@DavidCarlisle You are very good at politely describing that something s***s ;-)

@marmot the class removes the functionlity but comes with an example document overleaf.com/read/bpqdkmgxznvh#/11388571 that uses the position argument in line 406. I think we should blame @PauloCereda as the main abnt representative here

@DavidCarlisle I am very fortunate not to know anything about ABNT and prefer to keep it that way ;-)

@marmot :)

@PhelypeOleinik You're so lucky. I wish I had that power. Where I work there are "honors thesis" (TCC) without a committee, but they "demand" ABNT, even if no one is going to read it.

@DavidCarlisle, yes, it's one of it's infinite reincarnations. Just note the 2013 "to-do list": 1. provide hyphenation for words with accents and cedilla. 2. fix problem with \uline and sections

@DavidCarlisle It's not the worst-looking one, but that hopeless environment nesting is at least bizarre.

@Joseph I wouldn't say lucky. It's taking me ages to get something good, stable, and usable for people with little LaTeX experience. At least my dissertation will not have a list of equations :P

@Joseph My situation is really unusual, as most postgraduate programs already have their rules. But the one I'm in was recently threatened to be closed by CAPES so a few teachers decided to take the coordination and make the program grow (after almost 20 years of bad administration).

@Joseph These teachers were the examiners of my TCC and they liked it, so it was easy to convince them to hire me (for free obviously :P) to do it.

@AlanMunn Here is a draft for an opt-out letter.

@DavidCarlisle Finally, I think I'm getting somewhere with the primitives

@DavidCarlisle I think the timer primitives should now work, and I also think I see how to get at least some of the file data ones working

@DavidCarlisle Of the set of primitives in pdfutils.ch, I think I'm now only short \pdffiledump: that one looks tricky so I might try to deal with a few others then come back to it

@DavidCarlisle Should mean though we don't have engine variations from TL'19 (and in MiKTeX even earlier)

I just saw this video about the typesetting system SILE, heavily inspired by TeX. Of course it gets quite a few things wrong about TeX, but it also gets a surprisingly large number of things right, and people interested in typesetting may like to see examples of the problems etc.: youtube.com/watch?v=5BIP_N9qQm4

@ShreevatsaR not seen the video but the author was active in the luatex lists for a while so I looked at some of the early sile examples. Yes it's an interesting system in practice more of a test bed to incorporate some ideeas back to luatex rather than being a mainstream system itself i suspect

@JosephWright sounds good

@DavidCarlisle See team list

@DavidCarlisle I've rebased and squashed the expanded branch into a single commit that matches pdfutils

@DavidCarlisle A bit of rebasing later, everything should be more-or-less logical

@JosephWright so do you have a single branch with all changes now?

@DavidCarlisle Not quite :)

@DavidCarlisle Still have the different logical branches, but they basically work together: xetex-tidy -> pdfutils -> expanded. So if you check out expanded you get all of the changes except the \Ucharcat one.

@JosephWright OK well expanded and pdfutils just got a "forced update" but updated cleanly with no scary conflicts from the rebase:-)

@DavidCarlisle I can run everything together, I guess: we are going to be submitting diffs anyway (can't pull request here)

@JosephWright do you plan to submit to texlive svn or xtetex sourceforge git (or both:-) ?

@DavidCarlisle Good question

@JosephWright watching that sile video...

@DavidCarlisle I will do in a bit

@DavidCarlisle Probably get everything sorted then ask KB

@JosephWright you will enjoy 3min 36 seconds in:-)

@DavidCarlisle I've rebased Ucharcat so it' now got everything

@DavidCarlisle When you can, it would be useful to know if the primitives actually work ....

@JosephWright so I switch to expanded, do a full build then try some tests?

@DavidCarlisle Please

@DavidCarlisle Should have all of the other primitives in and (fingers-crossed) working

@DavidCarlisle We can probably live without \pdffiledump ...

@JosephWright blurg git pull is upset about merge conflicts, I think I'll junk "my" expanded branch, and start again from another one tracking yours anthd en add the test stuff again

@DavidCarlisle Cool

@DavidCarlisle I've cracked \pdffiledump (with a little help from the (u)pTeX commit history): finalising everything now

@JosephWright OK I'm doing a full build now, but I'll pull and do a --no-clean build when you are ready

@DavidCarlisle Cool

@DavidCarlisle It only needed changes to xetex.web so should be safe enough (turned out I needed to declare a local variable)

@DavidCarlisle Once it looks OK, I guess it's time to ping KB :)

@JosephWright oh in mail you said everything is in Ucharcat branch, I'm building expanded right now?

@DavidCarlisle The only difference is the \Ucharcat change (for \active chars)

@DavidCarlisle expanded has all the new features

@JosephWright OK I'll do tests there then seeing as that's quite away through the build

@DavidCarlisle Great: with the bash set up here, actually building is no longer the big issue. But I'm not really set up to use the binaries in anything useful (I'd want w32tex ones)

@JosephWright well same here in a way as I don't have a full tex tree in the subsytem but I'll add some tests suitable for the make system

@DavidCarlisle Ah right

@JosephWright if you run make install then put install's texmf-dist/bin/* directory in your path you have a complete but minimal texlive setup (I copied plain.tex in locally and tried a few tests previously, although for make I write tests just using initex so they have fewer dependencies)

@JosephWright build just finished

@DavidCarlisle I think I'll leave this part to you: it's good to have a second person do this, anyway

@DavidCarlisle Great

@JosephWright sure I'll get something set up. ping me if you've added filehexdump

@DavidCarlisle That's not in the (u)pTeX set, hence the question on the team list

@JosephWright ok

@DavidCarlisle Do you mean \pdffiledump? It's done

@JosephWright i was one update behind I think, should be Ok now

@DavidCarlisle Cool

@DavidCarlisle s/Dentist/bug fix/g ^^^

@PauloCereda Do you still need talks for TUG2018?

@JosephWright Good question. I have to catch up the discussions, but it's always welcome. :)

@JosephWright I need to write a talk before.. tomorrow. Loads of time...

@PauloCereda You got mail ;)

@TeXnician Got it, thanks. :) I will reply later on today, do not worry. :) Lots of things to tell you!

@PauloCereda Oh, I do not worry. You are probably still "arriving back home" :)

@DavidCarlisle Same here: I'm doing Pint of Science. They only reminded me Friday night: I thought they'd dropped me ...

@TeXnician Body is 50% functional, thankfully. :)

@PauloCereda 50% functional??? That does more sound like the positive results of my unit tests, but not like a good percentage in any relation to health…

@TeXnician :D

@PauloCereda are you back home?

@yo' Yes! <3

@PauloCereda good to hear!

@yo' I was afraid in the last day: the flight back to Brazil was delayed in about 5 hours!

@PauloCereda well, I'm flying to Paris next Sunday and Air France is on strike :-/

@yo' oh no

Btw, check this, I'm sure you know it :) music.tomh.eu/arr/162

@yo' ooh very nice, Tom!

@PauloCereda I should make a recording so that @barbarabeeton can enjoy it, too!

@yo' ooh please do it!

@JosephWright modulo a sed script to add or remove \pdf in places this test gives the same results in pdftex, xetex,euptex if \expanded is in the existing test have I missed any new primitives?

\catcode`\{=1
\catcode`\}=2
\catcode`\#=6
\def\typ#1{\immediate\write-1 {#1}}

\typ{START}

\typ{TEST 1: RNG}
\show\pdfsetrandomseed

\show\pdfrandomseed

\pdfsetrandomseed 123456789


\show\pdfnormaldeviate

\edef\bbb{\pdfnormaldeviate 100}

\show\bbb

\typ{FILE TEST 1: filedump}
\message{\pdffiledump length 10 {ftest.txt}}

\typ{FILE TEST 2: filemoddate}
\message{\if !\pdffilemoddate {ftest.txt}!empty date\else good\fi}

\typ{FILE TEST 3: filesize}
\message{\pdffilesize {ftest.txt}}

\typ{FILE TEST 4: mdfivesum}

@DavidCarlisle \pdfcreationdate: it's in (u)pTeX and doesn't seem to be PDF-specific

@DavidCarlisle So we are good to go then :)

@DavidCarlisle What do you think about the names? Are we OK without pdf?

@JosephWright oh yes I'll add a test for that also \pdfuniformdeviate I think I missed. I'll get that added to make file and make a pull request. as adding to the makefile is harder than writing the tests I was planning to have just one test file for all these that could be extended later. I could include the existing expanded test and only have one new test or do you think keeping \expanded separate would be better?

@JosephWright yes

@DavidCarlisle Put it all together: I've committed separately largely so the history makes it clear what was going on, but we want them all to go in together

@JosephWright that simplifies things as I just need to rename the expanded test file (or even simpler would be to not rename the test file but that might be confusing?)

@DavidCarlisle My plan is to make a series of patches for XeTeX (so they can go into the master Git repo keeping the messages), but for pdfTeX it will be just one change file

@DavidCarlisle Perhaps call it pdfutils?

@DavidCarlisle In the pull request, keep the name the same, then I can change it

@JosephWright I'd thought of that

@JosephWright Ok, let me write a couple of more tests then re-arrange for the makefile and check make doesn't fail:-)

@DavidCarlisle All sounds good

@AlanMunn ^^ did you see this clip? :)

@JosephWright yeh, positive feedback on xetex

@JosephWright make no longer terminates:-)

@DavidCarlisle Good stuff

@DavidCarlisle Not so great

@JosephWright oh it runs when I run it by hand so it will just be sed or xetex hanging on file input somewhere, the makefile copies all the files somewhere strange and runs the tests while hiding the log so it's just a matter if working out what it's done. eat first I think.

@DavidCarlisle Sounds good

@DavidCarlisle I'll start on PR requests to XeTeX probably this evening

@JosephWright I hate that makefile:-)

@DavidCarlisle I'm sure

@DavidCarlisle I'm putting PRs in for XeTeX ...

@JosephWright I've worked out why the test is failing in xetex (the input is a zero length file) but it isn't zero length at the start of the run and make -d wasn't exactly informative:-) It must be something stupid I've done somewhere

@DavidCarlisle Hmm

@DavidCarlisle I'm not sure we should worry too much: actually getting this into TL probably means sending a test file and asking them to add directly

@JosephWright yes i was already going to suggest that for the *ptex ones, I checked the test passes for euptex by hand but I wasn't going to try to integrate it as they probably want to manage the interrelations between the variants. for xetex similarly i know the test passes if i run it, so getting the makefile to work is mainly a learning curve for me rather than anything useful. I think I'll back it out for xetex and just give you a PR with the extended test running in pdftex

@JosephWright let's see what travis makes of that....

Back from the mass. :)

@DavidCarlisle sourceforge.net/p/xetex/code/merge-requests

@JosephWright yes I get mails. Meanwhile my Makefile it passed with no-clean so I pushed to github (my fork) where it failed and so I did a full build here and that hangs after doing the tests have to kill make, so I reverted back to your branch and checking things pass again, (full build again:-) but I may need to stop and write that talk...

@DavidCarlisle Much the same here: slides to prepare

@JosephWright xetex will end up getting \expanded before pdftex:-)

@DavidCarlisle I'll sort a .diff for Karl I suppose

@DavidCarlisle Need to do the manual there ... in ConTeXt MkII

@JosephWright you can get @WillRobertson to do the xetex manual...

@DavidCarlisle Sure

@DavidCarlisle oh no

@PauloCereda you don't trust him? (it has his name at the top!)

@DavidCarlisle more work. :)

@DavidCarlisle Nailed it

@PauloCereda Funny ending. But my interest in Céline and superheroes is about zero. :)

@AlanMunn :)

@marmot Thanks. There's also a post from SO saying they're looking into an electronic version, claiming they had no idea the arbitration clause would be so controversial.

@PauloCereda So you're back from PT. Did you have a good trip?

@AlanMunn It was a lot of fun, but a bit stressful in the end. My flight had a 5-hour delay.

@PauloCereda Yes, that can be stressing. Was it a direct flight or did you have connections? That always makes things worse for me; a delayed flight almost always leaves me in a hub with no way to get home.

@AlanMunn Thankfully it was a direct one. :)

@PauloCereda More than once I've ended up in Chicago or Detroit and just rented a car rather than spend a night waiting for another flight. it's very common in the summertime in Chicago because there are big thunderstorms that screw all the flights up.

@AlanMunn oh if it was a domestic flight I'd surely find a way.

Chicago is O'Hare?

@PauloCereda Yes.

That \HUGE airport?

@PauloCereda That's the one. Almost all of our flights go through it, since it's an American Airlines hub and we fly American most of the time.

@AlanMunn Interesting. The flight from OPO to MAD was fine. But then MAD to GRU was terrible. :)

@PauloCereda Apparently MAD is living up to its airport code. :)

@AlanMunn Exactly! :D

@AlanMunn And I definitely hold a GRUdge. :)

@PauloCereda I see what you did there. :)

@AlanMunn <3

@JosephWright sorry it's not as finished as I'd have wanted but spent too long on it today off and on:-)

@DavidCarlisle No problem: submitted patches to XeTeX and pdfTeX sources anyway

@JosephWright yep it was more or less a personal challenge:-)

@DavidCarlisle -- i've just watched it, and agree with his analysis of what's left out of tex. but i have a big grumble about one of his examples of how it can be done easily in sile. just after giving a justification of why grid typesetting is necessary, he shows an example of how to do a drop cap easily that totally ignores adherence to a grid. (might you know how to contact him? with an address, i would like to register a complaint.)

@barbarabeeton well so long as you make your drop cap frame a multiple of the grid spacing i expect it'll be fine.

@barbarabeeton a lightly obscured version of his email is at the top of this posting of his to the xetex list tug.org/pipermail/xetex/2014-September/025539.html

@DavidCarlisle -- of course, and it would have been easy to do, so he should have. thanks for pointer to address.

@barbarabeeton anything about these two in your backlog of amsmath issues? github.com/latex3/latex2e/issues/50 and this one (which has discussion for both) github.com/latex3/latex2e/issues/49

@DavidCarlisle -- i'm on laptop (in alligator country) until next monday, and the wretched thing complains that it can't get a secure connection. i'll try another method later, but now it's time for the dinner bell. i also can't easily get to the bugs list, but if i can see the reports, i may remember whether i've seen them. (getting youthe ams list is really high on my list, but i figured you'd be occupied with getting the kernel straightened out first and would rather be left alone to do that.)

@barbarabeeton no rush, have your holiday:-)

@PauloCereda well, if only you were PRaGmatic :)