« first day (5066 days earlier)      last day (27 days later) » 

6:34 AM
@samcarter It's not a hack - it's the official approach
@JosephWright so slow (and for those few words)
@Skillmon :)
7:34 AM
@JosephWright @Skillmon Thank you both!
 
3 hours later…
10:17 AM
@JosephWright is using here lots of braces around the €\, to avoid the error a "sane" solution? Or should one use \multicolumn?
\documentclass{article}
\usepackage{siunitx}

\begin{document}
%errors:
%\begin{tabular}{>{€\,}S}
%{header}\\
%\end{tabular}

%errors:
%\begin{tabular}{>{{€\,}}S}
%{header}\\
%\end{tabular}

% works
\begin{tabular}{>{{{€\,}}}S}
{header}\\
\end{tabular}

% works
\begin{tabular}{>{€\,}S}
\multicolumn{1}{l}{€\,header}\\
123
\end{tabular}

\end{document}
10:48 AM
@UlrikeFischer hmm, two groups I can see goes wrong - I can take a look but I'm not sure there's a better way
@JosephWright I only want to know which advice to give. If the additional braces are too fragile I would go for the multicolumn approach. (And perhaps the siunitx docu should mention this case somewhere. Using >{..} is not unusual ...
Bär is a bit nervous if visiting the UK is safe: bbc.com/news/articles/c4gly3l8y45o
6
@UlrikeFischer hmm... >{\mbox{...}} (it was the standard workaround suggested for similar problems in tikz options...)
11:50 AM
@JosephWright I'm currently at dayjob, so can't look it up right now, but regarding tex.stackexchange.com/questions/726354/… There was a policy change to use the generated variants more than adding low-level optimised code paths to some functions. I was just unsure whether \str_if_eq:eeTF was hit by that policy change.
 
1 hour later…
1:03 PM
Doing some clean up after my latex2e to expl3 implementation of my letter class. I'm wondering, is it better to use \keys_define:nn or \DeclareKeys to define the class options?
@daleif depends :) But if you're in expl3 coding regime anyways stick to the expl3-variant for performance reasons. \DeclareKeys is mostly for the 2e-level.
1:26 PM
@Skillmon This is an Expl3 class, so I'll switch to make it faster.
BTW: for documenting the source code (instead of explaining the choices in the cls), it is better to use l3doc or just the normal dtx setup? Note this is only for documenting and explaining the source code (for me), it will not be used for the documentation, this lives in a separate document.
The sources are rather large and complicated and I had to re-learn what my thinking was when I made the code some years ago.
1:49 PM
@Skillmon so \DeclareKeys -> \keys_define:nn and \SetKeys -> \keys_set:nn, what aboiut \ProcessKeyOptions is there something we should use instead? or just leave it.
@daleif Leave it, that's just part of 2e.
 
2 hours later…
3:56 PM
@daleif As @Skillmon says, option processing is a 2e concept so just has the one interface
 
3 hours later…
7:09 PM
The TUG’24 proceedings are online: tug.org/TUGboat/tb45-2
3

« first day (5066 days earlier)      last day (27 days later) »