« first day (4870 days earlier)      last day (53 days later) » 

12:05 AM
@AlMa1r up to you, I'm not desperate to get more points, so long as they don't go to egreg:)
12:27 AM
@UlrikeFischer excellent choice
@DavidCarlisle Alright. This time you get the checkmark, and next time, he gets the checkmark. Interleaving semantics :-).
7 hours later…
7:14 AM
@UlrikeFischer His phrase "Do ye the little things in life" ("Gwnewch y pethau bychain mewn bywyd") is today a very well known phrase in Welsh (en.wikipedia.org/wiki/Saint_David).
8:08 AM
@DavidCarlisle oh no :)
@DavidCarlisle we have a preliminary code to generate a native image, but it's still really in early stages :)
1 hour later…
9:15 AM
@PauloCereda breakfast
@DavidCarlisle oh no
@UlrikeFischer BTW in case I did not mention it, the reason why I break up into fontspec and pdflatex is then I can support a path=dir feature on engines that support fontspec making it easier to use these fonts on OverLeaf for example. (adding a texmf tree is possible on overleaf but I would nor call it easy). Just checked the rest of the fonts work on overleaf under both xelatex and lualatex.
9:50 AM
@daleif well sure, but you make it more difficult to use it, if you have to do different things to install and use it depending on your engine. Does it have to be a font? Can't you make 4 pictures?
10:03 AM
@UlrikeFischer the whole point of the university to make a font of these four logos is then the user can control the color. And as mentioned lualatex works fine. I just need to know what the xelatex equivalent to \symbol{61488} is and if we can tell xelatex not to complain about a missing U+0020 in the font.
You should see som of our templates for Word, they seem to use a bitmap image for some of these logos and you can image what that looks like when printed.
I have a LaTeX version of our Letter design what can draw the logos via tikz, but that can be kinda heavy when some of the logos are very complicated. So I'd like to use the actual fonts in the future.
BTW the fonts are freely available, one can just not redistribute them.
@daleif well without an example it is difficult to test ;-)
medarbejdere.au.dk/fileadmin/www.designmanual.au.dk/hent_filer/…, it has several fonts, here it is AULogoReg.ttf which I renamed to AULogoRegular.ttf to have a consistent naming
% this does not work
% thus works under lualatex, the number came from lua cache

In the font the names of the four chars are uniF030 to uniF033
10:25 AM
Ohh, that is interesting, xelatex now comlains about \symbol{61488} : Missing character: There is no  (U+F030)
10:36 AM
@daleif the missing 0020 seems to come from the engine when you define the font:
\font\test=[AULogoReg.ttf] xx
@UlrikeFischer I know, apparently xetex expects all fonts to have a space char.
@daleif you can use the symbols like this. But I don't think that you can use symbol.
\XeTeXglyph\XeTeXglyphindex "uniF030"
\XeTeXglyph\XeTeXglyphindex "uniF031"
\XeTeXglyph\XeTeXglyphindex "uniF032"
\XeTeXglyph\XeTeXglyphindex "uniF033"
@UlrikeFischer that was what I was looking for just could not figure out how to combine those information. It is no big deal that we use these under xelatex, I define the same macro for each anyways so uniF030 is just \AULogo. The code for the seal is looooong when converted into svg.
And we'll just have to accept the warning about the missing space char.
The font probably also have an "unusual" size, \fontsize{1cm}{1cm}\selectfont\XeTeXglyph\XeTeXglyphindex "uniF033"\rule{1cm}{1cm}
The next test is of course to see if it works when I add the texmf tree that has the mapfile for pdflatex.
But that is for another day
11:06 AM
@UlrikeFischer brilliant suggestion:
20 hours ago, by David Carlisle
@daleif you can/should access the glyphs by name rather than private use slots?
@DavidCarlisle I know, but could not figure out the syntax needed.
Now that I know that, why is this not possible?\newcommand\AULogo{{\logoFont\XeTeXglyph\XeTeXglyphindex "uniF030"}}, the line break is part of the syntax?

{\tracinglostchars=-1 \logoFont
\XeTeXglyph\XeTeXglyphindex "uniF030"
\XeTeXglyph\XeTeXglyphindex "uniF031"
\XeTeXglyph\XeTeXglyphindex "uniF032"
\XeTeXglyph\XeTeXglyphindex "uniF033" }

@daleif ^^ no warnings about space (even in the log)
@DavidCarlisle yes I stole from a master mind
@daleif leave a space between the " and the } see my example above without the space before } you get an end occured in a group error (because xetex is weird)
@UlrikeFischer saintly
@DavidCarlisle ahh, yes I did get a warning about levels. That \tracinglostchars is that only xelatex?
11:18 AM
@daleif no its classic texbook tex but a new level 3 (to make it an error) was added to all engines last year.
It seems that now that these glyphs work xelatex no longer complains about the map file from pdflatex (more testing needed)
@daleif no that is in all engines and gives the missing chars warning/error. But by setting it to -1 you suppress the warning
@DavidCarlisle That is better. I missed the tracinglost inside the body in the first go. It seems to be all working now. It will be released on our website (we have b een given permissing to repack and rename the font files as long as it is only available from an AU website)
1 hour later…
12:25 PM
\documentclass[border=10pt, multi=draw]{standalone}


\cs_new:Npn \mytext_draw_circle:n #1 {
\draw_transform_shift:n { #1 }
\draw_path_circle:nn { 0cm , 0cm } { 1cm }
\draw_path_use_clear:n { stroke }


\color_stroke:n { black!20 }
\draw_path_grid:nnnn { 1cm } { 1cm } { -5cm , -5cm } { 5cm , 5cm }
\draw_path_use_clear:n { stroke }
\draw_path_circle:nn { 0cm , 0cm } { 2pt }
\draw_path_use_clear:n { fill }
^^^shouldn't all three circles overlap?
12:55 PM
@JasperHabicht Can you make a pgf equivalent? Usually that's the best way to see where I've messed up :)
@JosephWright Right. I'll do, check and then (if the result really differs) directly file an issue ...
@JasperHabicht Thanks
How to make something simple much more complicated: pbs.twimg.com/media/GHiJ6IYXMAA3Wtz?format=png&name=medium
@JosephWright Maybe one day I write a perl or python script that automatically converts PGF to l3draw and vice versa ... =)
@JasperHabicht should not be that hard
1:11 PM
Result from PFG:
Oh no =D
I'll compare agan ... just to make sure
1:22 PM
PGF, of course ... long day, so maybe I am just not getting things right ...
@JasperHabicht Thanks
@daleif When I was a student (some twenty years ago) there was a powerpoint presentation of that :-)
The statement was however slightly different: let us make this trivial expression mathematically more sound.
@campa Then we are probably the same age
3 hours later…
4:37 PM
@JasperHabicht Got it
@JosephWright So, it was indeed a bug? I was unsure, honestly, because I haven't discovered that earlier
@JasperHabicht Now working out waht to change: I tried something quick, that wasn;t right - working on it
5:11 PM
@JosephWright At least for the example your fix seems to work
@JasperHabicht :)
@JasperHabicht I'd got the matrix entries mixed up
@JosephWright Nice the package becomes more and more flawless ✨
@JosephWright I had the idea to maybe add some notes to the documentation about the logic of using tuples, like you already put some notes into the code. What do you think?
There are some notes on that already, but a bit sparse, I think
Also, I found the command \draw_point:n quite helpful to make sure that some tuple expression is parsed as 2-tuples. But the command is not mentioned in the doc, as far as I know ...
But I understand that this is not top prioritiy at the moment. first get the code working, right? =D
5:49 PM
@JasperHabicht Cool
@JasperHabicht From memory, that's one I've been on-and-off about including
@JosephWright It does not much and you won't need it normally, but if you want to do more complicated things, such as calculating trigonimetric things etc, it comes in handy. It could be mentioned in this "how to work with this code in more complicated contexts" section.
I'll think about this again in more detail. I think I could help with this, but someone who really knows how things work should probably assist =)
3 hours later…
9:20 PM
Feb 28, 2022 at 23:09, by Ulrike Fischer
user image
@CarLaTeX ^^^ Bär knows
9:52 PM
@UlrikeFischer Yay!
1 hour later…
11:09 PM
@daleif @WillRobertson if you use the epigraph documented flushepinormal in memoir it errors since 2002 ... github.com/wspr/herries-press/blob/master/epigraph/… vvv
A: Getting rid of overfull in epigraph

David CarlisleYou are indenting a full width rule by \parindent (15pt) so it is too wide by that much as can be seen in the output. You should never need ragged2e justify the epigraph manual documents flushepnormal should be used but that gives an erro rhere as the package is not really loaded, the log shows ...

11:29 PM
@DavidCarlisle flushepinormal? Sounds as if some german ate the spaces ...
@UlrikeFischer australiansdontlikespaceseither

« first day (4870 days earlier)      last day (53 days later) »