« first day (3491 days earlier)      last day (1433 days later) » 

cis
5:21 AM
@Skillmonlikestopanswers.xyz No, but I can create a github-account.
Surely.
--------------> I am cis555 there....
...
...
...
I have changed an existing font with Fontforge.
(All I've done so far is to delete the B and the C.)
I used the following code to call up the font.

What I should know:
Can I use the Unicode number instead of typing B or C, e.g. `\UTF{0042}` in that case.

I can also ask a question in the forum, but then I probably have to upload the ttf file.
% arara: lualatex
    \documentclass[a4paper]{article}
    \usepackage{fontspec}
    \begin{document}
    Test:
    \newfontfamily{\myfont}[]{test04.ttf}
    ABCD\dots, but \myfont ABCDE
    \end{document}
It is better, if I ask in the forum:
0
Q: How to call up characters of a font by their unicodenumber

cis I have changed an existing font with Fontforge. (All I've done so far is to delete the B and the C.) I used the following code to call up the font. What I should know: Can I use the Unicode number instead of typing B or C, e.g. \UTF{0042} in that case. If necessary, I can also upload my ttf ...

 
 
1 hour later…
cis
7:02 AM
@UlrikeFischer Meine Sache! ( https://tex.stackexchange.com/questions/545603/how-to-call-up-characters-of-a-font-by-their-unicodenumber )
I think, this can't be translatet in English: "my thing" does not have the great clout.... :()
 
7:17 AM
@cis well either it is relevant then you need to explain it or it is not then why mention it at all and confuse people?
 
cis
@UlrikeFischer I am just creating a new font and I use some of the existing ones.
 
7:41 AM
Is pdflatex's startup time getting slower and slower with each year? I have a Mathematica package that can compile small LaTeX snippets and import them to Mathematica. I am looking at the documentation example, which show the output of commands at the time when they were originally run, some years ago. The timings have more than doubled. I wonder why.
 
8:07 AM
@HenriMenke It's coursework/homework for some LaTeX-course somewhere. I don't know more than that.
 
cis
8:24 AM
....
....
I have a table here in Firefox:
https://de.wikipedia.org/wiki/Unicodeblock_Hiragana

How can I find out which (ttf?) font was used for the Japanese signs?
 
@JosephWright vv
@MightyTyGuy I think the correct fix is should be in beamer over which I have no control but I could add a guard to colortbl's hhline I guess (but really hhline should have no reason to expect to need a guard there, I think) — David Carlisle 45 secs ago
@cis on my system it's Microsoft YaHei just right click on it select inspect element and then look in the fonts tab of the inspector
@Szabolcs check that you don't have directories that are being searched that are not cached in ls-r files, as that can slow things down a lot if they get big. There is UTF-8 processing in latex now but that's not much overhead if your text is ascii, and if your text is non ascii utf-8 you need it anyway
 
cis
@DavidCarlisle Ahhhyaa.... thx.
 
9:05 AM
quack
 
@PauloCereda breakfast
 
@DavidCarlisle oh no
 
10:06 AM
@yo' メールを送信 Overleaf
 
10:27 AM
@JosephWright davidcarlisle.github.io/ll/en/language-01 almost works (you just need to remove the auto-added comment)
 
@DavidCarlisle Wow
 
@JosephWright I put in a support call asking they only add that comment to .tex files or special case latexmkrc and use # rather than % for the comment
 
10:43 AM
@DavidCarlisle and no arara support... :(
 
@PauloCereda it's bad enough writing html with embedded liquid templating that generates javascript that adds URL encoded perl to customise latexmk to call platex instead of latex without throwing arara rules in there as well:-)
 
cis
...
...
@Skillmonlikestopanswers.xyz What do you say now? Is this good or is this good?
% arara: lualatex
\documentclass[a4paper]{article}
\usepackage{fontspec}
\usepackage{pgffor}
\usepackage{xintbinhex}

\newfontfamily{\myfont}[Scale=1]{Split-IPAGothic.ttf}
\begin{document}
\section{Stroke-Order-Font Test}
\Huge
\foreach \n in {0,...,106}{%
{\myfont\symbol{"\xintDecToHex{\n}}}
}%
\end{document}
I don't like this font here "IPAGothic.ttf" yet.
I like "MS YaHei.ttf" better.

But now that I know how Fontforge works, it can be used for other fonts too.

I wanted to add some little arrows or something like that with TikZ.
 
yo'
@DavidCarlisle ?
@DavidCarlisle Ah I see now. Interesting...
 
