@UlrikeFischer Wow! <3<3<3

@samcarter how does the hairstyle look?

@UlrikeFischer ooh

@DavidCarlisle: LuaTeX with HarfBuzz

@samcarter ^^^^ the mohican Mr T-duck version.

@UlrikeFischer ooh

@DavidCarlisle I see we have movement on geometry

@JosephWright do you want write access :-)

@DavidCarlisle I don't think that's necessary :)

Need to book flights to Brazil ...

@UlrikeFischer very nice!

@JosephWright this l3build config stuff gets complicated: github.com/davidcarlisle/geometry/blob/master/build.lua

@DavidCarlisle :)

@DavidCarlisle With the new +x approach, we could actually skip the file entirely and have --module=xxx (or nothing at all for tests)

@JosephWright I thought there was a no build.lua version but I couldn't remember if we had just talked about it or if you'd done that already

@DavidCarlisle I've not done it, just thought about it

Now I have to remember what I was going to change

@DavidCarlisle Frank was suggesting l3build init would make a skeleton build.lua

@samcarter we had cheese today:

@JosephWright is there a good l3build way to check that pdfinfo (or anything else) says the result of this file has size 100x200 postscript points?

\documentclass{article}

\usepackage[papersize={100bp,200bp}]{geometry}

\begin{document}
x
\end{document}

@DavidCarlisle Would have to be a PDF-based test, which I really need to rework

@DavidCarlisle I've not tackled this as it's not been a priority: I guess it is now :)

@JosephWright yes but I don't really want to compare against a reference pdf in this case, I just want to test the metadata in the file

@DavidCarlisle There was the question about testing PDF streams ...

@JosephWright not really a priority, I have run the above in pdftex and luatex and checked what pdfinfo says, so that's sort of fine, but I just thought I'd add it to your stack of jobs, when you have finished l3draw:-)

@DavidCarlisle :)

@DavidCarlisle l3draw will be done by TUG2018 ...

@UlrikeFischer know anything about the geometry package?

@DavidCarlisle First I need to sort out __ business

If any of you have files using geometry and want to try breaking this, please do...

github.com/davidcarlisle/geometry/commit/…

@UlrikeFischer Oh! Was it good?

@JosephWright blurg I suppose I ought to detect old luatex and make if follow pdftex in that case.

@DavidCarlisle sure, one can't avoid to look into it from time to time.

@samcarter very. The cheese situation in our town is quite depressing so from time to time we have to get some from happier towns ;-)

@UlrikeFischer I'll ask you what I just asked Joseph then...

@DavidCarlisle Hm. I don't remember any open problems with geometry apart from the luatex stuff you handled already so I won't probably break anything ;-).

@UlrikeFischer yes plan is just to add luatex support, but preferably not break anything:-)

@DavidCarlisle can't you search for /MediaBox [0 0 100 200] in the uncompressed pdf?

@UlrikeFischer only if Joseph lets me.

@UlrikeFischer point is can I make l3build check do that (or anything else)

@DavidCarlisle Log an issue

@JosephWright OK

@JosephWright done

@PauloCereda Are there any arara rules for htlatex?

@ChristianHupfer Please consider adding your pizza-duck to tex.stackexchange.com/q/422989/36296

@samcarter I was looking at the tikzducks code and it seems kind of unmodular: to add a new duck I need to modify the style itself. Is this right or am I missing something?

@samcarter I was expecting a duck answer here

@UlrikeFischer If you ever happen to be near the origin of this cheese, there is whole shop in the city centre waiting for you :)

@samcarter I know. My husband was there last year and brought back some cheese ;-)

@AlanMunn If I simply want to have multiple ducks in one tikzpicture, I wrap them inside a scope and shift this to the position I want. Or do you want to define new duck shapes?

@samcarter my husband says you should exclude animal experiments as answer ...

@samcarter I meant adding new duck style.

@DavidCarlisle I was too busy reading loopspace.mathforge.org/CountingOnMyFingers/Ikea

@UlrikeFischer Or allow only those experiments which did not harm any ducks!

@AlanMunn Do you have any suggestions how to improve the package?

@samcarter you could use my picture mode ducks instead of tikz

@DavidCarlisle Your aeroplane has an even bigger potential - think of all the applications in aerodynamics!

@samcarter true

@JosephWright @UlrikeFischer geometry pushed to ctan

@DavidCarlisle What is a duck answer?

@AndreasStorvikStrauman Lately, just an answer. :) But more properly, any answer which somehow manages to use the tikzducks package in its example code.

@AndreasStorvikStrauman There are a number of people here who are somewhat obsessed with ducks.

@AndreasStorvikStrauman But it's not yet a requirement for participation. :)

@AlanMunn THAT PACKAGE. It's amazing.

@AndreasStorvikStrauman Then you'll fit right in. :)

haha :p

@DavidCarlisle Cool

Are spaces allowed in package options?

@JohnDorian I suspect it depends on what package you use to process them.

@AlanMunn Ah OK. So it seems without specialized packages, only \DeclareOption and \ProcessOptions, they are not allowed.

@JohnDorian With the standard method, the spaces are removed. So if you have \DeclareOption{option with spaces}{...} the log shows Unused global option [optionwithspaces].

@AlanMunn Thank you :-)

@JohnDorian The most likely candidate for a package that would support them would be @JosephWright 's pgfopts, since pgfkeys definitely supports spaces in keys.

@JohnDorian And here's the definitive ansswer: No. An important point to notice is that the key names do not contain a space. This is because the LaTeX kernel removes spaces from options before they are passed to the class or package. Spaces can occur in the path to the key, but not in the key name itself. This restriction only applies to keys used as options.

@AlanMunn We revised things a bit, but the kernel space stripping can't be got around

So, uploaded a new version of standalone to CTAN today. Has a couple of bug fixes

@samcarter Done! ;-)

@JohnDorian they are not an error, but are ignored

@JosephWright fmtutil [INFO]: Failed to build: 2 (pdftex/pdfcslatex pdftex/cslatex) hmmm

@DavidCarlisle File loading order business (again)?

@JosephWright haven't checked exactly, i assume it's because I changed default encoding and the cs* variants assume the enctex stuff, texdoc cslatex says The CSLaTeX package is obsolete and it will be maintained no more but texlive builds it by default ... I'll sort it out, maybe tomorrow

@DavidCarlisle It's not supported by it's author, or by us ...

@JosephWright that doesn't necessarily imply I should break it (I'm also wondering if I'll need to add guards for *ptex variants, I'll see....)

@DavidCarlisle Ah, now pTeX is different

@JosephWright I was thinking I'd get it checked in and tested as far as possible but then open a gh issue and ping the usual suspects and ask for review

@JosephWright I think I'll skip if \mubyte is defined.

How does one determine what revision of the TeX Live sources' upstream SVN repository one's local TeX Live installation corresponds to?

(I'm on MacTeX 2017, if that helps.)

@RandomDSdevel Barring very exceptional cases, the TeX engines are never updated during the lifespan of a TeX Live release.

@egreg Hmm, tug.org/svn/texlive must be the entirety of TLnet, then. I'm just trying to figure out what the SVN revision number was when it got frozen for the year (the repo logs aren't being quite as helpful as I'd hoped.)

The development schedule listed at the bottom of tug.org/texlive say March 5th was the freeze date, which jives with me updating to that revision via TeX Live Utility within the next few days, but the first commit message I'm seeing in the repository's SVN logs is March 1st.

Either that, or I'm misinterpreting things.

I see a lot of activity recorded in the logs for March 5th, but that seems like it might be post-freeze.

(The reason I need a manual, out-of-line copy of the TeX Live sources is tex.stackexchange.com/a/422246/71079. I'd clone the entire repository's complete history, but I would still have to figure out which revision to check out even then and know from bitter experience that it's…rather large and thus takes too long to clone over my network, particularly since it tends to drop at night.)

So I should just check out the 2017 branch, then, and leave it at that, instead of trying to get something that matches what I already have on my system?

@RandomDSdevel Sorry, not an expert in SVN

@RandomDSdevel probably, it depends why you need sources for a frozen system

@egreg (Nods, shrugs.)

@RandomDSdevel oh bibtex, well that's not changed for years so it doesn't really matter which source you get I would have thought?

@DavidCarlisle OK, cool; I'll just check out the 2017 branch, then, and not worry about whether or not anything else in it won't mirror what I already have installed on my system. Thanks! (Wanders off to tinker…)

Does Heiko's answer at tex.stackexchange.com/a/416779/4301 work for anyone else. I seem to get Undefined control sequence: shortintertext ...th }\normalbaselines \if_dim:w. I thought perhaps there was a package update that I did not yet see so thought I would wait but it still fails the same way, so wondeirng if there is someting wrong with my setup.