TeX, LaTeX and Friends

12:36 AM

Can anyone think of a case where \lccode`\*=`\* would break something?

7:23 AM

Might it be that there is a duplication of the section about unicode soft hyphens in ltnews32 (page 2 left column, page 3 right column)?

Ulrike Fischer

7:38 AM

@TeXnician I will look.

TeXnician

7:56 AM

@UlrikeFischer Thanks. If you're at it, there is in the version in my TL also a testting and a |ini| (which should probably be verbatim and without the bars).

Ulrike Fischer

8:12 AM

@TeXnician thanks I have corrected this.

Joseph Wright

8:32 AM

@UlrikeFischer I'm reading over xcolor and colorspace to sort spot colors and multi-model support

David Carlisle

@HenriMenke break isn't the right word necessarily but it changes the hyphenation of wibble* (potentially)

Ulrike Fischer

@JosephWright ah good. it would be good if we could integrate spot colors - patching colorspace to avoid that it clashes with the page resources is quite a pain.

Joseph Wright

@UlrikeFischer That's definitely on the list: it's just ... tricky

@UlrikeFischer Probably first will be multi-model support, as that makes the spot color CMYK fallback easier

Ulrike Fischer

@JosephWright what is tricky?

Joseph Wright

@UlrikeFischer The mixing spot colors business: I need to get to grips with creating PDF color spaces

@UlrikeFischer I'll email you: I remember now what held me up :)

Ulrike Fischer

8:52 AM

@JosephWright I'm not sure if mixing makes sense with spot colors. I tried with colorspace \color{spot!50!red} and got a capacity exceeded error. And mixing two spotcolor gives in the pdf only one color.

cis

@AlexG Aha, and wath does this sign mean?

Joseph Wright

@UlrikeFischer It's logically achievable: the issue is that the current approach is all patching internals

@UlrikeFischer First I need to cover cmyk equivalence for spot colours, and that links to multi-model support

Ulrike Fischer

@JosephWright that should be not so difficult. But what does it mean in the pdf is you want to mix two spot colors? I will have to read the reference.

Joseph Wright

@UlrikeFischer On the multiple models: I'm thinking of using the xcolor syntax, which would allow { spot / cmyk } { Name / c,m,y,k } as well as { rgb / cmyk } { r,g,b / c,m,y,k}

@UlrikeFischer On mixing spot colors: you have to create a new color space that is that mix, then use it as the 'color' (that's what's tricky)

Ulrike Fischer

@JosephWright that is for defining the colors?

Joseph Wright

9:00 AM

@UlrikeFischer Yes, exactly

@UlrikeFischer Same as xcolor, basically

Ulrike Fischer

@JosephWright should we allow "additive" definition? so that if you define violet for rgb and later also for cmyk both are combined? Or should the second command remove the rgb settings?

Joseph Wright

@UlrikeFischer My initial thinking is that it's not additive: if you want to specify cmyk and rgb, you do it at the same time

Ulrike Fischer

@JosephWright yes I think it should be the default behaviour - in case two unrelated packages define colors it could have suprising effects if the definition were additive - but perhaps an "add_spec_to_color" would be useful. Then one can extend existing definitions.

Joseph Wright

@UlrikeFischer I'll get the basics set up then we can work on detail :)

AlexG

@cis Black Horse (jisho.org/search/%E9%A9%AA%20%23kanji)

daleif

9:20 AM

@yo' why does overleaf suggest \footnotesize{} when you start writing \footnote? That ends up leading users to think \footnotesize takes arguments.

cis

9:35 AM

@AlexG Ahhhaaa.

AlexG

@cis I didn't know the meaning either, until you asked. I took the original static SVG from github.com/KanjiVG/kanjivg/releases/download/r20160426/… It is the largest file in the ZIP.

cis

@AlexG Yes, I know that file. Seems like mouse-written in Inkscape. I wonder, why they do not create a font for the stroke-orders.

AlexG

@cis Fortunately they didn't. So I was able to animate the strokes using Ghostscript, Perl and animate.

cis

10:07 AM

@AlexG I want something like that (eventually in sf-font):

strokeorder.com.tw/%E9%A9%AA.html

Apparently they have broken down the sign of a font into individual parts.

AlexG

10:19 AM

@cis You want to start from a font file (OpenType, I guess), and decompose a glyph into its individual strokes and then animate the whole thing?

cis

10:29 AM

@AlexG Yes, I would like.