10:58 AM
@yo' :-)
 
yo'
@DavidCarlisle nice feature, isn't it? :-)
Will also work very well with dtx :-)
 
@yo' yes I think just adding it to .tex should be safe enough
 
yo'
@DavidCarlisle and it's also the sensible way to go really
@PauloCereda oh no! We won't ever meet again then! :-(
 
@yo' Don't worry about these signs. :)
 
yo'
11:19 AM
@PauloCereda It's tempting...
 
@yo' hey sign me in. :)
@yo' let's have pizza when I visit you in Praha.
 
yo'
@PauloCereda oh well we can have much better things here :-)
 
@yo' ooh two pizzas
 
yo'
goulash with dumplings, fried cheese, home-made waffles, ...
@PauloCereda or a large pizza!
 
@yo' yay!
@yo' we could have those things as toppings of a large pizza
:)
 
yo'
11:23 AM
@PauloCereda :)
 
An answer to a question, almost 9 years later.
Thank you for the diagnosis. It sounds like a bug report is warranted. Can you submit one? Though, for a package that was apparently last updated a long time ago, it's not clear if that is an option. Could a third party fix it? — Faheem Mitha 1 min ago
Would it be possible for someone to fix soul?
 
@cis interesting. I haven't cared much for font choices for Kana, yet.
 
Hi all, I have a question that one of you might be able to help with.
I found a bug in the soul package that should be easy to fix, but I don't know who to contact about this.
There's no contact information in the documentation as far as I could tell.
@FaheemMitha Oops, I didn't see you were also here.
 
11:42 AM
@Circumscribe I mentioned that I posted on Chat. :-)
Like I said, it would probably be possible for someone to apply a fix, assuming that TeX Live is all under version control now.
 
cis
@Skillmonlikestopanswers.xyz I assume that the choice of font in the package will be an option.
The parser then looks for the correct Unicode character. ;)

I just don't want to create the stroke-order as a png, not as an svg, but as a correct font. A font, of course, which I can then use in a package.

Why do I want this at all? I plan to create Genko-Yoshi-like worksheets. But I'm not there yet.
 
@BambOo Oh sorry, forgot revisiting that question
Will do in some days
I remember my first tests having shown good success with your solution, though.
@DavidCarlisle Today I found out I cannot use an itemize environment with tabular, but tabularx.
So even if you don't need X columns, there might be a reason to use tabularx.
 
@ComFreek You can in a p-column.
 
And m columns?
 
11:57 AM
@ComFreek Yes.
 
Well then it doesn't work
It fails with lots of error messages
I'll rather stick with tabularx then
 
@ComFreek No it doesn't:
\documentclass{article}
\usepackage{array}
\begin{document}
\begin{tabular}{m{5cm}}
\begin{itemize}
\item foo
\item bar
\end{itemize}
\end{tabular}
\end{document}
 
Then it's specific to my file.
 
@cis The parser currently doesn't link against Unicode, but against the font table of one specific font. I intend to recreate this a bit more versatile, with the option to link against Unicode characters (which then need to be handled correctly by something else, e.g. inputenc for pdfTeX, or the Unicode engines for which the user specified an appropriate font. For that there will be different symbol tables and the parser links against such tables (so I add an internal abstraction layer).
@ComFreek then try to reproduce it in an MWE and ask a question (or find the error while you're creating the MWE).
 
Not sure what advantage I have from doing that :)
It's typesetting fine, just with tabularx.
How do I best write "3-COL" in uppercase font without a big space appearing surrounding the minus sign?
 
12:05 PM
@ComFreek understanding (La)TeX better.
@ComFreek 3-COL?
 
yes
 
@ComFreek Have you tried it?
 
@ComFreek in math mode?
 
okay \textsf{3-COL} looks good
 
12:07 PM
@ComFreek why should you write that in math mode?
 
Because you usually write "G \in 3-COL"
for a graph G
 
@ComFreek what's 3-COL supposed to mean?
 
It's the set of all 3-colorable graphs
It's complexity theory in CS.
3-COL is the agreed upon convention for decades
Perhaps you know SAT - that one is better known
As in SAT solver
 
@ComFreek ok (not my field, so I had no idea :)
I think it's a strange name to use as a variable/set in equations...
 
Actually, there are too many such sets to reserve special symbols for them
Ah, the halting problem is such a set and conveniently denoted by H.
 
12:13 PM
@ComFreek I'd define a symbol for it at the beginning of my thesis/paper and use that one... :)
 
@Circumscribe Thank you for the fix. i hope someone can apply it. You did test it, right?
 
hello, I have a question here, should I use amsmath or mathtools with amssymb ?
I mean, are there any drawbacks to either approach ?
 
I use all three oO
 
@FaheemMitha I did, but not extensively.
 
does mathtools not replace ams maths (except for symb and thm) ?
 
12:22 PM
hey
 
@T.D. mathtools is an extension of amsmath. It already loads amsmath so you don't need to load both (but it does no harm).
 
