« first day (5195 days earlier)      last day (34 days later) » 

07:34
@Rmano nice one
08:08
@daleif Thanks! Although you know, the comma... 😏
08:26
@Rmano the comma is just a figment of your imagination
 
2 hours later…
10:16
user image
7
Well it seems that gold badges follows a clear trend... let's see for 55555 ;-)
10:32
@Rmano WOW!
@CarLaTeX 😊
10:49
@Rmano not for me (1 with 70k)
 
1 hour later…
11:50
i really hate to admit it, but the Expl3 capabilities are pretty awesome…
@Lupino :)
12:09
@Lupino Why do you really hate to say that?
@mickep because from a reading-psychological perspective the syntax is abhorrent…
3
@Lupino That is only there to scare the weak users to other macro packages (or back to Word).
@Lupino you get used to it, although xii.tex plain tex is more readable, I agree
(i had to laugh harder than i'd like to admit)
12:27
@Lupino it really isn't. It's much easier to read than 2e
(once you spent some time there)
13:43
@Skillmon I'm going to look into the \AssignTemplateKeys business
13:58
@JosephWright One could use a \tl_act:... loop :P If only it was a public function and not removed from the kernel
@JosephWright better idea: You're only searching for a single token. Just \usepackage{etl} and you're good
yo'
yo'
14:26
@Lupino I don't agree; the syntax is just fine, but it's highlighting is way too wrong in most IDEs (incl. Overleaf, just to shoot our own leg!)
@mickep MS Office seems to render it with a +ve vertical gap
@Skillmon Current plan: do a \str_if_in:nnTF test first, only use \regex_match:nnTF if there is a hit - should be a lot faster for the setup but not change behaviour at all
@mickep Here is another font with both vertical and horizontal gaps set to 0
@yo' vimtex is ok I think...
14:37
@JosephWright we should l probably fix this. Also does l3 version do better? tex.stackexchange.com/a/735446/1090
@DavidCarlisle Uses the string mapping for files, so should be reasonably robust
@DavidCarlisle But no, no change to the basic idea - it's all done in TeX
@خالدحسني Yes, and I do not see that one would like them to be split more like in that figure. Or, maybe I just misunderstand the purpose.
@خالدحسني This looks already better to me.
14:58
@JosephWright but you need to turn off the engine level file encoding handling otherwise there is no input at all to tex
@JosephWright I'll try later:-)
@JosephWright New plan: \RequirePackage{etl} then \str_if_in:nnTF and if true \etl_token_if_in_deep:nNTF
(jokes aside, yes, sounds like a very reasonable plan)
@Skillmon I see the point and we perhaps should look again at 'action' loops - they are handy
@Rmano mostly.
@Skillmon Half-way through - have done one timing run, now doing the comparison
@JosephWright well, if it is not inside this should be faster by roughly a factor 1000, if it's inside it should be slower by a factor 0.001. So seems like a very reasonable operation :P
15:03
@Skillmon Sure, but I want to see the 'real world' impact - in this case on the test suite run time
@DavidCarlisle Oddly, I've not done that
15:19
@DavidCarlisle I could sent you the eps if you want for testing? (I didn't try yet, was working on something else)
@UlrikeFischer you could thanks, although somewhere I should have some test eps files with tiff previews, it's come up before (although not for years:-) but why is dviluatex so popular this week?
15:35
@DavidCarlisle I try to forward but thunderbird shows only an empty mail, perhaps you will see something
@UlrikeFischer mail but no atachment, not to worry I can probably find some here
(/usr/local/texlive/2024/texmf-dist/tex/latex/graphics-def/dvips.def)))
(/usr/local/texlive/2024/texmf-dist/tex/latex/l3backend/l3backend-dvips.def)
(./tiff.aux)
! String contains an invalid utf-8 sequence.
<read 0>
   �������^^^^^^^^@^^@^^@^^C^^@^^@^^@^^@^^@^^@^^@^^@^^@^^@!^^@^^@�1^^@^^@1^^@
^^@%!PS-Adobe-2.0 EPSF-2.0
l.7 \fbox{\includegraphics{tiffeps.eps}}

