« first day (4863 days earlier)      last day (61 days later) » 

3:21 AM
Just noticed that lualatex and lualatex-dev are both symlinks to luahbtex. Are they then identical?
 
4:13 AM
Suggestion, inspired by my tex.stackexchange.com/questions/170547 : add the [nonfrenchspacing] tag.
Wrong link above. Correction: tex.stackexchange.com/questions/710284 .
 
 
2 hours later…
6:26 AM
@dedded No
@dedded The way this works is that TeX engines example arg[0] to see what format to load, so lualatex => luahbtex &lualtex but lualatex-dev => luahbtex &lualatex-dev, i.e. they use hte same engine but load different format files
 
 
2 hours later…
8:14 AM
@JosephWright Hmmm ...
        \draw_begin:
     	\draw_path_circle:nn { 0 , 0 } { 1.25cm }
     	\draw_path_use_clear:n { fill }

     	\draw_path_rectangle_corners:nn { -1cm , -1cm } { 1cm , 1cm }
     	\draw_path_replace_bb:
\draw_end:
so, something like this?
Output seems to be ... strange
 
8:29 AM
:65233215 I think that being able to force the BB would be useful in several cases... I do not think that clip is the same. An equivalent to \path [use as bounding box] (a,b) rectangle (c,d); would be useful --- example : gist.github.com/Rmano/c68789c5efe8feca897da66d4e991be8 @JosephWright
 
@Rmano Yes, that's the idea.
 
@JasperHabicht I had not time to look at the proposed implementation --- so it could be already like that. The semantic in pgf is a bit complex, I would prefer to have it last in the drawing and force the BB no matter what (sometimes you want part of the drawing stick outside of the BB, like a partial overlay).
 
@Rmano The current idea is (if I understood Joseph right) to have a command \draw_path_replace_bb: that uses the current path to update the bbox. It will discard the current path, but you can put some \draw_path_use:n { ... } before.
But currently, it seems, that the command does not use the path to update the bbox, and the bbox just becomes a zero-sized box
 
8:46 AM
@dedded similarly latex, pdflatex, pdftex, etex,xmltex are all symlinks to pdftex
 
9:01 AM
@DavidCarlisle pdftex is a symlink to pdftex?
 
@Skillmon :)
 
9:40 AM
@Skillmon it is if you choose the "make symlinks in /bin" install option (which is clearly what i was refering to, as I'd never have mistyped a list)
 
@DavidCarlisle :P
 
10:02 AM
The fill and stroke commands from the l3opacity package only work with the pdfmanagement module which ... does not exist currently?
 
@JasperHabicht \DocumentMetaData should load it, iirc.
 
@JasperHabicht as @Skillmon says, you need \DocumentMetadata{} (small d ...).
 
@UlrikeFischer whoopsie :)
 
@UlrikeFischer Ah, thx!
 
@JasperHabicht transparency need resources in the PDF that must be managed to avoid incompabilities between packages like this one tex.stackexchange.com/a/253417/2388. That is one of the main point of the PDF management.
 
10:13 AM
@UlrikeFischer Yes, I understand this. I was trying to find out on my own, but I got stuck at some error pointing to \pdfmanagemant_add:nnn and then I tried to find where this is defined ... but I got lost
 
10:26 AM
@JasperHabicht texdoc pdfmanagement or texdoc -l pdfmanagement
 
@UlrikeFischer I searched on GitHub. There, it is in latex3 / pdfresources
 
@JasperHabicht yes it started under this name.
 
 
4 hours later…
2:29 PM
@JosephWright Ah, I figured that was the difference between lualatex and luatex, but I always thought that ...-dev referred to a new engine.
 
@UlrikeFischer can we redefine \pmb in latex-lab-amsmath ?
 
@DavidCarlisle I just wanted to look at that ;-). Any suggestion for a better definition? (Btw: I fixed the two bible issues in tagpdf/dev.)
 
@UlrikeFischer can we just tag the first and tag the overprints as artifact?
yes I saw the diff mail although haven't rebuilt yet.
@UlrikeFischer you could in theory get the same from \bm if it can't find anything better and decides to do \bm@pmb which is similar to amsmath \pmb (but better, of course)
 
2:57 PM
@JosephWright With the code of your last commit, \draw_path_replace_bb: first does \draw_reset_bb: which sets \g__draw_(x|y)(min|max)_dim to \c_(min|max)_dim. Then, \__draw_path_use_bb: is called, but the comparison \dim_compare:nNnF { \dim_use:c { g__draw_ #1#2 _dim } } = { #3 -\c_max_dim } in \__draw_path_use_bb:NnNN will now always result in TRUE, so the bounding box won't get updated.
... if I am not lost in the code, that is ...
 
@DavidCarlisle I think that should work. Or do you think that we disturb spacing here somewhere:
\DocumentMetadata{uncompress,testphase=phase-III,debug={log=vvv}}
\documentclass{book}
\usepackage{amsmath}
\begin{document}
\makeatletter
\def\pmb@@#1#2#3{\leavevmode\setboxz@h{xxx#3}%
   \dimen@-\wdz@
   \kern-.5\ex@\copy\z@
   \tagmcend\tagmcbegin{artifact}%
   \kern\dimen@\kern.25\ex@\raise.4\ex@\copy\z@
   \kern\dimen@\kern.25\ex@\box\z@
   \tagmcend\tagmcbegin{}%
}

\def\pmb@#1#2{\setbox8\hbox{$\m@th#1{#2}$}%
  \setboxz@h{$\m@th#1\mkern.5mu$}\pmbraise@\wdz@
  \binrel@{#2}%
  \dimen@-\wd8 %
 
@JasperHabicht Yes, but that's also what pgf does - I must have missed a step somewhere
 
@JosephWright Why is the comparison? To prevent \g__draw_xmin_dim from being too large?
 
@JasperHabicht I think I know what's up
 
3:19 PM
@UlrikeFischer I don't think so as there is no math magic spacing here it's all in hboxes positioned with \kern and \raise so the spacing is pretty much in explicit control
 
@JasperHabicht Ah, found a couple of issues
 
@UlrikeFischer can you patch bm from latex-lab or do you want a modification in bm itself to give you a hook or socket or some such ?
 
@JasperHabicht Fix in hand I hope
 
@DavidCarlisle If I see that correct only has to redefine \bm@pmb@@? Imho it would be more sense now to that in latex-lab. I would use sockets in "external" packages only when we are sure which one we really want.
 
@UlrikeFischer OK sounds good to me
 
3:42 PM
@JosephWright Cool! I'll check later, but I will =)
 

« first day (4863 days earlier)      last day (61 days later) »