@T.D. mathtools loads amsmath as its basic and only builds upon it.
 
@Skillmonlikestopanswers.xyz youtube.com/watch?v=--moiivJ6VY
 
@Circumscribe @FaheemMitha well basically you have the options: - find the maintainer and convince him to fix the package - fix the package and convince ctan to take you at the new maintainer, see ctan.org/help/become-author, - fix the package under a new name, eg. new-soul, and upload it to ctan. Whatever you do: there is another long standing bug in soul regarding xcolor which would need a fix too.
 
@T.D. (Just loading mathtools sort of hides the fact that amsmath is being used, so loading both explicitly actually seems preferable.)
 
12:35 PM
@PauloCereda impressive.
@UlrikeFischer imho, the complete approach of soul is inherently bugged. It is impressive and there is not much one could do better, but it adds so many pitfalls to it that I'd never use it personally. It's just too unstable, imho.
 
@UlrikeFischer Hm, since soul seems quite buggy I'm not so sure how I'd feel about becoming its maintainer. :(
@UlrikeFischer How about option 4: convince someone else to become the new maintainer and ask them to fix it? :P
 
@Circumscribe Ok.
@UlrikeFischer Is it possible to apply a fix and do a new upload without becoming the maintainer? Akin to Debian's NMU.
 
@Skillmonlikestopanswers.xyz I'm not using it either, I would use lualatex and the lua-ul package, but it is used quite often so getting the obvious bugs fixed would be too bad.
 
@Circumscribe I see we are having conversations in parallel. :-)
 
@FaheemMitha sometimes it is possible, contact ctan and discuss this with them. But even for a one-time upload you will have to create a complete zip, you can't upload simply the sty.
 
12:48 PM
@Skillmonlikestopanswers.xyz 6 questions about soul this year. So there are still people using it.
If it so buggy, shouldn't it be depreciated, or something? With a view to removing it eventually?
 
@Circumscribe the other bug is this one: tex.stackexchange.com/a/48502/2388
 
@UlrikeFischer Anything else that needs fixing?
 
@FaheemMitha I remember only this.
 
@UlrikeFischer Ok.
@Circumscribe Are you up for a one-time upload? Yes/no/maybe.
 
cis
@Skillmonlikestopanswers.xyz Cool
 
12:55 PM
@Circumscribe BTW, whereabouts is your location? I don't recall seeing you in chat before.
 
@FaheemMitha Yes/maybe
 
@FaheemMitha yes, we found that too some time ago, but if you look at the issues you see that David didn't get any answer. There is also a postal address in the dtx so you could try with a real letter.
 
@FaheemMitha I'm in the Netherlands. I've been in chat a couple of times, but not often.
 
@Circumscribe That's a bit blurry. Do you use soul yourself?
 
@FaheemMitha I'll create an issue there.
 
1:05 PM
@Circumscribe I guess you're doing relatively well with the virus situation.
 
@Circumscribe davidcarlisle/soul-fixed
 
@Circumscribe I've no idea if that is an official repo, so check first.
 
@PauloCereda A URL would be better.
 
@FaheemMitha well, it does not exist, it meant just as a joke. :)
 
1:07 PM
@PauloCereda Oh, right.
 
@FaheemMitha I don't use soul myself. I answered your question because a new question about the same bug was posted recently.
 
@Circumscribe Oh. Huh.
The two questions should probably be linked or merged, or something. Close as dupe?
 
@FaheemMitha: I actually found four duplicates of your question (of which one was on stackoverflow.)
 
@Circumscribe LOL
@Circumscribe I can offer a small amount of assistance, though I'm a very ignorant TeX user. If you need someone to assist, that is.
 
@FaheemMitha I did that. You'll see three of the questions linked to from yours were marked as duplicates yesterday :).
 
1:10 PM
Sounds like it should be fixed. Probably someone should trawl through existing issues and try to fix as many outstanding things as possible.
 
@ComFreek no that is wrong.
 
@Circumscribe Oh. Is mine the oldest then?
Anyway, AFK for a bit. @Circumscribe, you can generally reach me here. On you can alternatively ping me on that question.
 
@FaheemMitha Yes
@FaheemMitha Alright
@FaheemMitha We seem to be doing okay
 
@Circumscribe actual physical address and email are in the file, no idea whether either is still valid (@FaheemMitha)
 
@UlrikeFischer Sending a physical letter about a software problem is an intriguing idea. :)
 
1:16 PM
@ComFreek never use tabularx without an X column
 
@DavidCarlisle I'll try the email address.
 
@FaheemMitha nothing in a LaTeX installation ever gets deprecated and removed it seems. Just look at the vast number of package. All those templates for thesis at a single chair or institute. None of them is of any use for 99.99% of the users of TeX Live. Still it gets included, because someone thought it's a good idea to publish it on CTAN, instead of providing it on a different route so that the default installation doesn't get unnecessarily bloated :(
 
@Skillmonlikestopanswers.xyz I wish ctan/texlive would refuse to publish one-department thesis classes, it's really an abuse of the system, trouble is everyone claims it isn't really a one-department class but a template for how a thesis class should be done and esily configured for other institutions.
 
@DavidCarlisle time for a new version.
 
@Skillmonlikestopanswers.xyz of tx, what would it do?
 
1:23 PM
@DavidCarlisle that's the reason why mine isn't published on CTAN, even though it contains code that could benefit others as well. At some point I might consider forking those parts out of it and publish them.
@DavidCarlisle throw an error if there was no X column :)
@DavidCarlisle providing a second environment tabularx* that doesn't.
 
@Skillmonlikestopanswers.xyz it would break 99% of all existing uses probably:-(
 
@DavidCarlisle I never used it without X.
 
@Skillmonlikestopanswers.xyz although it does start by counting how many X there are to get a guess of the width so I could warn if it starts off at 0
 
@DavidCarlisle you really should (the error was merely a joke, but a warning seems appropriate)
 
@Skillmonlikestopanswers.xyz it literally just calls tabular* anyway :-)
 
1:26 PM
@DavidCarlisle yes, but only after doing its job, sooo.
 
@Skillmonlikestopanswers.xyz I could remove some guards and let the users get a divide by zero error. would serve them right
 
@DavidCarlisle that seems not like a good idea (cryptic errors for things you can easily catch at the cost of ~0 time are never a good idea)
 
@Skillmonlikestopanswers.xyz you're no fun
 
@DavidCarlisle oh, I am. Most of the time (look at the errors and warnings in ducksay), but considering the user base size of tabularx and the fact that you actually check for this error currently and at some point in time were as unconsidering as not throwing an error, you should throw a warning.
 
“Address not found
Your message wasn't delivered to a8603365@unet.univie.ac.at because the address couldn't be found or is unable to receive email.”
 
1:30 PM
@Circumscribe that's hardly surprising it looks like a "this year's class" throw away student email address
 
@DavidCarlisle yeah, I tried anyway
 
@DavidCarlisle well, my only guaranteed to function E-Mail address at the university is similarly cryptic, but won't change.
 
@Circumscribe this looks more likely members.aon.at/mfranz/terminal
 
@DavidCarlisle I tried it and it appears to have been delivered successfully (by which I just mean that it wasn't bounced).
 
1:47 PM
@Circumscribe google will find most people, in the end:-)
 
@DavidCarlisle at some time we probably will need the contact too for crop.sty.
@StefanKottwitz I get again a certificate error from texwelt.de
 
@Circumscribe let us know if it works, then @UlrikeFischer can bug him about crop.sty :-)
 
@DavidCarlisle Will do!
 
@UlrikeFischer working on the sites
@UlrikeFischer fixed
 
2:03 PM
@StefanKottwitz works now again.
 
thanks!
 
2:17 PM
@JosephWright and @DavidCarlisle: I focussed so much on tabular for lesson 8, that I totally forgot about the table environment... My suggestion would be (as L08 is pretty long already) to move L07 after L08 and including the table environment there (as the title is "[...] and making things float").
@cis I just noticed that the stroke order of "wo" is missing a strike in your table... The step from the second to the third isn't correct, because there was one "flick" missing.
 
2:36 PM
@DavidCarlisle Why shouldn't I use tabularx without an X column?
(Seems like a design flaw of tabularx.)
 
@ComFreek tabularx does quite some work to measure the tabular to calculate the width of X, that is rather pointless if there is no X isn't it?
 
@ComFreek er why? The point of tabularx is (a) give it a target width and (b) specify some columns whose width can be varied to achieve the target width. If you specify a width but specify no X columns then the code has absolutely no way to achieve the goal and can do nothing useful. It is user error not a design flaw.
 
I feel like this goes against the spirit of almost all utility functions in programming, say, in functional programming the map function. It also simply does nothing on empty lists.
But apparently tabularx still does more if it doesn't work with tabular in my case.
 
@ComFreek no it is like saying that accessing the third element of an array should do something if the array has no values. what do you want it to do?
@ComFreek of course it works with tabular, tabularx just calls tabular anyway. You are presumably using a c column, that wouldn't work in tabularx either if you put a list in it.
 
Well, I think that analogy is far-stretched. To arr[3] with len(arr)=0 you cannot give useful semantics, but to tabularx without X, you can give a canonical semantics: just the one from tabular.
 
2:44 PM
@ComFreek no what about the first argument, which is the main point of tabularx that specifies the width?
 