David Carlisle

@wtsnjp (@JosephWright @yo') for your information Overleaf just deployed an update for me and now the "open in Overleaf" button at davidcarlisle.github.io/ll/ja/language-01 should compile the project with platex :-)

Joseph Wright

@DavidCarlisle Wow

@DavidCarlisle Wow wow wow

David Carlisle

@JosephWright it's all coming together nicely:-)

Ulrike Fischer

@DavidCarlisle that is true. Really nice.

AlexG

@cis I see. Seems difficult, in particular filling the hollow shapes with bends in them. This may involve quite a bit of manual work.

daleif

10:50 AM

Anyone knows anything about the servers that overleaf uses for compilation. I have a user where if I compile on my system (using latexmk, thus multiple runs) it takes 30sek, on overleaf it times out. The user has a LOT of tables using siunix (and we are using parse-numbers=false)

David Carlisle

@daleif @yo' :-)

@JosephWright I could set up a stub ja directory (like the fr one with no translation at all) but with that ja/language-01 now it works on overleaf the test file at my ll copy is unstable as I break the javascript there more often....

Joseph Wright

@daleif Could you do me a favour and try it with v3? I've still to finish the code, but it would be handy to know if this kind of case is helped by the faster code

@DavidCarlisle Sounds good to me

@DavidCarlisle By the time we get to TUG2020, I might be able to 'launch'

Ulrike Fischer

@DavidCarlisle will it work with latexonline too?

David Carlisle

@UlrikeFischer it doesn't crash (but it doesn't run) I have asked for support for (u)platex but he's looking at the looping on firefox first.

Ulrike Fischer

@DavidCarlisle that is certainly more important.

daleif

11:14 AM

@JosephWright is that available on overleaf?

Joseph Wright

@daleif Not at present: I meant for your local test

daleif

@JosephWright URL for v3?

Joseph Wright

@daleif github.com/josephwright/siunitx

daleif

Perhaps luckily the user is on a Mac so I've asked them to intall MacTeX 2020 and run the thesis offline instead of on overleaf.

Joseph Wright

@daleif There's an alpha version on my blog: I'll find the link

daleif

11:17 AM

Having an appendix with 70 tables might not be the best choice.

Joseph Wright

@daleif github.com/josephwright/siunitx/releases/tag/v3.0.0-alpha.2

@daleif That will certainly push things

David Carlisle

learnlatex.org/ja/help

yo'

@DavidCarlisle It shouldn't be that bad.

Ulrike Fischer

@yo' I had this recently too, if you remember I asked you how one can trace what overleaf is doing.

yo'

@UlrikeFischer yeah. One way to see what Overleaf does is using a wrapper document and bashful. That's not optimal though, and slightly cumbersome.

daleif

11:31 AM

@JosephWright Hmm, it does not seem to like TL 2019

! Undefined control sequence.
<argument> \char_codepoint_to_bytes:n
{"03BC}
l.4395 ... \__siunitx_unit_non_latin:n { "03BC } }

or TL 2020

daleif

11:44 AM

Directly making a clone and compiling it from there seems to work

@JosephWright Some quick tests that I probably need to run several times. I ran a clean in between each test (latexmk -c *.tex) and is compiling this thesis using TL2019 using time latexmk -pdf file.tex

Siunix from TL2019 (parse-numbers enabled): 2min 6 secs

Siunitx from github otherwise same: 24secs

I'll see if TL20 makes a difference

TL2020 does not make much of a difference

TL2020, parse-numbers=false, normal siunitx: 24 secs

TL2020, github siunitx, parse-numbers=false: 17secs

cis

12:09 PM

@MarcelKrüger I dont't understand this with the 'COLR font'.

If I have

What have I to do? @UlrikeFischer

Joseph Wright

@daleif Ah, oh, forgot about that, yes

cis

1:13 PM

...
...
...
@Skillmonlikestopanswers.xyz Whats new with the Kana/Kanji/quak/github-parser?

I recommend the following syntax:

or short version: \kanakanji{...............}

Skillmon likes topanswers.xyz

1:32 PM

@cis I'm against an environment, the idea shouldn't be to input large parts of a document but to provide a quick solution to input a small amount of Japanese into a document. For other things there are already other solutions to which I can't compete.

wtsnjp

1:42 PM

@DavidCarlisle Wow, that is quick! I could confirm that it works nicely :)

cis

1:56 PM

@Skillmonlikestopanswers.xyz The main thing was the syntax for Kanjis:

`{<unicodenumber or character>}[<options>]`

between some Kanas.

