TeX, LaTeX and Friends

1:15 AM

@cfr Quack! <3

@cfr: How are you? :)

@egreg: I always forget to rename the variables too, sorry. :)

cfr

1:40 AM

@PauloCereda Still here, I guess ;) I feel as if I've been updating a set of slides on abortion for at least a year, although I know that cannot possibly be the case. And you?

6:54 AM

Morning all.

7:57 AM

@JosephWright As the author of achemso, what do you think about that question? latex-community.org/forum/viewtopic.php?p=93493#p93493

@Johannes_B 'You are using the wrong class: the AIP journals require REVTeX'

@JosephWright Make that an answer ;-)

@Johannes_B You've already said that

@JosephWright ;-)

9:08 AM

@cfr In a hurry, but fine. :) Trying to race against time and work on my thesis draft. :)

Christian Hupfer

@PauloCereda I will not buy this thesis -- it's not finished :-P

@ChristianHupfer ooh :)

@PauloCereda Your thesis is full of eels

@egreg <3

9:48 AM

@DavidCarlisle I wonder if it would be worth looking at setting up Travis for L2e ...

10:11 AM

@JosephWright posibly or just simply suggesting on luatex-list that they take the test suite from svn and run it locally before release. having no regression testing at all is not really funny any more.

@DavidCarlisle True: it would need to be set up to test for their changes (which we'd miss on a simple Travis arrangement)

@DavidCarlisle By the way, did you know that with open_any=p LuaTeX opens no file unless in the current directory or below? :)

@egreg Yes, saw that: another one for the LuaTeX list

@JosephWright A bit too paranoid. :)

@egreg So which of us is going to report to Hans? @DavidCarlisle, @UlrikeFischer, me, ...

10:24 AM

Hello everyone :)

@WillRobertson Hi

@JosephWright Wasn't sure whether to expect people here saturday

@WillRobertson Hi!

@WillRobertson Thesis writing. :) I need @David to keep prodding me.

I see the LuaTeX quality assurance team is hard at work

2

(You can interpret that a couple ways...)

@WillRobertson you can't open files, you can't hyphenate words and you can't do any math, but apart from those minor issues, quality is assured.

10:34 AM

@JosephWright Ask them to install subgit? :)

@WillRobertson Indeed

@DavidCarlisle :-)

@JosephWright I don't think it's that crazy a suggestion.

@WillRobertson Probably not: there's already an unofficial GitHub mirror

@WillRobertson they don't really need to track our changes much they could use git or svn or simply take a static copy they just need to run the test suite on their changed files

@DavidCarlisle I think @WillRobertson meant that with SubGit we could arrange to have Travis do the job each time they commit

10:48 AM

@DavidCarlisle I was probably conflating the idea of subgit and the travis stuff

Does subgit actually give us anything though? Now I think again

@WillRobertson Travis and similar work with Git not SVN so you need some Git version

@JosephWright headdesk. I had been thinking of "git svn", but I think that bridge is crumbling.

(headdesk me)

@DavidCarlisle — now what on earth is this useful for (from 2e):
\expandafter\DeclareMathDelimiter\@backslashchar
{\mathord}{symbols}{"6E}{largesymbols}{"0F}

@WillRobertson Which file?

@WillRobertson git svn is not so useful, but SubGit works fine

@JosephWright fontdef.dtx

@JosephWright Yes :)

@JosephWright But is LuaTeX third-party hosted somewhere anyway? I guess they can't just install it.

@WillRobertson Not sure what the deal is (they might have it anyway)

10:58 AM

@DavidCarlisle — I guess just so that people can do this if they really like:

\documentclass{article}
\begin{document}
\makeatletter
\[ \left\@backslashchar\frac{x}{y}\right\@backslashchar \]
\end{document}

@WillRobertson You know, for \[\begingroup\escapechar=-1 \expandafter\endgroup\expandafter\left\string\\\frac{1}{2}\right/\] ;-)

@WillRobertson an extendable coset thingy \left. H\right\@backslashchar although a \mathchardef

@WillRobertson Would be possible to arrange an SVN cronjob then to do SubGit on some other machine