@ComFreek m columns in tabularx and m columns in array package are the same code so if it works in one it will work in the other
 
well it doesn't compile with tabular
The file is self-contained, you can try yourself.
 
@ComFreek then you did something wrong
@ComFreek it works as expected. \begin{tabular}{m{4cm}|m{6cm}|m{6cm}|m{6cm}|m{6cm}|m{6cm}} and \end{tabular}
 
oh
I guess I must have left the width argument of tabularx in it
Apologies
 
2:48 PM
@ComFreek I do know how my own code works, most of the time:)
4
 
Hah, I wish I could say this too about my code
 
@DavidCarlisle ooh
 
3:12 PM
@Skillmonlikestopanswers.xyz I'm surprised there is no depreciation system at all.
Debian, for example, is constantly removing packages, yes the number of packages continue to grow steadily.
 
@FaheemMitha things get deprecated eventually (eg latex2.09 isn't built by default any more) but it's hard. If you deprecate say color in favour of xcolor what would you actually do? are you volunteering to go through every document in arxiv and update it?
 
@DavidCarlisle No, of course not. But a reasonable first step would to change the package to issue a warning to switch to a different package, or whatever.
 
@FaheemMitha programs come and go but documents live for thousands of years and latex has to be stable enough to support that.
 
@DavidCarlisle Thousands of years?
 
@FaheemMitha well it's not reasonable if getting several million warnings scares the arxiv maintainers from updating to the next texlive.
 
3:17 PM
@DavidCarlisle If it has that effect, it would indeed be unfortunate.
 
@FaheemMitha there are bits of papyrus lying around that old
@FaheemMitha well why do you think people are always moaning on the site that arxiv is an old texlive? updating for them is a massive task and they do re-run their collection
 
It's nice to imagine that TeX would survive for thousands of years, but the smart money says the human race doesn't have long to go. 100 to 200 years tops. And that might be too optimistic. Unless Elon Musk manages to colonize Mars first, of course.
@DavidCarlisle What, re-run all the files on their system?
 
@FaheemMitha I haven't seen any money smart or not betting on that timescale:-)
@FaheemMitha yes
 
@DavidCarlisle It's an expression. But if good options existed for getting off this planet, I think there would be some interest.
@DavidCarlisle That sounds like a terrifyingly daunting task.
 
@FaheemMitha I don't actually know if it's all or whether it's a representative sample, but either way some human would get some thousands or millions of deprecation warnings and have to decide to ignore them all, so what's the point?
 
3:24 PM
@DavidCarlisle quack
 
@FaheemMitha there is only a need to warn about deprecation if there is a realistic chance of something being removed
 
@DavidCarlisle Well, I suppose it's unlikely that whoever it is would try to fix those warnings. But perhaps new users of those packages won't ignore the warning. So that would be the point.
 
cis
is Yes, I will fix that.
But I think I will make a 2nd version new with a nicer font tomorow.
 
@DavidCarlisle Yes, of course.
 
cis
 
3:25 PM
@FaheemMitha I guess you don't do much tex support:-) the chance of users not ignoring warnings is vanishingly small, Just getting them not to ignore errors is a losing battle.
@PauloCereda dinner
 
@DavidCarlisle oh no
 
@cis looks way better :)
@DavidCarlisle especially with convenient IDEs which ignore errors and show you some built PDF...
 
@DavidCarlisle with the exception of the people asking why they get an error and then show some warning in the log ;-)
 
@DavidCarlisle I admit I do ignore overflow warnings, and have done so for 25 years. Thogh sometimes I feel bad about it. But I do pay attention to other warnings and try to fix them.
And I haven't done much TeX support at all. Very little, actually.
 
@UlrikeFischer mind you it's hard to tell people to read all warnings when someone fills it up with chatty user-facing messages like:
Lua module: luaotfload-loaders 2020-05-06 3.14 luaotfload submodule / callback h
andling
Inserting `luaotfload.define_font' at position 1 in `define_font'.
Lua module: luaotfload-database 2020-05-06 3.14 luaotfload submodule / database
Lua module: luaotfload-unicode 2020-05-06 3.14 luaotfload submodule / Unicode he
lpers
Lua module: luaotfload-colors 2020-05-06 3.14 luaotfload submodule / color
Lua-only attribute luaotfload_color_attribute = 4
Lua module: luaotfload-resolvers 2020-05-06 3.14 luaotfload submodule / resolver
 
3:38 PM
LuaTeX is definitely rather chatty. Though perhaps the information provided is useful. I wouldn't know. But it could probably use a verbose flag.
 
@FaheemMitha That isn't Lua the system. It's ... name omitted to spare embarrassing Ulrike:-)
 
@DavidCarlisle Oh?
 
@DavidCarlisle blurb
 
@FaheemMitha they are perhaps the most important warning to not ignore
 
cis
@PauloCereda quack.sty ...
 
3:42 PM
@DavidCarlisle hm ;-). Well I think the file loading message are ok - other files show up in a similar way, that is not worse than the file messages from tikz. The callback messages could perhaps be made optional. Are they from luaotfload or luatexbase?
 
@UlrikeFischer actually I guess they are from ltluatex so me/Joseph but I blame you anyway
 
@DavidCarlisle ;-). What is the current way to use dtx + hyperref? still hypdoc?
 
3:58 PM
@UlrikeFischer hmm in base we just add ihyperref in ltxdoc.cfg dont we?
 
@DavidCarlisle ltxdoc.cfg and l3doc.cls contain \RequirePackage{hypdoc}. So it is still used.
 
@DavidCarlisle They are? Normally it's doesn't make a significant difference to the appearance of the text.
Also, I don't know how easy such warnings are to fix.
 
@FaheemMitha it means that the text overruns into the margin and isn't really usable
 
@DavidCarlisle -- Yippee! A properly-used colon!
 
@DavidCarlisle Are there standard ways of fixing this?
As I said, I've been ignoring these warnings for 25 years, so it would be quite a departure.
 
4:13 PM
@FaheemMitha so don't suggest we add more warnings until you take note of the ones that are already made.
@barbarabeeton all my colons; are properly used.
 
@DavidCarlisle As I said, I ignore overflow warnings. I don't ignore other warnings.
 
@FaheemMitha strange choice, so why do you think people wouldn't ignore deprecation warnings. a deprecation warning isn't telling you anything is wrong just you are using an old form, an overfull box warning is telling you your document is typeset incorrectly. It is far more reasonable to ignore the deprecation warnings.
 
@DavidCarlisle I guess I'm weird. I don't ignore depreciation warnings. Maybe it's my way of keeping up with the Jones's.
Does anyone know what relationship github.com/drehscheibe/soul has with the soul package?
 
@FaheemMitha he wrote it
 
@DavidCarlisle Oh? The name on the documentation is Melchior FRANZ.
 
4:26 PM
@FaheemMitha exactly, click on the profile link
 
@DavidCarlisle Oh, right.
 
@JosephWright ^^
 
cis
4:43 PM
@PauloCereda @PauloCereda But, we need a periodic system with the corresponding Japanese Kanjis too..... :(
 
5:03 PM
@BambOo I can reproduce your answer works in the MWE. But it doesn't work in my original source, not even after having integrated both lines from egreg's answer.
@BambOo Not sure what to do. My original source spans multiple directories and files.
 
5:14 PM
@UlrikeFischer Do you know the minimal LuaTeX version required to load current versions of luaotfload?
 
Hi, is thee anyone with experience on Mac that can help here: tex.stackexchange.com/questions/545643/…?
 
@MarcelKrüger no, I could probably try with some older texlive. Why?
 
@UlrikeFischer I was doing some stuff in luaotfload-main.lua and came across a check where an error is issued if the LuaTeX version is below v0.95.0. Now I strongly doubt that luaotfload works with 0.95.0 so I thought about updating it, but I don't know which threshold would be reasonable.
 
@ComFreek can you send the link to your question again ?
 
@MarcelKrüger we are currently writing this as a requirement "The current luaotfload needs luatex 1.10 (present in TeXLive 2019 and a current MiKTeX). harfmode need luahbtex 1.11.2. The development targets the engines luatex and luahbtex and the version that are in TeXLive 2020." And if needed we can require for some feature a newer version.
 
5:23 PM
@UlrikeFischer Thanks. Then I will update the test to 1.10.0 to correspond to the documentation.
 
@MarcelKrüger we can't really test older systems anyway, so basically luatex + luaotfload is a bit of a combo.
 
5:38 PM
Yes, but I think the test makes sense anyway to avoid people from reporting very odd errors when they try it with versions which are known to be incompatible. So failing early with a reasonable error message seems useful.
 
6:00 PM
@MarcelKrüger yes, if we know that it fails for a version, it would be not bad to test it. I could try tomorrow with a few older texlive and look which breaks
 
6:14 PM
 
6:47 PM
A nice game :

https://www.youtube.com/watch?v=_SzKwS-kqXE
 
 
1 hour later…
8:15 PM
@T.D. -- I know I'm late to the party, but it seems you might think that amsmath incorporates amsthm. It doesn't -- they were designed to work together, but neither one makes the other available. If you want both, you need to load both (or something else that loads them, like mathtools loads amsmath). amssymb is also independent (although it loads amsfonts).
 
8:27 PM
@Circumscribe I researched that this morning. The author seems to no longer be at the University of Vienna. Was probably a grad student when the package was written.
 
