@barbarabeeton :D So funny! Thanks for sharing!

I will just leave this here: twitter.com/FeelGoodPage11/status/1385873308961280006?s=20

@daleif I can't believe it's true

@CarLaTeX just imagine being the photographer: No-one is going to believe this.... I haven't seen anyone debunking it yet.

@daleif If it's true it's really astonishing

@CarLaTeX What I want to know is whether it was a case of "Dude, I've got a crazy idea..." or "Lol, this will be funny..." :p

@Plergux :)

... hey folks. I have a meta-related question: In tex.stackexchange.com/questions/1319/… I once saw a pretty neat thesis/project document from someone at Aarhus University. I recall that it had some really great looking graphs, all set in a sans font. IIRC it was typeset with strong similarities to the classicthesis layout. That is pretty much all I can remember. Would someone happen to which document I am talking about?

I have been looking everywhere, also on google with "aarhus site:tex.stackexchange.com", but so far I haven't found it. :(

@henry any idea what the subject was about?

@barbarabeeton FYI, Will has contacted me about the sans non-bold Greek.

@samcarter_looks_forward_TUG'21 looking at another of Sandra Boynton's videos I found the following snippet (in youtube.com/watch?v=Z1f9b7sX_XY)

@UlrikeFischer :) that's a very reasonable statement!

@UlrikeFischer The cow video is from mid 2017 - matches perfectly with the ducks first getting something together for xmas '17 :)

@samcarter_looks_forward_TUG'21 ;-)

@Rmano TEXDOCS=/home/romano/texmf/doc// texdoc -l circuitikz should work (you need to add "/doc//" at the end)

Hello!

Can I ask how to enter the coordinate (0, 2^(37)) in pgfplots?

Something like \addplot+ [blue, mark=o] coordinates {(0, 2^(37))};? I am not sure if I did it right as there are some errors about this.

@wtsnjp YES! Thank you!

@soupless \addplot+ [mark=o] coordinates {(0,0.5e12) (1, {2^(37)})};

You have to hide the () inside the coordinates with {}

I don't know why it didn't work for me before. Thanks!

I'd appreciate any thoughts on github.com/josephwright/siunitx/issues/371 - siunitx issue about uncertainty output

@JosephWright I'm not sure how to express my uncertainty here, could you implement an (eeeerrr) factor? ;-)

@UlrikeFischer 1.34(ish)

but we should make the ish i18n, which I think it's a lot of work

@JosephWright my 2 ¢ over there. I like things that even non-expert can understand, so I normally refrain from anything different from the 1.2±0.2...

@Rmano I tend to think the \pm syntax can look like a tolerance rather than an standard uncertainty, but of course it very much depends on the field

@JosephWright yes, this is also right. But your proposal seems to cover every usage

@Rmano That's the plan: key for me is that having a single separate-uncetainty can't cover it, so I'm best having uncertainty-mode and filling in as many 'obvious' cases as I can

@daleif Only rough and nebulous memories... I think some graphs were either dealing with flight patterns and/or signals...

And I recall that some graphs (i.e. their lines) had a dashed or dotted style. Looked very neatly.

(Or was it Aalborg... hm. :/)

@UlrikeFischer I've adding a few tweaks to v3 of siunitx: I might go for beta-2

@JosephWright my statistic didn't break (and it is faster now).

I just converted my CV which is mainly a bunch of longtables and bibliographies with biblatex to .odt using tex4ht. Very impressed with the conversion.

@daleif Or it might have been a project work thing.

@MarcelKrüger I moved that mathml intent document, we decided they were not supposed to be so hidden mathml-refresh.github.io/discussion-papers (@UlrikeFischer's sword came in handy)

@DavidCarlisle ;-)

@UlrikeFischer as you are here, I was a bit tied up with work stuff last week, what do I need to try the mathml examples you and Marcel have concoted?

@DavidCarlisle you need Marcel's code from here git.math.hamburg/zauguin/mathml, and this example: github.com/u-fischer/tagpdf/blob/develop/experiments/… (it is a bit crude, I'm just wondering how to improve the pdf object part).

@DavidCarlisle or you can use the test file in Marcel's repo (mine tried to add it as associated file).

@UlrikeFischer thanks let's see...

@UlrikeFischer I ran lualatex-dev test_tex.tex got no error but got no mathml either?

@DavidCarlisle The current version writes it into a PDF stream object and only hows you the object number.

You can open the PDF in an editor to see it, but I'll add a command to show it directly. Wait a minute.

@MarcelKrüger ah I saw the numbers:-) OK so I should count that as a success and try Ulrike's test over that?

@MarcelKrüger ah yes OK see it in the pdf thanks

@MarcelKrüger why can one use \luamml_get_last_mathml_stream:e{} only once? at second use I get an error.

@MarcelKrüger only two <math elements?

@UlrikeFischer Because the object is written then and you can't set the directory content after the object is written.

Also the idea was to discourage people from using it long enough after an equation that it might accidentally get reset in between.

@DavidCarlisle What do you mean?

@MarcelKrüger if I look in the pdf I see two math elements but if I look in the tex there are three \[ and align ?

@DavidCarlisle Add \ShowMathMLObj after every equation you want to have converted. (Objects you never ask for never get created)

@MarcelKrüger ah I see, I hadn't realize that it also writes the stream, luatex like syntax ;-).

@MarcelKrüger ah OK I thought that was just showing the numbers in the log not actually doing the convertion, thanks

@UlrikeFischer v

@UlrikeFischer sorry ignore that, needed to install tagpdf as well

@MarcelKrüger should the stream also write /Type /EmbeddedFile?

@UlrikeFischer so how do I see the associated file thing? this? << /Type /Filespec /AFRelationship /Supplement /EF<</F 20 0 R/UF 20 0 R>>/F<FEFF006500780061006D0070006C0065002D0069006E007000750074002D00660069006C0065002E007400650078>/UF<FEFF006500780061006D0070006C0065002D0069006E007000750074002D00660069006C0065002E007400650078> >>

@UlrikeFischer I couldn't see any reason why it would be useful to have an explicit type, but I can add it if you prefer it.

@DavidCarlisle the associated file is in the structure: The /AF key here (which should point to the filespec dictionary:

<<  /Type /StructElem /S /math /P 17 0 R /K [<</Type /MCR /Pg 25 0 R /MCID 0>>  22 0 R] /AF 18 0 R /NS 13 0 R  >>

@UlrikeFischer hmm and which pdf readers can read this? (xpdf and firefox don't seem to show anything)

@DavidCarlisle ngpdf.com can use it (but it can't really be called a PDF reader)

@DavidCarlisle that is the open question ;-). Neil believe that adobe will support it soon. But you can try this ngpdf.com, you can upload the pdf there and then go to the editor and look at the html output.

@UlrikeFischer ok or I stick to emacs and just get better at reading the tag structure, which is probably good for my eductation?

@DavidCarlisle ;-). I mostly look at the pdf. But I think we need also some lua code which writes out the structure overview. Shouldn't be so difficult.

@MarcelKrüger which math environments are currently supported?

@UlrikeFischer Mostly align*/align and everything which is really just a primitive displayed equation (so equation/equation* probably). Everything which shares code with align (alignat/xalignat/...) will act as if it were an align.

@DavidCarlisle I added '\tracingmathml` to make it easier to look at the output. Basically \tracingmathml=1 writes all equations to the output which are also written into PDF streams, while \tracingmathml=2 also prints equations which were never asked for.

@MarcelKrüger can one catch inline math if one use a command e.g. \inlinemath{a=b}?

@UlrikeFischer If the command uses \luamml_get_last_mathml_stream:e after the equation (so after the closing $) it should just work.

@MarcelKrüger wow, it does.

@MarcelKrüger ah thanks I got sidetracked reading Japanese

Two things I love: benchmarks and sarcasm.