@egreg ooh a cheering fraction! \o/

3

10:59 AM

@egreg It's funny there's support for it but no interface

@WillRobertson Just do \[\left\backslash\frac{1}{2}\right/\] and you'll get the same

although why we have that as well as

\DeclareMathDelimiter{\backslash}    % for double coset G\backslash H
   {\mathord}{symbols}{"6E}{largesymbols}{"0F}

@WillRobertson I'm not sure, I suppose to match < and \langle

@DavidCarlisle Probably just one of those things that slipped in :)

@WillRobertson it's from plain TeX \delcode`\\="26E30F

@DavidCarlisle Ahhh. I guess that explains it

Ulrike Fischer

11:08 AM

@JosephWright Hans will probably say that he doesn't use kpse anyway ... (side remark: if you start context with --safer it doesn't run). But I found an old thread in the texlive list with this problem tug.org/pipermail/tex-live/2011-February/028612.html. Perhaps Akira knows something about the details.

@DavidCarlisle I saw the words coset here and I cannot explain the excitement it generated in me!

@kan Ciao!

(As we speak, I am writing some notes for my number theory lecture and I am considering some double cosets to explain some interesting prime decomposition law in extensions of number fields!)

@egreg Hello Prof. EGreg! :-)

@kan How are you doing?

@kan Still in freezing Ontario?

@egreg I am doing fine! Yes, I am still in Ontario, but I might be moving to the more freezing Chicago! :-)

I am really excited, but I have been waitlisted at UChicago.

@egreg And, thank you for asking! It really feels nice to be back! I hope to frequent here again!

11:22 AM

@DavidCarlisle Thanks ;0

@kan :)

@kan glad to be of service:-)

11:40 AM

@kan: quack!

@PauloCereda Quack quack! Hello Paulo!

11:56 AM

@kan Hi! :)

Manuel

Hello chat.

About \verb in movable arguments, what if \verb input the whole .tex file and then somehow just took its argument and removed the rest of the code. Is there a clear barrier in this approach? If that were possible, we would have a completely flexible \verb command (still not expandable though, but that's less of a problem).

@egreg: I updated my code, this time without properties list. Only with divs and mods. :)

Manuel

Okey, silly idea :) I'm a bit asleep.

@Manuel Quack! <3

Manuel

@PauloCereda :)

12:10 PM

@Manuel don't talk to @PauloCereda he's no time to chat he has to write his thesis :-)

@DavidCarlisle ooh.

Jan 12 at 12:15, by David Carlisle

@PauloCereda finished your thesis?

@DavidCarlisle This will haunt me for a while, won't it? :)

@PauloCereda I wouldn't have, but you did request to be prodded a few lines up....

@DavidCarlisle <3

@David: you and @barbara have an important mission. :)

1:26 PM

@JosephWright — (l3build) hmmm, lua files that need to be executed as part of an unpack process (potentially after having been extracted from a dtx)? Just load up the .ins file with \immediate\write18 ?

@WillRobertson can't you add the execution to a local function in your local build.lua? (rather than enabling write18?)

@DavidCarlisle Yes…but then I'd need to overload unpackbundle() :)

I don't think there are convenient hooks to do that yet

On a totally unrelated topic: it looks like I should track LuaTeX changes properly. How are you keeping up to date? Checking out svn and building from source?

Q: How to exclude ligatures from hyphenation, if their constituent characters have non-zero lccode?

@WillRobertson er hook I had in mind was "over-write some l3build function to do something" :-)

meanwhile (@JosephWright) while testing this

0

Is there some trick to make LuaTeX hyphenate ``xxaaxx'' in the following example? \language255 \righthyphenmin=2 \lefthyphenmin=2 \hyphenation{xx-aa-xx} \parindent=0pt \hsize=1pt \hfuzz=100pt \lccode`\'=`\' \hskip0pt ``xxaaxx'' \bye

it seems that hyphenation only fails for language 0, other languages seem to hyphenate OK.

@WillRobertson more or less, actually I rsync from the texlive head rather than use svn. Joseph uses a pre-built binary from Akira's win32 distrib

1:43 PM

@DavidCarlisle I'll try set something up tomorrow. So LuaTeX already (now) has \Uhextensible! Curious whether we could cross-port it to XeTeX in time for the next TL...

@WillRobertson oooh so it has! hadn't spotted that,

@barbarabeeton ^^^^^^^^^^^^^^^^^^^^^

@WillRobertson have you tried it, does it work?

@DavidCarlisle In the mean time I'll probably add the macro version of the arrows to unicode-math… if I get time to update that package the rate things are going (head a bit foggy at the moment).

@DavidCarlisle No, I need a newer LuaTeX :)

Also funny that \Uvextensible wasn't added at the same time

@WillRobertson were you just reading the manual or did this hextensible thing come up somewhere in another context?

@DavidCarlisle I wrote to the LuaTeX/XeTeX mailing lists to mention it

@DavidCarlisle Hans replied very quickly

@WillRobertson ah yes in my mail:-)

1:54 PM

Gotta love them classicthesis questions.

Note that the file classicthesis-config.tex coming with the distribution is just a collection of (badly called) packages with options that you'll probably won't need. Leave it alone. — egreg 1 min ago

@egreg Agreed

2:11 PM

@PauloCereda -- how's your thesis? (you must really be concerned, to be up and working on it at 04:30.)

@egreg -- is it just my eyes playing tricks, or does that backslash have a slightly different slope than the forward slash?

@barbarabeeton Ask Don ;-)

@DavidCarlisle -- thanks for pointer. i'll pass this info on to stephen. (will has been invited to join the stix fonts community. can't see that he's responded yet.)

2:31 PM

@barbarabeeton I replied to the email but I think Stephen needed to add me once I'd done so?

@WillRobertson -- probably so. thanks for joining. when you're on, please share the info about the extension facility (even if it's not yet available in xetex).

@DavidCarlisle, @WillRobertson I guess I need to add more primitives to l3names

@WillRobertson Any idea how the script=math stuff works in luaotfload? I'm not sure if there is lots of complexity or it's a simple 'tell the loader' thing

@JosephWright Sorry, I never looked :)

@JosephWright All the LuaTeX support was done by Khaled and Phillip Gesang, while I was away for TeX for a while

@WillRobertson Yes, sure, I just wondered

@WillRobertson I guess I'll have to ask on the list (which will be fun as usual)

From the discussion site on hyphenation of the german wikibook: A: That should work with T1 fonts, right? Maybe add it to the article. B: "Honestly, i don't have a clue. What is T1?" The user page of B states Unter anderem habe ich das LaTeX-Wörterbuch aus dem Boden gestampft.

@barbarabeeton ^^^^^

2:42 PM

@Johannes_B -- doesn't sound like a happy camper.

@barbarabeeton What do you mean?

@barbarabeeton good afternoon (or morning or whatever). Snow gone?

oof, I should be asleep — good night all

@WillRobertson good night:-)

@Johannes_B -- if he's tossing latex reference books onto the ground (and probably stamping on them as well), he is frustrated.

@WillRobertson -- g'night. sleep tight. (don't let the bedbugs bits.)

2:50 PM

@barbarabeeton I could give you a luatex manual to hold, I'm sure you wouldn't toss that ;-)

@DavidCarlisle -- morning here. still lots of snow, and very cold (20F, -6C), but going to be lots colder tomorrow (-5F, -20C) with wind chills below that (-30F, -35C)

@DavidCarlisle -- don't assume too much ...

@barbarabeeton Translation mistake: He created that one, from the ground up.

@barbarabeeton did you see my ping above Will pointed out that they have just added new primitives to access opentype extensible arrows so exactly what stix2 needs, so it isn't all bad (of course doesn't help unless amsmath actually uses the new primitives) and I haven't tested to see if they work yet, will do that this afternoon possibly.

@Johannes_B -- aha! that'll teach me to look in a real dictionary! thanks.

@DavidCarlisle -- yes. that looks promising. we can hope they become available with xetex soon. and that amsmath, well, ...

3:39 PM

@barbarabeeton thanks.

@DavidCarlisle I'll update the LuaTeX primitive list in expl3 in a bit

@DavidCarlisle -- no, thanks to you. will be offline for a bit; have to reboot workstation, and that takes a half hour or so to get back to a workable state.

Doing template support is fun: latex-community.org/forum/viewtopic.php?p=93511#p93511

@Johannes_B tell you what: Joseph and I will do templates, and you can support lualatex.

@DavidCarlisle I think LuaLaTeX is even more fun, so thanks, but no :-)

@ChristianHupfer I was just writing a comment that the template is old and scratched, but ... question deleted.

4:47 PM

@barbarabeeton Oh it's going fine, I believe. :) I just need to focus, as I distract very easily. :)

6:20 PM

I encountered a problem so complicated that I'm thinking of writing a package only to make the solution easier to use.

@egreg tex.stackexchange.com/questions/292722/… wanna write something?

@Alenanno and 30 years later you'll be getting feature requests for it to be extended:-)

@DavidCarlisle Ahah yeah. Although I don't think it'd be a very useful package. I mean, I doubt it'd be used that much. I have to think about it.

@Alenanno I have packages that explictly say "don't use this" and people use then anyway:-) I wrote afterpage as a joke to myself one day avoiding real work and then I see people using it 25 years later in real documents:(

@DavidCarlisle What's it for?

6:29 PM

@Alenanno sticks stuff into the document at the next page break.

@DavidCarlisle Oh and why do you dislike it? :D

@DavidCarlisle May i print that out in italics and frame it and hang it on my wall?

@Alenanno it's massively fragile: poking into places that a document-level command shouldn't be poking, people put tables with the major result of their thesis into \afterpage{...} and I always fear it'll disappear into a black hole and never to be seen again:-)

6:53 PM

@Johannes_B Done. I only refrained from saying “forget about classicthesis”.

@Johannes_B And got a tick. :)

@egreg upvote deserved ;-)

7:27 PM

Hi all. I have decided to make public my letter to the TUG Board:

https://groups.google.com/forum/#!topic/comp.text.tex/_Ir3CaL_xQs

Please publicize it to other TeX groups as you see fit. Thanks.

3

@Kaveh1000 Noted

@Kaveh1000 Should your address be redacted?

@jo

A while since I posted here!! I thought about redacting Joseph, but better leave fully open. Everyone is welcome to come over to my place. ;-)

@Kaveh1000 This is of course your call: just wanted to be sure you meant this to be visible

@JosephWright very much appreciated, and I did give it some thought, but in this case I thought let's get away from any redactions at all!

8:35 PM

@Kaveh1000 Thanks for sharing this, it is an interesting insight. I see you've decided to go public with the affair, so I hope you don't mind me replying in public. At the same time, this is nothing against your person, it's just my thoughts written in words.

It's impossible to "take sides" in the problem, speaking sites between you and VR, as well as you and the Board. In the first case, because I simply do not know the opinion of the other side, and I have no way to verify that whatever you say is trye. In the scond case, because I believe the board acted in what they considered the right way, but at the same time, I understand your appeal.

The only thing that hasn't been said is, in my opinion: Would you have been elected the TUG president if TUG members in general had known about the lawsuit? While I don't think that the whole thing should affect people's decition for the elections, it could, and I understand that the board felt this information could have been important.

9:17 PM

Hi all

Is there an analog of \nicefrac for right cosets (quotients "on the other side", i.e. H\G)?

9:34 PM

@Danu Hi, I doubt it. However, PLEASE DO NOT DO THAT :) The slash / backslash operators are really just binary operators that get a different spacing; they have got a left and a right operand, not the top left and bottomr right (or whatever) operands.

@yo' Actually, I was wondering whether there exists a "canonical solution" to this (common) problem of typesetting quotients nicely

I've seen some questions on the site about it, but no really good solutions.

But maybe I just missed the right thread.

Do you have any idea?

@Danu the canonical solution is $G/H$ and $H\backslash G$ . It is nice. (Says your math copy editor. :D )

Now I gotta go, or I get up in 8.5 hours and I need to catch some good sleep.

@yo' But I'm dealing with "larger" things, e.g. quotienting $[(I,0),(X,x_0)]$ by $p_*\pi_1(X,x_0)$

@PauloCereda :-D

that standard solution works rather terribly for those cases, in my opinion.

@yo' Would you care to discuss this tomorrow?

9:42 PM

@Danu $\bigl[(I,0),(X,x_0)\bigr]\big/ p_{*} \pi_1(X,x_0)$

@Danu May be. Or ask @egreg, he's among the top experts on math typography in the world IMHO (together with @barbara of course)

I would like some vertical offset :\

So, @egreg, do you have any opinion on this? ;)

@Danu and that's what I consider ugly, removing it from papers I copy ed, unless I see the authors seem to be keen on their typography.

Right.

@Danu I consider \nicefrac and similar good for recipe books, not for math. ;-)

hahaha

I'm still in the "formative stages", I guess.

You both agree there should be no vertical offset?

9:46 PM

@Danu I do.

The next question is: Where did I get the idea that there should?

Is it even standard to have some?

@Danu Note that good typography is a mixture of historical precedence, good reasons and personal taste, so you can do what you wish.

(sorry, my sudden absence was caused by a minor problem: )

@yo' Of course. I wasn't happy with my current (\nicefrac) solution, but as I said I somehow thought the vertical offset was universally agreed-upon.

wut.

9:50 PM

@Danu No, it's not that way. In handwriting, maybe, but in typed material, the offset is superfluous; it's replaced by proper spacing etc. Also, you would introduce a "inbetween" size: the operands would be niether in text size nor in script size

@Danu my thought exactly, the rest of the plug is still in the wall outlet, on the life wire.

@yo' That's crazy!

I gotta go now, will solve this tomorrow (with the circuit off probably)

@yo' Yeah, maybe...

The only annoyance is that I might have to introduce some parentheses.

@Danu indeed. Well, I plug the cable in average more than once a day, I've had it for 2 years....

Things like $X\cup Y/A$ are ambiguous

9:53 PM

@Danu yeah, but note that without them, the text is not really clear...

@Danu indeed they are

With vertical offset it is...

I think that that was actually my original motivation for having it

@Danu either the offset is not enough (to be distinctive) or it's too much (and becomes really ugly)

The latter is happening for me :\

It all gets so small

@Danu I wouldn't consider $B \cup \nicefrac{X\cup Y}{A}$ non-ambigous...

@Danu btw, this is not ambiguous, it's clearly X\cup (Y/A)

now I really go. Good night.

@yo' Yeah, right. But if I naively change my provisional \quot command it won't mean that, but it'll look like it!

9:57 PM

@JosephWright did you notice \letcharcode? Hans is killing off the one use \uupercase has:-) (@barbarabeeton ^^)

@yo' Sorry for delaying you. Goodnight!

@DavidCarlisle No, haven't looked over the list yet. Will do

@DavidCarlisle s/uu/up

@Danu no worries. Night. Night all.

@yo' -- please do flip the circuit breaker or take out the fuse before doing anything! (how did you manage to do that?)

@barbarabeeton I don't know. I plugged in, and then my computer died of low battery. Only then I noticed a problem. And don't worry, I know how to operate electricity safely :) Also, thanks for caring!

10:05 PM

@DavidCarlisle Not sure I see the point (it's not adding anything to TeX, and I thought Hans didn't want new primitives if they could be covered by ...)

@JosephWright mail....

@Danu In the book on algebra I own, I've never seen it.

@DavidCarlisle I don't get it (when we get told we can't have X as it can be done in Lua or ...)

@egreg As yo' pointed out, it may be a difference between handwritten and typeset documents.

@Danu Big.

10:16 PM

@JosephWright basic design of the whole system is "seemed useful to Hans at the time":-)

@DavidCarlisle How many of the new names do we add to expl3? Not all are documented ...

@yo' what do you expect if you use such wimbly plugs, you should use a proper plug like this ^^^^

@JosephWright dunno, generate the list from lua then look. all of them, probably:-) It's probably worth putting the generated list into an l3 test file similar to the callback one I added for 2e so we notice when it changes again

10:55 PM

Thanks for your comment @yo'. I have made my letter public because I felt TUG members were getting frustrated with a lack of information, and because I have not heard from the Board in two months. And in any case transparency is generally considered a good thing, particularly in a group using open software like TeX. I have not made public all my correspondence with TUG but happy to do so.

Regarding "who is right", I can only give you my side. It is up to others to judge.

And regarding the legal case, this has been common knowledge and not a secret. As I have said in my letter, the board w…

11:42 PM

@Kaveh1000 Not a secret? Possibly. Common knowledge? I don't think so.