TeX, LaTeX and Friends

cfr

07:00

does anybody know if checkencodingsubset is expected to produce sensible results for fonts installed with fontinst?

Paulo Cereda

09:13

Ducks are very good friends with fish: x.com/shouldhaveanima/status/1858653918005457019

David Carlisle

09:44

@cfr yes for suitable definitions of "know" and "expected"

@PauloCereda yes they go well together sethlui.com/ng-soon-kee-fish-duck-porridge-singapore

Joseph Wright

10:28

Anyone got any idea what's going on here:

\documentclass[draft]{article}
\usepackage{unicode-math}
\begin{document}
\ExplSyntaxOn
\dim_zero:N \l_tmpa_dim
\int_step_inline:nnn { 0 } { 9 }
  {
    \hbox_set:Nn \l_tmpa_box { $#1.$ }
    \dim_compare:nNnT { \box_wd:N \l_tmpa_box } > \l_tmpa_dim
      {
        \dim_set:Nn \l_tmpa_dim { \box_wd:N \l_tmpa_box }
        \iow_term:e { #1~\dim_use:N \l_tmpa_dim }
      }
  }
\ExplSyntaxOff
\end{document}

In particular @DavidCarlisle, @Skillmon, @egreg, @UlrikeFischer ^^

Spotted doing some tables in siunitx, wondered what was happening

I can fix easily by using 7 as the 'model' for width of numbers, but why is it different with a Unicode font - issue in Latin Modern Math?

@mickep ^^ See the same in ConTeXt with LMMath??

David Carlisle

....\TU/latinmodern-math.otf(1)/m/n/10 7
....\kern0.13 (italic)
....\TU/latinmodern-math.otf(1)/m/n/10 .

@JosephWright ^^

@JosephWright all the digits are 5pt width I think you just need to isolate the .

Joseph Wright

11:00

@DavidCarlisle But there shouldn't be a kern here ...

@DavidCarlisle One of the design aims in v3 is to produce underlying (La)TeX that is as close as possible to what you'd type manually: so here what's odd is the kern, you'd not manually type a number with extra groups around the . - sigh

mickep

@JosephWright I have no clue what that code is doing :(

David Carlisle

@JosephWright well i wondered about that, although while digits are monospace . isn't so it's not necessarily in error even if it's unhepful. The kern seems odd given the shape of 7 if you kerned at all I'd expect it to be negative to kern in, but this kerns 7. out

@mickep just do \showoutput \hbox{$0.$} \hbox{$7.$}

Ulrike Fischer

@JosephWright with luatex you can use \hbox_set:Nn \l_tmpa_box {\mathitalicsmode=2 $#1.$ }

mickep

David Carlisle

11:16

@JosephWright well agreed, but I just meant the issue was the kern, not (as i thought you meant originally the digit 7 was wider)

Joseph Wright

@DavidCarlisle Yeah, I was scratching my head

@UlrikeFischer Nice

@mickep Thanks

David Carlisle

@JosephWright but I know you want to support xetex :-)

Joseph Wright

@DavidCarlisle :)

@DavidCarlisle You raise a serious point - this is a Unicode engine issue only

I don't think it would come up in a classical 8-bit situation

David Carlisle

@JosephWright and the ongoing issue about getting updates/fixes to the lm/tex gyre opentype fonts

Joseph Wright

@DavidCarlisle Yeah, that beign it's a non-starter

Ulrike Fischer

11:22

@DavidCarlisle the kern is also used if you do e.g. 7(. (I wonder if math mode can actually do different kerning for different pairs)

mickep

Isn't there something "funny" about the italic 7 in cm as well?

Joseph Wright

@DavidCarlisle I might seriously go for @UlrikeFischer's suggestion, and simply note that XeTeX may give some slightly odd results due to font issues out of my control

David Carlisle

@mickep oh look 0.13pt ... (@JosephWright)

(CHARACTER C 7
   (CHARWD R 0.511108)
   (CHARHT R 0.644444)
   (CHARDP R 0.194445)
   (CHARIC R 0.135556)
   )

Joseph Wright

@DavidCarlisle Ah

David Carlisle

@JosephWright not that the upright 7 in opentype should be based om cmti10 but....

Joseph Wright

11:30

@DavidCarlisle Indeed :)

I've logged an siunitx issue for this: github.com/josephwright/siunitx/issues/759

David Carlisle

@JosephWright I thought you'd say that

Joseph Wright

@DavidCarlisle Tracing my code right now :)

Ulrike Fischer

@JosephWright do you get the same with other math fonts?

David Carlisle

@JosephWright the fonts are open source, we should really lean on @mickep to fix them rather than patching them in context-specific Lua code. It could be a secret plan to just keep mentioning issues until he thinks of fixing the fonts without us having to say anything.

Joseph Wright

@UlrikeFischer No, at least not with the couple I checked quickly

@DavidCarlisle :)

mickep

11:41

@DavidCarlisle You only want that because then you can blame me when people notice bugs...

David Carlisle

@mickep is that a bad reason?

mickep

@DavidCarlisle Not for you, perhaps.

Paulo Cereda

@DavidCarlisle oh no

Lupino

Hello again. Where do I report a bug with luaLaTeX and hyperref (and \enlargethisage)?

David Carlisle

@Lupino if you need hyperref to show the problem, then the hyperref repo

@Lupino or just blame @UlrikeFischer

Lupino

11:47

yeah, pdf generation breaks down hard when there is an unstarred enlargethispage and a link breaks across pages.

David Carlisle

@Lupino links across pages are always "interesting" but github.com/latex3/hyperref/issues

Lupino

@DavidCarlisle Issued an issue: github.com/latex3/hyperref/issues/368

mickep

@DavidCarlisle I suggest instead that you go to BachoTeX and talk to the polish guys. Bring some Adobe money...

samcarter

13:05

@mickep How many polishfoodies.com/polish-roast-duck-recipe will the adobe money buy?

mickep

13:24

@samcarter Oh, Mr Duck might know... :)

Skillmon

13:57

@samcarter maybe we closed that second question a bit too quickly. It's about having \begin{frame}[bad idea] instead of \begin{bad idea}...

samcarter

@Skillmon I'm certain we have a duplicate for that, too :)

@Skillmon updated the link

Skillmon

@samcarter Saw it :)

samcarter

@Skillmon But they are essentially all a duplicate of "Bad idea!"

David Carlisle