} }
  }

Paulo Cereda

@cfr: Cwac!

yo'

@cfr clearly internal; you never want :w eird stuff in your public interface unless you have to

cfr

10:17 PM

@PauloCereda Cwac!

yo'

the point is: do you expect another package to need to call your _print:w function? If yes, it should be public, if not, it should be internal.

cfr

@yo' But why aren't they all internal, then? The public interface is using TikZ.

yo'

@cfr because if I were writing my class in expl3, I would call \chronos_set_dateformat:n and not \chronos@setdateformat, but I still wouldn't dare touching \__chronos_print:w

(but maybe @Joseph will be better at explaining...)

cfr

@yo' But \cs_new:Npn \__chronos_replace_Y:nnn #1#2#3 {#1} where there is no w ...

yo'

@cfr looks like some internal fu; I won't be calling "replace Y with something", I'll be calling "format this date for me" or "setup the format I want for me"

the site is down... :-(

cfr

10:27 PM

@yo' OK. Thanks. I don't really get it, I don't think. As far as I'm concerned, you probably shouldn't call anything from my code. But I think there's some concept here I don't get ... (Apart from the other concepts here I don't get). :-)

yo'

@cfr that's why I started with the non-fact that everything is private :-)

The problem is that your package has two types of users: (1) users and (2) other package authors. You shall probably provide an expl3 interface to all your functionalities so that when I write my package and rely on your package, I don't have to leave the expl3 world at all

the site is up again :-)

cfr

@yo' Yes but you shouldn't ever rely on code I write!

yo'

@cfr lol but that's a completely different matter :-)

@cfr I just hope you get the idea of internal vs. public at least a bit

cfr

@yo' I get the idea in other people's code but I don't understand it for mine. I don't understand how to decide or in what circumstances I even need to decide, I guess.

@yo' But thanks for your help. I'm sure it is not your explanations which are the problem, but my density.

yo'

@cfr nah the concept is not easy, really. One last thing that can hint you: the expl3 guidelines say that the public functions and variables have to be documented including their usage, possible values etc. So things for which you feel it makes sense to be in user documentation, these things should probably be public.

cfr

11:07 PM

@yo' Since I'm not actually dealing with package code here, this would be none of it :-). Thanks.

yo'

@cfr that's possible, yeah :-)

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends