« first day (4867 days earlier)      last day (81 days later) » 

12:27 AM
Does anyone else do slides in class article? Every couple of years or so I start looking at the docs to Beamer, decide no, and go back to article.
1 hour later…
1:41 AM
@dedded You're missing all the fun :)
@samcarter I'm a simple man, with simple tastes. :-)
It's not that I'm too lazy to read 200+ pages, couldn't be.
2:32 AM
@dedded No, but I create slides in markdown and have pandoc make the beamer code for me. Fewer errors, and much faster, given beamer's high markup to content ratio.
5 hours later…
7:24 AM
@dedded I did them in beamer a few times, and even wrote a derived/parasitic class around it. But I'm not quite happy with it. Luckily I don't have to do slides anymore (at least not in the foreseeable future), so that's a non-issue for me :)
@dedded There are various tutorials for making slides in e.g. KOMA, so yes, I am sure people do - in the end it's a question of what works for you
1 hour later…
8:34 AM
@dedded Hmm. I had the same sensation for years, but now (after I had my skeleton presentation ready), it's just as fast as every other thing... easy things are easy, and difficult ones are possible. (I even have a duck style ;-))
8:46 AM
@CarLaTeX Replied :)
9:21 AM
9:50 AM
user image
@AlanMunn ^^
2 hours later…
12:12 PM
has anyone any advice as to whether one should use markdown -> beamer for simple presentations ?
my rationale is being able to convert it to other formats
@Minsky Very much a 'it depends' question: I'd for example say presentations should not really have any words in them .. so not so suitable ;)
@JosephWright Ohh, only formulas (or emojis)
12:30 PM
but being realistic i mean
12:54 PM
@mickep only live demonstrations of exploding things :)
@samcarter Oh, the joy of being a chemist.
What is a good way these days to display all available chars from a ttf font, only using the ttf file as input (I thing fonttable needs there to be tfm data?). I can use fontforge but keep getting surprised by random chars I find way up in the array with most other slots being empty
@Minsky Really depends on you, honestly
@daleif For a quick check, I sometimes use fontdrop.info/
1:12 PM
@JasperHabicht that's an interesting one, thanks
@PauloCereda I've never used it, so I'm doing well. :)
@daleif I guess you can sort in fontforge, and also only go compact mode (only showing the slots that are non-empty).
@yo' Thank you!
@daleif "quick" of course depends on the font size and the internet connection ... I once needed to check something in a Chinese font ... waiting ...
@CarLaTeX ... so your tcolorbox code now works on overleaf, too?
1:28 PM
@AlanMunn ooh :)
holidu.co.uk/magazine/pizza-cities Shouldn't all the UK cities be disqualified due to pineapple contamination?
1:41 PM
@mickep ahh, did not think of that. I was looking at the AU Peto font. I think that whas used when you were at AU (it is that strange font where the AU logo comes from).
@daleif Oh, that funny one. Should be pretty sparse, I guess.
@samcarter Yes, it works perfectly!
@CarLaTeX Glad to hear!
@mickep it has 226 chars. I never looked at the lower case letters, thought it was only upper case.
@samcarter This ranking is ... surprising
1:53 PM
@mickep given that the look of this font is so strange, it could be nice to have a doc that shows the normal chars and then these.
@JasperHabicht indeed!
2:11 PM
@DavidCarlisle enjoyed my question about penalties?
@UlrikeFischer :)
@UlrikeFischer Germany win? ;)
2:44 PM
@CarLaTeX Yay!
@PauloCereda const % = 0.01; // as simple as that :)
@yo' ooh :D
3:08 PM
@UlrikeFischer you confused me by adding the cubpenalty settings
@UlrikeFischer has it really never worked, and no one noticed?
why is there mathoverflow.net and math.stackexchange.com ? is the former not part of the SX network?
@daleif Ive seen an opentype version somewhere fontspec perhaps, but it could be a big table...
@JasperHabicht History
@JosephWright I only today see that! Which one should I refer to if someone asks on this site how to write something as mathematic fomula? Both?
@JasperHabicht I think mathoverflow is strictly professional-level maths
3:14 PM
@JosephWright Alright! Thanks!
Q: Math Stack Exchange versus MathOverflow for a beginning graduate student

DCTAs I understand it, MathStackExchange is for homework level questions and MathOverflow is for research level questions. As a student, I always posted in MathStackExchange unless I had some weird question that came up in undergraduate research. However, as a beginning graduate student, I'm starti...

@DavidCarlisle Oooh! Okay, noted.
3:48 PM
@DavidCarlisle found some perl code that can convert data from ttfdump into an actual string of chars, that can (after one have removed special char and math chars) be printed using \text_map_inline:nn (via some egreg answer), then run that through lualatex with a proper handler macro on each char
4:11 PM
@PauloCereda This is the logic of that well known English expression "Two of one, sixth of a dozen of the other".
4:37 PM
I know you are not a ConTeXt user but there are some nice modules for similar things. For example:

user image
(Long list, several pages. If you only give one font in the list, you only get that one of course...)
@JasperHabicht ...or Iosevka... (a.k.a. lualatex killer)
@Rmano Oh my, what a font!
@mickep how does on run context these days?
@daleif "context file.tex"
@daleif context ...
4:47 PM
@mickep very nice, I will save it! (How can I see a font in a .ttc file?)
@Rmano I just tried to change.otf to .ttc. I got 944 pages. :)
ahh tried running luametatex which failed context did something
That was super useful
@mickep Ah yes, didn't find it the first time around. Wow.
Why is this font using private slots, having a black square in fd058.
@daleif The font creator can put whatever they like in the private slots. Iosevka had many many of them.
4:54 PM
Ahh well of to todays walk home. Nice way to get my daily 5km (take the bus in the morning). We got snow in Aarhus as the begining of the year, took more than a week before the busses where running again and the roads were unsafe for bikes, so warked 4-5km twice a day.
@mickep I know. AU has now a font with their logos and seal. They placed them in the private area. For pdflatex support I just remapped them to A,B,C,D and made a package that provides names for those four chars with support for pdflatex, xelatex and lualatex. Trying to do something simiar with the Peto font using autoinst where I'll just replace the generated .sty and write one manually that uses fontspec under xelatex and lualatex.
@daleif Nice with exercise!
@daleif Ah, it was more like a "Why on earth did they add a black square at a random private slot?". :)
5:29 PM
@DavidCarlisle it looks so.
@UlrikeFischer ah fmi not wspr I knew we had one somewhere
@UlrikeFischer So, if I need to put \DocumentMetadata{} before \documentclass if I want to use l3opacity, how would I be able to write a package that loads l3opacity? Maybe I should check what \DocumentMetadata{} actually does ...
@DavidCarlisle :)
@DavidCarlisle FMi wrote for TLC3
@UlrikeFischer wow that's the benefit of obscure documentation, no one finds blatant bugs for 30 years
@DavidCarlisle Seen that l3graphics got a prod? ;)
5:39 PM
@JosephWright no I've been deep in xml land all day just surfaced now
@DavidCarlisle Cool
@DavidCarlisle The PR I wrote a while ago: muzimuzhi had a related one about some of the setup, but if we merge l3graphics to l3kernel it's redundant
@JosephWright ah I would have caught up with email at some point. may not be able to look until later in week
@DavidCarlisle It's OK, not urgent, just interesting :)
5:48 PM
@JasperHabicht you can test with \IfDocumentMetadataTF it is has been loaded and error or warn if not.
@DavidCarlisle ;-) It also shows that nobody missed it. I only found it because I pondered over unbalanced footnote columns in relemac and wondered if there is a way to force an even number of footnote lines ...
@UlrikeFischer Hm, so an idea could be to create a package option to activate opacity support where I could test and warn ... yes ...
@JasperHabicht We are looking to integrate into the kernel, but for the medium term I think you are best going with checking for meta-data loading
@JasperHabicht look at how transparent.sty handles it.
@UlrikeFischer @JosephWright Alright, thank you both!
6:16 PM
@DavidCarlisle There is a penalty bug?
7:07 PM
@mickep yes
@mickep unless you have some insight into how \interlinepenalties is supposed to work
7:39 PM
@DavidCarlisle How would you find such a bug! We all know that TeX is much more knowledgeable than every human being when it comes to typesetting. If it says that a penalty is appropriate, of course, we believe it! =)
@DavidCarlisle Supposed, no. But it seems that \interlinepenalties 3 12 34 56 adds 12 between the first two lines, 34 between the second and third and 56 between the rest. (At least here.)
But I guess you mean some more complicated case?
@mickep Using?
@JosephWright luatex/luametatex.
(ConTeXt, I do not know how to trace such things elsewhere)
\showboxbreadth\maxdimen\showboxdepth\maxdimen \tracingoutput=1

%\interlinepenalty 100
\interlinepenalties 3 12 24 36

\clubpenalties 3 66 77 88

a\break bb \break c \break d \break e

@mickep ^^^
@mickep How do you test LuaMetaTeX ... Hans doesn't provide a format for testing :)
@JosephWright See my msg after...
7:45 PM
@mickep not with pdftex or luatex
@JosephWright I did not try in plain, but I get in context with luatex. 78(12+66), 101(24+77), 124(36+88) and 2124(2000+36+88). Here, the numbers in parentheses are my guesses how it is added, in the tracing I switched on I only see the sum...
@DavidCarlisle What? No problem in those engines?
@mickep that's odd. with this example with no explicit penaties
\showboxbreadth\maxdimen\showboxdepth\maxdimen \tracingoutput=1

\interlinepenalty 123
\interlinepenalties 3 10 20 30
%\clubpenalties 3 66 77 88
\hsize 6cm

a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
with luatex texlive 2023 I get
$ grep penalty club.log
..\penalty 273
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 123
..\penalty 273
@mickep how do you switch on tracing in context? (And how do you force the use of luatex?)
so \interlinepenalty used not the array, and if you set that to 0 you get no penalties between the lines
@UlrikeFischer context --luatex for the bit in brackets ;)
7:50 PM
@UlrikeFischer % engine=luatex on first line will give luatex. Or as @JosephWright suggests. The tracing here was turned on with \showmakeup[penalty].
@DavidCarlisle Maybe interlinepenalty is somewhat global (what do you get if you add another paragraph) while interlinepenalties is not. But I'm guessing.
@mickep Ah, that looks plausable
@mickep nope
\showboxbreadth\maxdimen\showboxdepth\maxdimen \tracingoutput=1

\interlinepenalty 0
\interlinepenalties 3 10 20 30
%\clubpenalties 3 66 77 88
\hsize 6cm

a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a

a a a a a a a a a a a a a a a a a a a a
a a a a a a a a a a a a a a a a a a a a
with luatex:
$ grep penalty club.log
..\penalty 150
..\penalty 150
..\penalty 150
..\penalty 150
@DavidCarlisle Try without the blank line before the first block of text
$ grep penalty club.log
..\penalty 160
..\penalty 20
..\penalty 30
..\penalty 30
..\penalty 30
..\penalty 180
..\penalty 150
..\penalty 150
@JosephWright and with pdftex
$ grep penalty club.log
..\penalty 160
..\penalty 20
..\penalty 30
..\penalty 30
..\penalty 30
..\penalty 180
..\penalty 150
..\penalty 150
hmm so a par resets \interlinepenalties but not \interlinepenalty or \clubpenalties
@UlrikeFischer ^
@DavidCarlisle Yes, seems so
8:03 PM
@DavidCarlisle yes just saw it too (and in context it is the same). But that doesn't make much sense imho.
@UlrikeFischer well no
@DavidCarlisle, @UlrikeFischer Ah, I do have a note about this in l3galley:
%    \end{macrocode}
%   No marker, so really insert a paragraph: the \cs{tex_par:D} is inside a
%   group to preserve some dynamic settings (for example
%   \cs{tex_interlinepenalties:D}). In vertical mode, that means just
%   inserting the primitive.
%    \begin{macrocode}
@UlrikeFischer, @DavidCarlisle Lots of good stuff in this area in l3galley ;)
@UlrikeFischer Since you test in ConTeXt, you can also try \frozen \interlinepenalties...
8:57 PM
\opacity_fill:n and \opacity_stroke:n are broken and produce broken PDFs with up-to-date MiKTeX, I think.
But if I use the current l3opacity package and upload it to Overleaf (TeX Live 2023), things work. So, the package is not the culprit.
\opacity_select:n works fine.
I should add that \opacity_fill:n and \opacity_stroke:n work fine with TeX Live 2023. So, something inbetween changed.
@JasperHabicht Could well be - this code isn't really tested ATM
@JosephWright I don't know how to properly test. I just use it, check the source files and try to understand things =) But opacity is somewhat opaque to me ...
@JasperHabicht can you show an example?
@UlrikeFischer I filed an issue at GitHub
But you can actually already reproduce with:
\opacity_stroke:n { 0.5 }

This produces a broken PDF for me
@JasperHabicht, @UlrikeFischer I see the issue
9:52 PM
@JosephWright missing space ? or not. That is probably intended
@UlrikeFischer No :)
@JosephWright yes just saw it. What is the problem?
@JosephWright ah, you miss the 0.5 in the resources.
@UlrikeFischer Two things, one incorrect parameter passed, one place I forgot we need to initiate some data - all fixable in an l3backend update
@JosephWright yes, the fill is wrong too.
@JosephWright I love uncompressed pdf ...
@UlrikeFischer, @DavidCarlisle Probably if l3graphics moves, l3opacity should too ...
10:12 PM
@JosephWright Thanks for your and muzimuzhi's incredibly quick work!
@JasperHabicht I'm updating the tests
@JasperHabicht Probably will hold off from a release until after the TL'24 freeze unless it's urgent
@JosephWright Sure, no problem

« first day (4867 days earlier)      last day (81 days later) »