« first day (4880 days earlier)      last day (70 days later) » 

9:35 AM
@UlrichDiez usually well, exceptions are if the previous argument type is p, in which case you have to use {/}.
 
10:16 AM
Fontspec question: Say I've defined a new fontfamily \FooBar, what is the proper way to locally redefine \rmdefault to use \FooBar? I have a setup that I'm testing and some of the data is set via \url (with \urlstyle{rm}) and I need to change that font as well, so just issuing \FooBar does not work.
 
10:46 AM
@CarLaTeX @egreg @Rmano Does one of you know an Italian guide for making minimal working examples?
 
@samcarter No, I never bumped into one...
 
Since it is only \urls rm style, and \urlstyle{rm} is littered inside the code it is probable best to do something like
\makeatletter
\def\url@rmstyle{%
\def\UrlFont{\FooBar}}
\makeatother
 
 
1 hour later…
11:55 AM
@samcarter No
 
@samcarter se il tuo esempio ha più di questo, non è minimo
\documentclass{article}

\begin{document}
Ciao mondo
\end{document}
@egreg impressed?
 
@Rmano @egreg Thanks for your replies!
@DavidCarlisle I don't know why I pinged the Italians and not the master of all languages!
 
@samcarter I did wonder
 
12:18 PM
Hmm, I'll see if I can get a mathtools update in before the TL build. There haven't been more development on it since january.
 
@daleif Maybe a stupid question, but why upgrade if there has been no changes?
 
Actually, learnlatex has a short section about MWEs: learnlatex.org/it/lesson-16#ricevere-aiuto
 
@mickep I haven't released the stuff that I fixed in january (I'm not Joseph)
 
@daleif Karl wants to take last ctan updates today, so it might be too late, you can always update after the release, which is safer anyway
 
@DavidCarlisle Too late, I just sent it off. Was more thinking about the users on OverLeaf. Karl kan take it if he wants.
@DavidCarlisle apparently I need your help with git. It seems I've sent my version number changes to master branch instead of main. I thought I had rebased or what ever my local repro.
Any idea on hot to fix this?
 
12:30 PM
@daleif Oh, that makes sense.
@DavidCarlisle Let us hope that what is in there now is stable :)
 
@daleif there is no master branch (for reasons of politics and social justice) if your local checkout is still on a master branch then execute thise to get on to main:
git branch -m master main
git fetch origin
git branch -u origin/main main
git remote set-head origin main
@daleif overleaf won't take the first cut of 2024 so any udates in the next few weeks will make it on to overleaf
 
I think that and git push origin HEAD helped. There is no longer a master pull note on github. Not sure why git push started to mention master.
 
@daleif You have just made a master branch on github
@daleif but the commit is now on main as well so if that looks good I assume you can delete master
 
@DavidCarlisle can I delete ti from the webinterface?
 
12:39 PM
@DavidCarlisle I saw that as well.
 
@daleif yes there is a bin icon at the right hand side of github.com/latex3/mathtools/branches
 
@DavidCarlisle looked the wrong place. Deleted
 
@daleif thanks
 
1:18 PM
quack
see my contribution chart: github.com/cereda
6
@JosephWright ^^ :)
 
@PauloCereda spray? I had this in my github timeline today :)
 
1:35 PM
@samcarter same lol
 
@PauloCereda :D
 
@CarLaTeX Perfect! Thanks a lot!
 
@samcarter You're welcome!
 
 
1 hour later…
3:44 PM
@PauloCereda months of carefully timed contributions? (or a faked git history?)
 
@PauloCereda ah so untrustworthy faked history
 
Do we have a guide or an answer on the site as to how one can substitute the math letters under lualatex by and italic text font from a non math font? (just for testing, wanted to see what was possible)
 
@DavidCarlisle very well crafted history management
 
@daleif this?
4
A: “Old” math font commands (\mathrm etc.) and unicode-math (“range” option)

Ulrike FischerIt you want really to change \mathbf and similar: This is done easily by (re)declaring the math alphabets. The problem starts if you want to adapt the \symXX commands which normally map chars to the glyphs on the mathematical plan. Here the implementation is rather broken. It works more or less...

A more reliable source than @PauloCereda
 
3:54 PM
@DavidCarlisle oh no
 
@DavidCarlisle Something like this, though I'd like to specify the font via filename (untested)
 
@PauloCereda also we need a version that will run on locally hosted gitlab so that it can be added to workflows at work
 
@DavidCarlisle ooh
 
@daleif that argument just goes through to fontspec so you can use filenames or the internal font name
 
@DavidCarlisle I think the target repo could be in URI form
 
3:58 PM
@DavidCarlisle it seems to work. Can one scale it as well (the setup is a but wonky, as some part has scaling uptions others does not, so smallest cdommon demonimator dictates).
Scaling works
Hmm, might has an ok math setup for one of our AU fonts under LuaLaTeX then. (using KpMath-Sans with letters and perhaps numbers substituted, need to see how the kerning ends up. I think I've played with this before and things have often turned up bad after some testing)
Any reason why this font should not work under lualatex? tug.org/FontCatalogue/couriertenpitchbt
\documentclass[a4paper]{article}
\usepackage{courierten}
\begin{document}
\begin{verbatim}
FooBar
\end{verbatim}
\end{document}
Gives
LaTeX Font Warning: Font shape `TU/CourierOneZeroPitch-TLF/regular/n' undefined
 
@JosephWright clearly a not-so-well-thought-through-interface-issue: tex.stackexchange.com/a/712811/117050. \keys_precompile:nnN should've always set a function, imho.
 
@daleif yes that's come up before with different packages. If I remember correctly the generator has started generating regular as the font shape but latex expects m and expecially expects m/n as the default fallback font so if it's not theer it mans (although won't actually matter unless you select a missing font and the fallback is needed) the package sould get fixed:-)
 
@Skillmon I'd thought about that: we always set variables, and of course this ultimately is a matter of taste
 
@JosephWright \cs_set:Nn doesn't set a variable :P
 
4:13 PM
@Skillmon Sure, yes - I mean things with a 'return value' :)
 
@JosephWright 'return value' is very arguable, \cs_set:Nn has a return value if you so wish to call it.
 
@Skillmon I know it's arguable - I did think about this when we added precompiling, honest
@Skillmon My point is that you can argue that a tl is entirely the right return form: you are after all going to insert 'some tokens'
 
4:44 PM
@JosephWright hm, just another reason why expkv is superior, it doesn't have to deal with that distinction and hence it is clear that it assigns to a function (also, you can use #1...#9 in precompiled keys, so maybe \keys_precompile_cs:Npnn?)
 

« first day (4880 days earlier)      last day (70 days later) »