?
@UlrikeFischer ^^
(./tiff.aux) <tiffeps.eps>
Overfull \hbox (30.43874pt too wide) in paragraph at lines 7--8
@UlrikeFischer fixed:-)
15:52
@CarLaTeX ohhh, not long until 66,666!
@DavidCarlisle what did you do?
\documentclass{article}

\directlua{
function convert_char (char)
  return unicode.utf8.char(string.byte(char))
end

function convert_line_bytes (line)
  return string.gsub(line, ".", convert_char)
end
}


\usepackage{graphicx}
\makeatletter
\def\Gread@eps#1{%
    \directlua{luatexbase.add_to_callback("process_input_buffer", convert_line_bytes, "8 bit input")}%
    \Gread@generic{#1}\Gread@eps@aux
    \directlua{luatexbase.remove_from_callback("process_input_buffer","8 bit input")}%
}
\makeatother
@UlrikeFischer if the above works for you, I'll update my answer. Basically "luainputenc-light"
@JosephWright ^
16:14
@samcarter I don't like that number, quite evil
@CarLaTeX it will be a step closer to friendlier numbers like 77,777 :)
@CarLaTeX but I had 666,666 and I'm not evil at all
16:30
@DavidCarlisle you are mean
@mickep I mean, it shows that my interpretation matches the implementation. The font data might be less than ideal, but that is then a font issue to fix.
Jun 29, 2017 at 16:15, by Paulo Cereda
@DavidCarlisle you are not mean :)
16:54
@UlrikeFischer the file works here, thanks
17:22
@خالدحسني Oh,I dod not question that. But I think almost all (all I have looked at) set it positive.
@DavidCarlisle you're mean for using that quote against @PauloCereda
@cfr remember you misspelled my name once: It could've been worse @Smillion
cfr
cfr
18:01
@Skillmon if it had been worse enough, I wouldn't have mispelt your name at all.
@Skillmon I think I stole a tick from @egreg.
@cfr Congratulations!
cfr
cfr
18:25
@Skillmon ;)
would anybody with Cambria be willing to try this example?
3
A: Text fractions in Cambria have too much space around solidus

cfrSee this answer for explanation and motivation. Tl;dr: \f@family will not match \rmdefault if any features are added to the font after definition, whereas \l_fontspec_family_tl will always return the base family. Essentially, when you set up the serif, you are potentially initiating a series of f...

The OP says the numerator does not work in serif unless a sans font is set explicitly. that doesn't seem to make any sense, but I don't have Cambria to test.
@cfr Because you answered the question the OP didn't ask. 🤪
cfr
cfr
19:07
@egreg you mean you didn't follow the links ;).
@egreg but the OP did specify Cambria, which implies fontspec, I think.
19:22
@samcarter :)
@DavidCarlisle Indeed you're not mean
@samcarter Did you see my ping on TopTeX?
@cfr I have Cambria, haven't seen the question yet, do you want me to test something? I'll look at the question...
@CarLaTeX no, I didn't get a notification. Let me check...
@CarLaTeX replied! I hope it helps you to make a decision (sorry for not replying earlier, I didn't see a notification)
19:52
@DavidCarlisle anything you can, to move the tick from the dodo...
20:20
@mickep you have secret mail
@DavidCarlisle Ohh! (Later, singing along to Bonnie Tyler with the oldest...)
@mickep Lost in France^H^H^H^H^H^HSweden ?
@DavidCarlisle Total eclipse of the heart
20:38
@Rmano I'm curious enough to ask. Which editor and highlighter is that?
@DavidCarlisle Interesting!
@enkorvaks neovim + vimtex with catppuccin-latte color scheme (with some color tweaks)
 
1 hour later…
cfr
cfr
22:03
@DavidCarlisle the OP complained of two problems in tex.stackexchange.com/questions/735223/…. the first is not a problem, as far as I know. however, the second sounds odd: that my code doesn't get the numerator in serif unless a sans font is set explicitly. I can't reproduce that with tg-pagella and I wouldn't expect it to happen with Cambria, but I was hoping somebody could confirm or disconfirm that.
@DavidCarlisle basically, does my code do what I say it does if you substitute Cambria for TeX Gyre Pagella?

« first day (5195 days earlier)      last day (34 days later) »