Donald Arseneau in action: tex.stackexchange.com/a/545746/2693
 
@FaheemMitha -- On CTAN, some stuff does get moved into "obsolete". Although that's not the same as removal, it's an indication that a package shouldn't be used for anything new.
@DavidCarlisle -- Just getting them to realize there's a log file is to some a major revelation.
 
@AlanMunn catoptions
 
8:43 PM
@DavidCarlisle Yes, I was just remarking on the "I never thought I'd see an actual good use of \outer" comment. :)
 
@AlanMunn yes that's self evidently true (although @barbarabeeton tries to defend it occasionally) (apart from its use in bm.sty of course)
 
@DavidCarlisle Do you need written permission from the author to use it? :)
 
@AlanMunn that would be good, then Karl would eject it.
 
9:07 PM
@DavidCarlisle -- Here's an example of why you need to stress that there's a real difference between text and math modes: tex.stackexchange.com/q/545571
 
9:50 PM
boo uplatex matches .*ptex :(
 
@DavidCarlisle ? really?
 
@UlrikeFischer er well matches .*platex I meant:-)
 
@DavidCarlisle Sounds good
 
@DavidCarlisle ;-). And where doesn't that fit?
 
Sorry for interrupting. Are the lexical, syntax and semantic analyzers used in TeX to produce PDF identical to those used in MathJax, KaTeX and CSharpMath?
 
9:58 PM
@MoneyOrientedProgrammer No
 
@MoneyOrientedProgrammer no mathjax katex and tex have no code in common with each other
@UlrikeFischer my parsing of !TEX comments always picked platex even if you said uplatex as it matched TEX.*platex (fixed)
@MoneyOrientedProgrammer tex doesn't really have separate lexical analysis phase like compilers of other languages, the lexical analysis is interwoven with execution
 
In my hypothesis, MathJaX, KaTeX, TeX and others must have common denominator. They are only different in rendering phase. For example, TeX renderer produces DVI (and then PDF), KaTeX renderer produces another thing displayed on a web page.
There should be a common denominator, they don't have it?
 
@MoneyOrientedProgrammer false hypothesis.
@MoneyOrientedProgrammeras I said they have no common code.
 
OK. Let me explain what I want to achieve. I am interested to show equations in 3D game created with Unity. There is already a component assets sold in Unity assets market called TeXDraw (https://assetstore.unity.com/packages/tools/gui/texdraw-51426).

I think the author of TeXDraw used some parts of the existing open source available here (https://github.com/verybadcat/CSharpMath).

The objective is to find the common code that can be shared or reused for many other projects such as displaying equations in Unreal Engine, Unity, Three.js, etc.
 
10:14 PM
@MoneyOrientedProgrammer OK so the C# one is similar: it shares no code with tex, mathjax or katex.
I have heard of unity of course but know no details what does it need as rendering specification?
 
@DavidCarlisle I have no idea as well. Because they have no common mechanic shared among the previously mentioned applications, reading Computers and Typesetting, Volume B, TeX: The Program will not help me to understand the mechanic of github.com/verybadcat/CSharpMath/tree/master/CSharpMath, right?
 
@MoneyOrientedProgrammer take something like amsmath align in latex that is implemented as a series of more or less complicated tex macros measuring boxes and doing tex-arithmetic and finally passing things to the tex \halign primitive to align the cells, then tex's back end generates pdf from that. in katex and mathjax exactly none of that code is used. ...
@MoneyOrientedProgrammer they just take the documentation of align and write some javascript that implements that as near as they can and lays out some things in a web platform css/html/svg rendering. the C# one is similar it is a direct implementation of a small subset of latex syntax directly to the .net graphics calls.
 
The parsing is different?
 
@MoneyOrientedProgrammer very
@MoneyOrientedProgrammer there is a reason why most mathjax questions on this site are closed as off topic, it is not tex.
@MoneyOrientedProgrammer so you could base your system on any one of those but the implemenation would be quite different in each case, and not knowing the game engines I can't say which is nearest to your goal
 
OK. Is there a common configuration file or setting file describing the size, distance between characters, numbers, symbols etc for rendering equations in KaTeX, TeX, MathJaX, TeXDraw, etc?
Is there such a file?
 
10:29 PM
@MoneyOrientedProgrammer no there is no common anything. katex and mathjax reverse engineer the spacing used by tex as far as they can, and if users complain it differs they tweak their javascript a bit more.
 
The last question for today. Is this settings (github.com/verybadcat/CSharpMath/blob/master/CSharpMath/Atom/…) copied from files accompanying TeX distribution?
@DavidCarlisle: Thank you very much for enlightening.
 