I thought this is easier, instead each time `\kanji[<options>]{<unicodenumber or character>}`

David Carlisle

@wtsnjp i'm hoping to get the latexonline one working as well but it's a free service supported in his spare time so I can't push too hard:-)

@wtsnjp could you check my google translate japanese here learnlatex.org/ja/help or suggest any other trivial text at that point, also the buttons would translate if you could supply text for:

var buttons ={
    "edit":             "edit",
    "copy":             "copy",
    "Open in Overleaf": "Open in Overleaf",
    "LaTeX.Online":     "Latex.Online",
    "Delete Output":    "Delete Output"
}

eg for de I have var buttons ={
"edit": "Bearbeiten",
"copy": "Kopieren",
"Open in Overleaf": "In Overleaf öffnen",
"LaTeX.Online": "LaTeX.Online",
"Delete Output": "Fenster schließen"
}

Ulrike Fischer

2:21 PM

@DavidCarlisle why isn't there a "learnlatex.org" title on the main english version start page? All the language version have it.

Skillmon likes topanswers.xyz

@cis I thought that maybe we could even provide different input models for {}, which can be changed with a prefix, e.g., {u:65e9} or with the list entry points of the kana table (forgot the name), or or or. I'm not sure about the options, that's something you'd need to specify what you'd need. I can code them.

David Carlisle

@UlrikeFischer the original default gh template puts the site title on every page except the top level page where that is already the page title, but that logic doesn't really work now I added the translations, I could change to add it always

cis

2:54 PM

@Skillmonlikestopanswers.xyz

Ok, I do not care whether `{u:65e9}` or `{65e9}`; I think the second one is easier and not to be confused with Kanas `kakitu KAKIKU`....

It would be nice to have a few optional options
`{65E9} [<options>]`
such as

text = red,

font = \Huge,

ruby = {<text above>},

ruby * = {<text below>},

or

translit = true | false
(that should also work, right?)

Skillmon likes topanswers.xyz

@cis one could have a default (e.g., defaulting to Unicode point) and only need a prefix for other input methods if we provide them.

@cis why ruby as text above and ruby* as text below?

cis

@Skillmonlikestopanswers.xyz I dont know, I thought the call oversetted text 'ruby' (?)

The Anglo Saxons call it so...

@Skillmonlikestopanswers.xyz Or what do you mean?

Skillmon likes topanswers.xyz

@cis I didn't know this...

@cis yes, I meant why exactly the key word ruby. It was very unintuitive for me.

Paulo Cereda

quack

Skillmon likes topanswers.xyz

@PauloCereda is now the name of a package project.

David Carlisle

3:04 PM

@Skillmonlikestopanswers.xyz w3.org/International/articles/ruby/markup.en.html

Paulo Cereda

@Skillmonlikestopanswers.xyz ooh

Skillmon likes topanswers.xyz

@DavidCarlisle funny that they use a word that can't be directly transcribed to Japanese... :)

cis

@Skillmonlikestopanswers.xyz @Skillmonlikestopanswers.xyz @PauloCereda I would like to emphasize that I voted for a factual name like 'kanakanji.sty'.

Paulo Cereda

@cis quack is definitely better

David Carlisle

@PauloCereda although it makes you feel hungry

Paulo Cereda

3:06 PM

@DavidCarlisle oh no

Skillmon likes topanswers.xyz

@PauloCereda @cis QUick And Casual Kana seems like it's pretty factual, imho.

cis

@PauloCereda I will use \renamepackage{}{} .... :()

Paulo Cereda

@Skillmonlikestopanswers.xyz exactly

cis

@Skillmonlikestopanswers.xyz See texdoc ruby too....

Paulo Cereda

@cis suit yourself, a rose by any other name would smell as sweet, says an English bloke.

David Carlisle

3:10 PM

@PauloCereda a ruby red rose, no doubt

Paulo Cereda

@DavidCarlisle ooh xcolor

David Carlisle

@UlrikeFischer three in one learnlatex.org/ja/language-01

cis

@Skillmonlikestopanswers.xyz Aahaaaa: en.wikipedia.org/wiki/Ruby_character

Ulrike Fischer

@DavidCarlisle ;-). jsarticle doesn't like lualatex ;-(.

David Carlisle

@UlrikeFischer grr I had a working version but messed up...

Ulrike Fischer

3:18 PM

@DavidCarlisle I thought it is the class. Does it have an option to run with lualatex?

