5:05 AM
How can I 'print' or download a piece of a complete document from Overleaf? I have something like chapter1 ... chapter2... section 2.1 ... subsection 2.1.1 ... and I'd like to download on pdf only 2.1.1 with their respective bibliography.
Is that possible?
The problem is that if I try to delete all of the rest of the text that I don't want to print, the numeration 2.1.1 goes to 1.1.1, and so on.

1 hour later…
6:22 AM
@JosephWright there already were omitted braces around single-token arguments, e.g., you (assuming that you wrote the entire code of l3keys) used \exp_not:o \l_@@_relative_tl a few times. Should I try to search for those occasions and introduce braces?

@Skillmonlikestopanswers.xyz Those are 'allowed': I was thinking more that places like \__kernel_msg_ is one we don't normally optimise speed; we have 'allowed' o-type arguments to be 'optimised' internally (as we use o rather than V for performance here in the first place)
@Skillmonlikestopanswers.xyz I guess I'm mainly thinking of the x-type expansions, which generally are not 'optimised'

@JosephWright ok, in that case I'll revert the commit and redo it without changing the x-types. Is n-types ok?
@JosephWright I really don't know what the agreed upon optimisations are here :)

@Skillmonlikestopanswers.xyz 'Things that look OK'
@Skillmonlikestopanswers.xyz We've probably not squeezed every possible optimisation in this area
@Skillmonlikestopanswers.xyz Probably one to raise on LaTeX-L to get a wider input, to get an 'agreed' position

@JosephWright since it's kernel code, the number of people who have to agree is rather small, I'd say the team. People who write PRs have to obey your style guide (compare to the policy in this regard Linux kernel).

@Skillmonlikestopanswers.xyz Er, yes, but a lot of the style is really 'whatever Joseph felt was right'

6:32 AM
@JosephWright seems like a good style :) So far most of the code looks good (I'd write some things a bit different, e.g., I don't omit the spaces in { #1 }) but other than that your coding style doesn't seem to different from mine (except for your loops, I write them very different, the reason l3keys is getting faster with my PR)

@Skillmonlikestopanswers.xyz l3keys core was written quite a while ago, and the aim was largely to get something I liked as an interface; I might well do things differently if I started from scratch
@Skillmonlikestopanswers.xyz OK, let's leave your brace optimisation as it is, and see if the rest of the team comment when I merge

@JosephWright maybe ask before merging? :)

@Skillmonlikestopanswers.xyz Probably not worth the time at a team meeting: there is justification for the performance gain as keyval gets used a lot

7:06 AM
@JosephWright I found some more micro-optimizations (all in all just a few \str_set:Nx which can savely be turned into \cs_set_nopar:Npx because they act on already stringified contents)
@JosephWright and \exp_args:Nx is slower than \exp_args:Ne (though the latter seems to have some memory-leak issues, benchmarking \exp_args:Ne \use_none:n { \test } with \test expanding to abc results in an out-of-memory error if benchmarked twice in a row under pdfTeX).

@Skillmonlikestopanswers.xyz Could be: the code is pretty simple, and expansion was never designed to be nested, which \expanded allows: could be some save/restore missing in the WEB
@Skillmonlikestopanswers.xyz One that @PhelypeOleinik and I should look at , I guess
@Skillmonlikestopanswers.xyz I'm not surprised: assignments take time

@JosephWright I'm not surprised either by the speed difference, I'm surprised by the out-of-memory error.
@JosephWright I'll try XeTeX, too.

@Skillmonlikestopanswers.xyz We should try to get that down to an example in plain, then discuss on the pdfTeX list: like I say, realisitically @PhelypeOleinik or me will likely do a fix
@Skillmonlikestopanswers.xyz Code is a straight copy-paste, so the result should be the same

@JosephWright XeTeX also fails
And LuaTeX doesn't fail and seems to give consistent speed, so I guess it doesn't have issues with the memory-allocation here (meaning it doesn't have to allocate more memory, which would take time, I guess).
@JosephWright well, benchmarking is just running the code bazillion times, so recreating in plain shouldn't be too hard... :)

@Skillmonlikestopanswers.xyz That's not a surprise as things were re-written a lot

7:37 AM
\def\zz#1{#1#1#1#1#1#1#1#1#1#1}
\def\zzz#1{\zz{#1}\zz{#1}\zz{#1}\zz{#1}\zz{#1}\zz{#1}\zz{#1}\zz{#1}\zz{#1}\zz{#1}}
\def\zzzz#1{\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}\zzz{#1}}
\def\zzzzz#1{\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}\zzzz{#1}}
\def\zzzzzz#1{\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}\zzzzz{#1}}
\def\zzzzzzz#1{\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}\zzzzzz{#1}}
@JosephWright ^^^^ will err on line 12, commenting out line 11 doesn't change this. Does this suffice as an MWE?
@JosephWright also @DavidCarlisle would be proud.
@JosephWright should we change the \exp_args:Nx to \exp_args:Ne or is this an unnecessary optimization? (Also we don't know how much fixing the bug will affect performance of \expanded, I'd guess not so much that it gets worse than \edef, though)

@Skillmonlikestopanswers.xyz I'd go for it: I'm using a lot more of \exp_args:Ne in newer code, as it's 'just there'

@JosephWright then I'll open a new PR :) With further mini-optimizations :)
@JosephWright is \msg_<type>:nnxxxx implemented with real x-expansion or just by omitting the \exp_not:n in the \message?

@Skillmonlikestopanswers.xyz Real x-type: \message does e-type expansion wrt to #, and we found that 'shotcutting' was a recipe for getting stuf wrong

@JosephWright I was just curious :) No \msg_<type>: should ever be performance-critical :)

8:00 AM

@Skillmonlikestopanswers.xyz unless you use a koma class ;-)

@Skillmonlikestopanswers.xyz If you look through the history, we had 'fun' with things like '\str_if_eq:xxTF' in the past, as it was not really the right type of expansion

@UlrikeFischer how's that? Oh, because of the amount of status messages, I got it.

@UlrikeFischer there should be a KOMA-option shut-up.

8:04 AM
@Skillmonlikestopanswers.xyz What's needed is some way to redirect messages ;)

@Skillmonlikestopanswers.xyz ;-). I just compiled an older document that uses koma along with fancyhdr and it tells me over 11 lines that and why I shouldn't do this.

@UlrikeFischer your fault for using fancyhdr, scrlayer-scrpage gives way better results.

@UlrikeFischer Makes FMi look positiively terse

Can anyone recommend a graphical diff utility that supports word wrap and works on macOS? I am aware of latexdiff, but I am looking for a plain text diff to use with latex files (not simply a content diff)

@Szabolcs vim -d <file-a> <file-b> ? What do you mean with "supports word wrap" ?

8:16 AM
@AlexG If a line does not fit in the window, it should be wrapped, just like in normal text editors.

@Skillmonlikestopanswers.xyz I found a way to suppress the warning: something like this with a list of package before loading the class: \makeatletter\def\class@shoulddisablepackagewarning@list{fancyhdr,tocloft}\makeatother. I should write a shutup-package ;-).

@barbarabeeton No, didn't know you were the keeper of the English hyphenation exceptions --- you sure are doing a lot of TeX-related work (and thank you for that, BTW). Providing instructions/pointers is definitely worthwhile. As things stand it's not obvious to an outsider (like me) where such things as hyphenation patterns/exceptions are being maintained, much less how to report issues or contribute. I'm all in favor of lowering the bar for contributing to the TeX ecosystem, here and in general.

I do in fact use Vim for this at the moment, as it does support wrap (:se wrap), but I find it irritating that it can only scroll vertically full-line by full-line (not wrapped line by wrapped line)

@Szabolcs you can, give me a sec.
" Treat long lines as break lines
map <Down> g<Down>
map <Up> g<Up>
" also in insert mode
inoremap <Down> <C-o>g<Down>
inoremap <Up> <C-o>g<Up>
@Szabolcs put that ^^^^ in your .vimrc and you can move wrapped lines with the arrow keys while jk still move full lines.

@UlrikeFischer Did you look at the color stuff? I didn't get round to the spot-colors-in-drawings, shuld do that today

8:38 AM
@UlrikeFischer just wondering what the best way to detect the new hooks are? \ifcsname AddToHook \endcsname? The file hooks seems to work just fine, though I did manage to find some errors in the memoir manual.

@JosephWright I did look at the commits and saw that you created \color_fill command etc, but didn't get to round to actually test it.

For manuals and stuff what are we calling the version where AddToHook will be added?

@daleif Well you can use \ifdefined
@daleif LaTeX2e 2020-10-01

@daleif I think we decided (to avoid problems with the rollback code) that the best is to test the format version

@JosephWright Ahh right, I was looking at the code for InputIfFileExists (that I will need to remove)
@UlrikeFischer isn't the problem with that (from the memoir side) that users may add an updated memoir to their old system and this a kernel macro that checks format version might not exist (I don't even know if there is on)

8:42 AM
@daleif in the hopatch code that I'm just updating I use this:
\providecommand\IfFormatAtLeastTF{\@ifl@t@r\fmtversion}
\IfFormatAtLeastTF{2020/10/01}{}{\input{hopatch-2016-05-16.sty}}
\IfFormatAtLeastTF{2020/10/01}{}{\endinput}
That should work with old code too.

@UlrikeFischer ahh, nice

@daleif FMi suggestion ;-)

@UlrikeFischer Probably I should add that to beamer

@Szabolcs In normal operation (not in diff mode), vim does wrap long lines by default. Though it doesn't wrap in diff mode, for good reason I guess. But if you still want them to be wrapped, you can get this with :set wrap, once vim is running.

9:23 AM
@ key-value-experts
0

Normally at a genealogytree node I have to say p[]{name=Jane, birth={2-2-1888}{somewhere}} How can I create a key, like persdata=JaneID which expands to name=Jane, birth={1-1-1777}{somewhere}, so that I only have to say: p[]{persdata=JaneID} A not working or correct MWE: \documentclass[landsca...

9:37 AM
@cis I'm not sure if you are doing yourself a favour by constantly pushing your questions into the chat. If I see a question posted in various places it normally diminishs my inclination to look at it.
2

9:55 AM
1. Not everyone is called @UlrikeFischer.

2. I have seen others do this; so I assumed that's so common.

But of course there is always something to complain about.

@cis It's not very common I'd say, and if someone does it a lot I think they often get asked to stop.
4

@cis I don't know anyone regularly posting their own questions here beside you - which is good because if more would do it the chat would become a copy of the main site.
6

10:12 AM
@UlrikeFischer I'm playing a little with the shipout/background hooks. Inside memoir I can easily define trimmarks using \AddToHook{shipout/background}{\trimmarks} (\trimmarks needs to be modified a little), and it my test doc I wanted to try \AddToHook{shipout/background}[bg]{\color{orange}\raisebox{-\stockheight}{\rule{\stockwidth}{\stockheight}}}, this of course then gives a problem as the hook marked bg is executed last.

@cis usually, questions are posted in chat because they require special attention, e.g., because there is a severe issue with it, it should be re-opened, etc. They aren't posted in chat just because they are new. So, yes, others might post questions in chat now and then, but only because of good reasons.

But shouldn't \DeclareHookRule{shipout/background}{bg}{before}{memoir} solve that?

@daleif yes, but the hook is a picture environment, so you should better use picture commands:
\documentclass[]{article}
\usepackage{color}
%\DeclareHookRule{shipout/background}{bg}{before}{trimmarks}

\begin{document}
abc
\end{document}

hmm, seems better to wrap each hook code in \put(0,0){code} then the reorder works as expected

(but the label should probably better be something like memoir/trimmarks and memoir/bg or so)

10:25 AM
@UlrikeFischer yes, that was what I misunderstood from the manual, I thought it was added to such a put by default.

@daleif no, but I think there are few typos in the manual as I just found out ;-)

@cis also this is not for the key-value-experts, but for the genealogytree experts. I don't see myself having any advantage on this question, even though I wrote a key-value package and rewrote/changed many parts of l3keys.

@UlrikeFischer In lthooks section 2.1.3 it says that for hooks in classes or packages it is best practice to not specifying a label (I agree, that I'd rather mave something like memoir/trimmarks as the label to make it more specific

@daleif this is still a bit in flux, what the best practice should be. It also depends on if you expect that the hook chunk will need a rule. I mean if you expect the need to move the trimmarks code around then it is better if it has an unique label, and then it should be something like package/trimmarks (you can use ./trimmarks if you define it in the class).

@UlrikeFischer I think that is best here. I'll probably not do it for the file hooks there memoir seems enough.

10:37 AM
@daleif but if you do it only for the color: As I wrote we want to adapt the \pagecolor command, it is imho clearly wrong currently.

@UlrikeFischer this was just a test to try the reordering (hadn't notice than I had to delay the ShowHook to see that the reordering was actually made. Moving the memoir trimmarks into the shipout/background is a good first step to fix the \pagecolor issue. This test shows we can make page colors without \pagecolor and make the order correctly if need be

@UlrikeFischer We have the issue of course that color needs to work with plain

@JosephWright sure, but one could always split here in a plain \pagecolor and a latex \pagecolor version.

@UlrikeFischer Indeed

@daleif hooks aren't sorted constantly, but normally at begin document or first use (there is some speed optimatization here, @PhelypeOleinik could explain all the fine details).

10:52 AM
Fedora 33 will have BtrFS as default!
And... /sigh nano will be the default editor for the terminal...
@JosephWright ^^

@PauloCereda Sounds good to me :)
@PauloCereda Well, really should be pico, but close enough

@JosephWright Get a XPS and install Fedora, then :)

@PauloCereda This is a good thing, no?

@JosephWright I am going to trout-slap you :)
@JosephWright I think it is. :) I have BtrFS in my laptops for a while, more than an year and half. In this workstation, I still have EXT4. :)

@PauloCereda does it also have an "expand acronyms automatically to something understandable" feature?
2

10:55 AM
@PauloCereda Ext4 has been around ... a while

@UlrikeFischer LOL
^^ :)

@PauloCereda my mobile uses BtrFS for years now (my work-box still EXT4 as well)

@JosephWright it is. I was a fan of ReiserFS. :)
I used OpenSolaris in one machine and they had ZFS, I guess. Pretty nice too. :)

@PauloCereda Ah, so was I: takes me back

@JosephWright (you probably know the story behind the author of ReiserFS, don't you)

10:57 AM
@PauloCereda Yes, I do!

@UlrikeFischer Might be an idea to note that in the manual, was bit confused when it said execution order not set yet

@JosephWright quite the drama, yes
@JosephWright BtrFS seems like a very nice, stable filesystem, so I am happy to learn it will be the default for F33. :)

@daleif yes but at best sent a mail with your comments or open an issue so that it doesn't get lost.

@UlrikeFischer is this just an issue on latex2/latex2e or ...

@UlrikeFischer BtrFS is an acronym for B-tree file system, which is based on B-trees for directories and the copy-on-write principle. It has some nice features. The part I like is how this acronym is pronounced: butter F S. :)
In Italian, it would be burroFS.
@CarLaTeX @egreg ^^ :)
Also, for @Rmano: mantequillaFS. :)

11:07 AM
@daleif for example. This is new stuff and feedback is welcome. You could also mail to the team.

@UlrikeFischer the team mail does not like my mailing system (outlook, no-one can figure out how to confgure it to only send text emails to a specific receiver, so I gave up on that), so for me issues on GitHub is better

@daleif Email is a plain text format ;)

@daleif the solution is easy, only send text-mail.

@JosephWright HTML emails is the default and my colleagues rely on it for colors in emails when editing texts ...
@Skillmonlikestopanswers.xyz I would, but it is not the preferred way by others.

@daleif oh no

11:12 AM
@daleif I suspect my colleagues have got the message that I'm a traditionalist: I have shifted my work email to reply-above, but that's because nobody seems to listen about that

@daleif others are wrong.
@daleif but to err is to be human.

@JosephWright it is SOOO annoying that one cannot get outlook to add my reply at the bottom

@daleif Ah, well there I simply don't use Outlook :)

T H U N D E R B I R D

@JosephWright We are not allowed to use anything else for security

11:14 AM
are go

@daleif Oh goodness, idiocy everywhere

@PauloCereda does not support our Two-factor correctly.

@daleif You on Office365 or some home-baked solution?

@daleif oh :(
@JosephWright ooh baking

@daleif But I'll bet you are not encrypting mail using PGP ...
@daleif You don't have any Linux users? We've had some 'interesting times' at work, but there are a core set of Linix users in maths/computing (the latter get a free pass on all of the AD stuff)

11:16 AM
@JosephWright MS365, I don't think they call it Office365 anymore. As a linux user I use their webmail. It isn't bad.

@UlrikeFischer @JosephWright also, F33 will feature something that @DavidCarlisle will love: they are replacing swap by... zRAM. :)

@daleif Hmm, didn't know it had two-factor: I've not seen that (I do use two-factor for e.g. GMail)
@daleif So you are basically 'not supported'?? (Webmail isn't regarded at my work as suitable for the Linux team)

@JosephWright We can use an app for 2-factor. It works with a lot of our systems including email and VPN. The app even have a key generator for when you don't have net on your phone

@daleif What do you do on your own PC? I'd hate to have to have separate programs running for email from different places out-of-hours (as an academic, that seems perfectly natural)

@JosephWright define not supported? The webmail for example is the goto if Outlook on windows is acting strange. Our webmail is run directly by MS.

11:20 AM
@PauloCereda mutt

@daleif As in webmail is not a Real client, so it doesn't 'count': for example, what happens when people join/leave, and need to move all their mail from their previous employer? I'd move across in TB, as I have in the past.

@JosephWright Webmail. Generally I don't check my work mail when I'm not at wok. I also have my obvious Gmail

@Skillmonlikestopanswers.xyz alpine :)

@daleif Ah, I guess I'm a more 'classical' academic: I just have 'all my email' in one place
@PauloCereda Sounds good to me
@PauloCereda I miss pine

@JosephWright I have something to cheer you up. :)

11:22 AM
@JosephWright Yes, there are features that we currently don't have in the webmail, so one might need real Outlook to get your emails out. But Our helpdesk can help with that. It is also my impression that MS is moving towards having everything online instead. So one can hope...

@PauloCereda never used it, I had the evolutionary path of Outlook (yes, I'm young), Thunderbird, mutt. Up till now I like mutt the most.

@daleif At one point I was using a WebDAV set up, as IMAP was not enabled to start with on our Office365 set up. But they fixed that :)

@Skillmonlikestopanswers.xyz ooh rabbit skills

@Skillmonlikestopanswers.xyz I started with a program called Pegasus Mail (was one of the first Windows-specific clients, I think)

@JosephWright it was a nice email programme. Had a collegue who retired this last new year who sternly refused to leave Pine when we move her to Windows (from Linux). Though she just neeeded to see what sehe could actually do

11:24 AM
@daleif What I miss is that it logged directly into the mail server, so it was reliable in a way nothing else is

@JosephWright We had IMAP available at first. But is now shutdown by the security team because of the missing two-factor support. I did get my Liinux box to actually use MS SMTP to send out via.

@daleif Hmm

@JosephWright It was nice when we did get IMAP. those users who used POP and forgot about it deleting on the server ...

@daleif Unfortunately, a lot of the 'security' people are not so clued-up: for example, we still get the 'you need to change your password' rubbish even though the probably of a rainbow-table attack nowadays is tiny

The main issue I have with using webmail is the mailto: links on websites. I cannot just click on them. At one point I did get Firefox to redirect mailto links to the correct OWA URLs, but that no longer works.

11:27 AM
@daleif Our big worry at work is GDPR/data leakage, but the way it's being tackled is daft: they are obsessed with 'not giving out Admin rights' (which I managed to get on my workstation), but we are still using Excel sheets to track data ...

@JosephWright It is not so bad here. But still a concern. Regarding password, we use rather long ones, but only have to change them every couple of years.

@daleif We get 1 year (on average)

The password isn't the main focus any more. Two-factor gives a much better security if someone phishes your password

@daleif It's not just Admin rights: we are using AppSense which prevents any binaries being run unless authorised
@daleif Sure, but that's if it's workable: can't imagine how you could do this with the students

@JosephWright We might be headed that way. When we got our current Anti-virus we had some issues with LaTeX being marked.

11:32 AM
@daleif I had an issue with biber, as it unpacks stuff, but like I say, I have made the case for being allowed admin rights (I've got loads of specialist stuff that has to be updated all the time)
@daleif The great irony for us is Computing are allowed to do what they like: it's totally unworkable if you write your own software

@JosephWright Had an interesting issue at one point. We got a new Xerox printer. I downloaded the drivers from their website for my linux. When I ran the installer I lost internet access! First of one needs to know that our printers sits on a closed printer network (all prints a auth'ed, students have to pay). The installer wanted to set it self up with whatever accessories the printer has by asking the printer directly. So it did a network scan.
The network detected this immediately and cut my internet access.

@JosephWright the next PR will be a lot more to review, sorry :) Stuff I'm changing: where possible use e type expansion instead of x, remove unnecessary \exp_not:(n/o) around things expanding to strings, remove unnecessary \(str/tl)_set:N(n/x) and use \cs_set_nopar:Np(n/x) on things where we already know it will expand to strings, use \cs_set_eq:NN instead of \tl_set:N(o/x) where possible, and (maybe the most drastic change, while being no drastic change at all) curry everything

In order to get the appropriate Windows driver they have to go several hoops for the installer to see the printer and create the driver.
@JosephWright exactly. at a university the worst thing IT can do is set limits on the researchers on what they can and cannot do

@daleif Oh my!
@daleif We keep being told we will move to shared printers, but at the moment I still one in my office (plus one in the lab)
@daleif 90% of the uni is just 'office stuff', but in science there is loads of specialist stuff

@UlrikeFischer it stands for Butter File System.

11:43 AM
@JosephWright When the other departments started moving to "students have to pay for printing" we began to see students from the Law department (of all places) come by to print teaching materials for free. So we moved to that model as well. Cut the number of printed pages by 60% and no-one leaves prints all over the printers

ooh let's rename the printing network Greta will complain about paper waste.

@daleif Oh, our students have to do that: the only printers we have are in staff offices or locked labs. The 'plan' is to have communal printers for academics, as we already do for students, but there are basically no spare secure spaces in our building
@daleif For our final-year project students, we let them use the lab printer (that's where things like theses get printed)

12:12 PM
@Skillmonlikestopanswers.xyz Oh, thatz bad. I'll try to find out what's happening

@PhelypeOleinik thanks :)

@PhelypeOleinik any place where I should file an official bug report?

@daleif Looks like \ShowHook should initialise the hook before showing it...
@Skillmonlikestopanswers.xyz tex-k list, probably

@PhelypeOleinik "the hook needs bait"

12:13 PM
@PauloCereda :-)

@PhelypeOleinik might be an idea.

12:25 PM
@PhelypeOleinik while looking at @daleif's example I noticed that in the lthooks docu there are some errors, starting with line 344. there is < instead of "before", NewHooks has in the second example wrong arguments, and the argument of the rule looks wrong too.

phelype@phelype ~/d/A/c/lthooks> git blame lthooks.dtx -L344,344
a255cdf6 (PhelypeOleinik 2020-07-07 22:33:53 -0300 344) % \begin{verbatim}
@UlrikeFischer Not my fault ^^
@UlrikeFischer < was kept with the same meaning of before (same for >---Frank prefers the shorter symbols). I'll fix the \NewHook syntax, thanks!

Stupid question: using pdflatex how does the PDF get the correct height and width? We had issues long ago where the sizes was not written correctly and the size was wrong. Not sure if it is setting \pdfpagewidth etc?

@PhelypeOleinik I got an error when I tried \DeclareHookRule{shipout/background}{bg}{<}{trimmarks} (but I'm not sure in which state my system is).
! Undefined control sequence.
<argument> \ERRORunknownrule

@UlrikeFischer Weird... Let me check

@daleif what do you mean by height here? the "pdf background"? That is \pdfpagewidth, it sets the media box, but you can set more.

12:35 PM
@UlrikeFischer something like that. was playing with \documentclass{memoir} \usepackage[a4paper]{geometry} and it kept coming out as letter, turned out to be because I also loaded xcolor later on.

@UlrikeFischer Ah, < is not implemented in \@@_debug_gset_rule:nnnn (line 2180; used for debugging only). I'll fix that as well. (I want to get rid of this debug-only set of rules)

\documentclass{memoir}
\usepackage[a4paper]{geometry}
\usepackage{xcolor}
\begin{document}
test
\end{document}

@PhelypeOleinik not sure about that, afterall it is a bug directly in pdfTeX and XeTeX, not in Kpathsea or other support infrastructure (or am I misinterpreting the scope of that list?)

With book it does the right thing
With memoir the size is correct if I load xcolor first.
Not that it is really my problem as memoir uses its own system for the layout, but sometimes users do use geometry

@Skillmonlikestopanswers.xyz Well, you can post to pdftex (and crosspost to xetex), but tex-k is general for web2c, so I think it's fine (that's what I would do), especially because \expanded is not precisely pdfTeX, but it's the web2c implementation.

12:40 PM
@daleif hm yes, the latex drivers detects memoir and so use the \stockwidth to set the page size, but geometry ignores this length. Probable it should set it too.

@Skillmonlikestopanswers.xyz I'd say go for it

@UlrikeFischer how is that related to xcolor? Remove xcolor from the example and the size is A4

@PhelypeOleinik when I got a short "break" from my day-job.

@daleif or the latex driver should detect that geometry is there and prefer \paperwidth

pdftex.def?

12:43 PM
@daleif color loads the graphics driver, pdftex.def and this sets the page size.

@UlrikeFischer exactly. Hmm, I'll just not care about it for now. Haven't had any bug reports about it ;-)
@UlrikeFischer yes, since geometry does not support the stock sizes in memoir it ought to just set \stockheight/width to paperheight/width if used with memoir

@daleif if I got it right memoir uses paperwidth for the layout, and stockwidth for the paper? So for A5 on A4 paperwidth would be the A5 value?

@UlrikeFischer that is the idea
I'll email the geoemtry maintainer at some point. I don't see it as critical at the momemt
Interestingly, a lot of out PhD students have their thesis printed in B5, but we actually just make A4 docs and then leave it to the printing office to scale and cut it nicely (they generally do a really good job)

@daleif well with geometry paperwidth would be A4 and \Gm@layoutwidth would be the A5 value.

2 hours later…
3:15 PM
@PauloCereda I prefer the false friend version (burroFS -> donkeyFS). MantequillaFs does not exactly transmit the idea of solidity... ;-)

@Rmano :)

@JosephWright you must be working in the same place I am... :-(

3:38 PM
....
.....
........

We now \IfFileExists is there sth. \IfFileHasChanged ?

@cis not too trivial but not that hard, either. You can use \file_get_mdfive_hash:nN and compare to something you wrote into a file (e.g., .aux) on the last run.
@cis to get something a bit faster, but less reliable you can use \file_get_timestamp:nN.

3:57 PM
@PauloCereda Since FS are the Italian railways, it sounds a bit strange

@CarLaTeX ooh

I have hacked genealogytree.sty
( https://tex.stackexchange.com/a/557508/46023 )
...
But I do not understand anyway why an input method from simple raw data (see table) is not provided by default. That would be my first thought as a package author.

4:25 PM
@Skillmonlikestopanswers.xyz I created a question but I am not allowed to point it out here.
( https://tex.stackexchange.com/questions/557511/how-to-query-iffilehaschanged )

@cis I don't have much time today, might have to wait until tomorrow evening.

@DavidCarlisle Now you can be depressed tex.stackexchange.com/a/556803/4427

@Skillmonlikestopanswers.xyz It's not your sole responsibility either. Perhaps I can solve it myself with research, which I currently doubt.

1 hour later…
5:43 PM
3
oh nein

@PauloCereda Something that can be only understood by German native speakers

@ChiefO'Brien Google Translate didn't help me.

@PauloCereda Have never used it yet

@ChiefO'Brien and naturally by our linguists.

@UlrikeFischer Most likely but does being a linguist mean that those people can understand any other language?

5:52 PM
@UlrikeFischer ooh
poor whale

@ChiefO'Brien depends on the linguist, our is fluent in every language.

@UlrikeFischer 德语很容易

@UlrikeFischer I see -- even Klingon language?

@DavidCarlisle ooh

@DavidCarlisle curiously if you tell google that it is japanese it translates it as "easy to use" ;-)

5:56 PM
I have the impression that most people in here come from Europe

@ChiefO'Brien quack

@ChiefO'Brien ^^^ that was a no.

@UlrikeFischer quack quack means yes then?

@ChiefO'Brien ask the duck or the linguist.

@ChiefO'Brien oh no, that would be double negative

5:59 PM
@UlrikeFischer I thought that would be the same person/animal ;-)

although quack might also be yes. :)

@PauloCereda which is positive, isn't it?

@ChiefO'Brien not in as in the likes of formal logic :)

@PauloCereda so this is duck logic?

@ChiefO'Brien quite. :) In some languages, double negative just emphasizes the negation
But I am just a duck :)
@DavidCarlisle is the official linguistic :)
Allow me to demonstrate
@DavidCarlisle quack

6:01 PM
@PauloCereda A duck with a baseball bat in its hand?

@ChiefO'Brien a cricket bat, actually :)

@PauloCereda Ah cricket
@PauloCereda: Cricket isn't very well-known in Germany I believe. I don't know any fan of it in person oder a club where people playing cricket.

@moewe are you there?

@ChiefO'Brien nor in Brazil. :)

@PauloCereda You're from Brazil?

6:09 PM
@ChiefO'Brien yes!

@PauloCereda I hope you are not in trouble with this Corona situation?

@ChiefO'Brien Well, the situation in the contryside areas is quite complicated at the moment. The wave is hitting us pretty hard and the current health infrastructure is insufficient to attend all people...

@PauloCereda Oh my. Keep safe and healthy. We wear masks at the office or work from home. But the number of infected people is rising again in Germany.

@ChiefO'Brien Thanks, you too. :) People are probably relaxing far too much (not sure if that is the correct verb).

@PauloCereda Thanks. Yes there was a demonstration of Corona protesters in Berlin last weekend, protesting against the shutdown policy, the government, the distance rules, obligation to wear masks and a lot of other issues which they say is a conjuration. Most likely they protested also against the green color of grass or the blue sky.

6:19 PM
@ChiefO'Brien So I heard! I think there has some protests in Brazil as well...
My parents and I are locked inside home.

@PauloCereda Oh my... looks rather like a riot than protests

@ChiefO'Brien indeed

@PauloCereda Well I have no sympathies for those protesters but as far as I have heard or read there was no riot situation in Berlin.
In Germany they are called Covidiots ;-)

@ChiefO'Brien ooh :)

In the example shown in this answer: tex.stackexchange.com/a/88939/6302 Is there a way to make "Parameters" (the first column's heading) vertically centred in its table cell?

6:40 PM
@PauloCereda dinner

@DavidCarlisle Mean
@DavidCarlisle Fewer users yesterday :(

@JosephWright not me
@JosephWright probably because I'm on holiday:-)

@DavidCarlisle oh no

@DavidCarlisle ooh Windows XP

6:50 PM
@egreg we don't need the Romans to build forts:-)

7:18 PM
@PauloCereda Do you have a lot of UFO sightings in South America?

@DavidCarlisle where is this?

8:10 PM
@PhelypeOleinik aaaand, I didn't :) should I even, if you and @JosephWright will take a look anyways?

8:30 PM
@Skillmonlikestopanswers.xyz No problem, it was just for the record (and for people who know more WEB than me---read: everyone in that list---to have a go first). I can do so, if you don't mind

@PhelypeOleinik you could as well report the bug, if you want to (I'm not subscribed to it). And I don't know a thing about WEB.

@UlrikeFischer Now I am.

@moewe I was wondering if you know a style that implements this tex.stackexchange.com/questions/83845/…

@UlrikeFischer I think I saw something like this before. The hard part is finding it in the thousands of biblatex questions. Let me have a look.

@moewe the main question is not some code, but if I can show a short example, which means a ready style.

8:39 PM
4

I would like to get the bibliography style of the Kelley's book General Topology: I've been reading the options that LaTeX has here and there is no one like Kelley's book apparently. Is it possible to get what I want? Thanks PD: I haven't said it but it is supposed to use the standard bibte...

@UlrikeFischer Oh, sorry I misunderstood. I think then the answer is no.

@moewe but the code is nice too. Perhaps I can make use of it.

@UlrikeFischer I guess you could ask Ivan if he'd like to include that in his philosophy-modern as an 'alternate' version.

@Skillmonlikestopanswers.xyz @cis I tried an answer --- this is my first expl3 code, so it's probably horrible. But I'll appreciate comments and rants...

9:29 PM

@DavidCarlisle looks nice.

10:04 PM
@DavidCarlisle I can build oberdiek, so I could upload together with the other two. But I should probably create a proper readme.md with license. But what to take as version number? Simply the date?

10:27 PM
@chsk -- I'm one of the few people who has had a paying job that depended almost entirely on TeX; in fact, before TeX, I was the technical contact with the provider of typesetting software to the American Math Society, so I've been in this racket for a long time. Re hyphenations, the first exception list appeared in TUGboat 5:1 (1984), p.15; it was very short. It has been updated semi-regularly ever since, and addenda published every few issues. The cumulative list is on CTAN. (cont'd)
(cont'd) Where would you normally look for such things? The full TUGboat contents are listed on tug.org on the TUGboat pages, and most of the exception list entries are listed separately. (And for the few early ones that aren't, I'll try to get that fixed.)

10:43 PM
@JosephWright -- Some people have no choice. That was my situation at AMS before I retired. It was terrible!

10:59 PM
@PauloCereda -- Good find. I should probably read it sometime to learn how to improve my writing. Thanks.

11:33 PM
@DavidCarlisle -- There's a very funny typo near the beginning of the second paragraph. One might even think that you might have written it. (Interesting place, though.)
@AlanMunn -- A grammar question. Is "pair" a count or a mass noun. I claim count -- I have four pairs of socks. My husband claims mass -- I have two pair of trousers. Maybe it's regional variation?