@MoneyOrientedProgrammer indiirectly I would guess, the names are mostly from latex packages as noted in the comments and new BinaryOperator is I assume \mathbin as in latex's fontmath.ltx file \DeclareMathSymbol{+}{\mathbin}{operators}{"2B} but I would guess it's more or less hand done rather than a direct translation
 
@DavidCarlisle: Thank you again.
 
@Skillmonlikestopanswers.xyz straight toc for you
 
10:59 PM
@DavidCarlisle :)
@DavidCarlisle much better.
 
@Skillmonlikestopanswers.xyz I also pulled the toc code out to a jekyll include file so easily updated across all language versions so the german toc is straight too: davidcarlisle.github.io/ll/de
 
@DavidCarlisle that looks pretty Anglo-Saxon German to me... :)
@DavidCarlisle otherwise looks good.
 
@Skillmonlikestopanswers.xyz try fr that's more French
 
@DavidCarlisle but I don't speak French.
 
@Skillmonlikestopanswers.xyz google translate it to German, what could go wrong?
 
11:06 PM
@DavidCarlisle I thought @UlrikeFischer and her husband were translating it right now?! Otherwise I could translate it as well and let @UlrikeFischer or even Frank proof-read it.
 
@Skillmonlikestopanswers.xyz but I have complete faith in google translate, so I'm sure my French pages are perfect.
@Skillmonlikestopanswers.xyz it's probably a bit early for non-mechanical translation to be honest, I would guess there are several smaller or larger adjustments to be made once more people actually try to use the lessons and comment.
 
@Skillmonlikestopanswers.xyz my husband translated welcome to velcome ;-). If you have the time to translate it would be fine, we can then do some proof reading.
 
@UlrikeFischer in that case there was a misunderstanding, my bad :) I think, @DavidCarlisle's actually right and we should give it a few weeks before translating it (in example, @moewe made some adjustments to the citations already).
 
11:33 PM
@UlrikeFischer /tmp/downloads/tmp_3991/main.tex:4: error: File xskak.sty' not found`
 
@DavidCarlisle ooh checkmate
 
@PauloCereda I could use the secret 2020 URL :-)
 
@DavidCarlisle ooh but it's a secret :)
 
@PauloCereda amazingly it still is, given that the ducks got to hear if it
 
@DavidCarlisle quack <3
 
11:43 PM
@PauloCereda If I get another update at Overleaf so it works there, next someone will be asking to fix this: Unknown command: uplatex
 
@DavidCarlisle :)
@DavidCarlisle Does csplain work?
 
@PauloCereda the documentation says the command parameter has to be pdflatex/xelatex/lualatex but I assume platexand uplatex are in the underlying tl2017 docker but just needs a hook to the URL interface to make it do uplatex then dvipdfmx to get the pdf
 
@DavidCarlisle ah I see, quite possible
 
@PauloCereda currently it fails at overleaf for a rather different reason (they use latexmk not arara:-)
 
@DavidCarlisle ooh :)
@DavidCarlisle speaking of birds, have you been following the recent discussion regarding Oracle's new licensing?
 
11:48 PM
@PauloCereda or more exactly don't put tex comments in perl files.
@PauloCereda no
 
@DavidCarlisle LOL
@DavidCarlisle let me find some summary...
@DavidCarlisle: "It is Oracle’s preference that customers purchase Java SE licences. Oracle have since transitioned Java SE to a subscription licensing model."
 
@PauloCereda er as in pay actual money?
 
@DavidCarlisle thankfully, there's a ton of far better JVM alternatives.
@DavidCarlisle Yes, it's rubbish.
@DavidCarlisle I don't use an Oracle VM for a long time now, probably since Java 5. :)
 
I haven't installed for a while I seem to have:
$ java -version
java version "1.8.0_131"
Java(TM) SE Runtime Environment (build 1.8.0_131-b11)
Java HotSpot(TM) 64-Bit Server VM (build 25.131-b11, mixed mode)
how far behind the times is that?
 
@DavidCarlisle surprisingly, you have the stable 8 series, which is good enough for 99.9% of stuff out there. Java is now in version 14...
 
11:54 PM
@PauloCereda so if I want to update I can't just go to java.com and download the latest anymore?
 
@DavidCarlisle You can, I guess. :)
 
@PauloCereda for free? I'm confused.
 
@DavidCarlisle yes, you can still download an official Oracle JVM.
 
@PauloCereda oh it's the jdk for compiling stuff that costs?
 
@DavidCarlisle to be honest, I am not following the discussions since last year. :) Since I use an opensource implementation, I really do not worry. And by using a manager like SDKman, I can get literally dozens of opensource JVMs from different vendors just by selecting them in a list.
 

« first day (3491 days earlier)      last day (1433 days later) »