David Carlisle

@UlrikeFischer no I didn't the last example uses article (and luatexja)

Ulrike Fischer

@DavidCarlisle yes that worked fine. I only tried what happens if I switch the engine in the other examples - I like to break things.

David Carlisle

@UlrikeFischer ここでは歓迎されません

Ulrike Fischer

@DavidCarlisle しかし、あなたはいつも私を責めることができます

David Carlisle

@UlrikeFischer I do

barbara beeton

3:36 PM

@Skillmonlikestopanswers.xyz -- I became familiar with this term, "ruby", when I was active on an ISO working group (yhe group responsible for SGML), with a number of active Japanese members. I don't understand that choice of terminology either, but it's been around for a long time. "Ruby" is also a traditional small type size. (That always struck me as rather Victorian.) That may possibly be related.

David Carlisle

3:51 PM

@barbarabeeton according to wikipedia that is exactly the reason, it was originally the specification of the smaller font size and came to be used for the actual annotation itself

cis

5:30 PM

@Skillmonlikestopanswers.xyz And is there "a github" now for quak.sty?

Paulo Cereda

6:07 PM

@DavidCarlisle ^^ we need a German taskforce

cis

@PauloCereda Yes, it doesn’t matter, the Japanese don’t set spaces and pronounce entire sentences as if they were a single word:

@PauloCereda What are the tasks of the German task force?

Skillmon likes topanswers.xyz

6:30 PM

@cis yes, I already invited cis555 (which you said was your user name on github).

@PauloCereda looks like a reasonable language, not like that space infested abomination called English.

Faheem Mitha

@PauloCereda Print statements are effective.

Paulo Cereda

@Skillmonlikestopanswers.xyz :)

cis

@Skillmonlikestopanswers.xyz Aha, maybe I don't check this page...

@PauloCereda So learn all about the German language
---------> https://www.youtube.com/watch?v=K3IImGiiY1Q
<---------------- Good explaination.

The German Language

►

@Skillmonlikestopanswers.xyz On the other hand: as exaggerated as the github side deals with their security measures (10 digit PW, email verification with every login, ...) I would not be surprised if there was an invitation for 5 minutes and then never again.

cis

7:17 PM

@Skillmonlikestopanswers.xyz Mybe you mean the 'List of jōyō kanji'

List of jōyō kanji

The jōyō kanji system of representing written Japanese consists of 2,136 characters. == List of characters == For brevity, only one English translation is given per kanji. The "Grade" column specifies the grade in which the kanji is taught in Elementary schools in Japan. Grade "S" means that it is taught in secondary school. The list is sorted by Japanese reading (on'yomi in katakana, then kun'yomi in hiragana), in accordance with the ordering in the official Jōyō table. This list does not include characters that were present in older versions of the list but have since been removed (勺, 銑, 脹, 錘...

I do not know, whether this is a official numeration in the first column.
I think in unicode-numbers. ---> https://en.wikipedia.org/wiki/List_of_CJK_Unified_Ideographs,_part_1_of_4
But why not to use that number too.

David Carlisle

8:46 PM

@barbarabeeton tex.stackexchange.com/posts/85115/revisions

@PauloCereda good idea: it could meet at Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch

barbara beeton

@DavidCarlisle -- aw, gee. (I had already upvoted it, so I can't do that again.) Glad to see the improvement.

David Carlisle

@barbarabeeton someone left a comment (which I didn't understand) on that answer just now so I happened to see both the math italic and your comment:-)

Ulrike Fischer

latest palindrom: 33th wedding anniversary ;-)

8

David Carlisle

@UlrikeFischer congradulations

barbara beeton

@UlrikeFischer -- from me too.

Ulrike Fischer

8:57 PM

@DavidCarlisle @barbarabeeton thanks - and happily we could go to a restaurant and have a nice dinner ...

David Carlisle

@UlrikeFischer no chance of that here for some time yet

barbara beeton

@DavidCarlisle -- Yes, that is indeed a puzzling comment. There's an answer further down that has a similar answer, but loads mathtools and the two arrows are different lengths. So that's a serious difference between amsmath and mathtools. Might want to look into it. (@daleif -- tex.stackexchange.com/a/103994 .) Why not add those "missing" forms?

egreg

@barbarabeeton The definition in mathtools is wrong, because it has 0055 instead of 0359

David Carlisle

@barbarabeeton well it's not really a difference other than the usual one that mathtools has a command not in amsmath, but as @egreg says the mathtools one looks a bit wrong, it's in a latex3 github these days though...

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends