« first day (5200 days earlier)      last day (18 days later) » 

01:26
After I finally updated tlmgr for TL 2024, my big German book that heavily uses cleveref failed to compile:
! Use of \cref@override@label@type doesn't match its definition.
\label@optarg ...el@type \cref@currentlabel \@nil
{#1}}\protected@write \@au...
Of course I could take the burden of producing yet another bug report and asking the community for yet another bugfix and then trying to extend https://tex.stackexchange.com/a/309602 . However, I feel that it's time to stop flogging the dead horse. Where to move to: zref-clever or abandon automation altogether and go back to manually spelt en
 
6 hours later…
07:20
@cfr That is lang-txt.lua.
 
2 hours later…
cfr
cfr
08:52
@mickep well, you said that wasn't typeset ...
@AlMa1r did you update all the packages and the format?
cfr
cfr
09:12
@AlMa1r it works ok here, but there is also fancyref which is somewhere in between in terms of automation. it is much less subject to breaking ...
@DavidCarlisle oh no
09:42
67078284 cleveref works for me with an up-to-date TL2024... are you sure you don't have some dangling .cls/.sty around?
@PauloCereda breakfast
@cfr Then there is a misunderstanding. The labels in lang-txt are typeset. I thought you were looking at the interface file (the one where one can change the names of the macros used).
10:43
@Skillmon trouble maker
@samcarter Oh, using the cow image <3
@DavidCarlisle sorry for spotting hidden features in your code
11:11
@mickep moo!
 
2 hours later…
12:59
Hi guys, bit of a weird question here: I don't understand the layout decision made when processing the following mwe
using latexmk -xelatex
I have done my best to remove all useless stuff
without the geometry requirements, the table is inserted middle of the paragraph, right after the sentence that introduces it
but for some reason with the geometry directive (and maybe some unfortunate wording? But I have tried rephrasing without success) one additional line is inserted before the table
Anyone with an idea what's going on?
h! doesn't help either for float placement
and without the 11pt font the issue is not there either
@nathdwek you are inserting the float in the middle of a paragraph. LaTeX will then try to insert it after the next line. Because of the space after Table~\ref{tb:accels}. the next line has already started and is filled up first. If you use Table~\ref{tb:accels}.% the output will be different.
oh
Wonderful
thanks
I thought it had something to do with the sentence ending exactly at the page width
Thank you very much
 
1 hour later…
14:29
@UlrikeFischer I ran tlmgr update --self --all (and it terminated successfully or at least without visible errors) before compiling my book.
@AlMa1r but are you sure you have no local copies of old versions in your path (that's the most common cause of failure)
@DavidCarlisle The used tlmgr and TeX Live 20224 is installed locally in $HOME. There is an older TeX Live version installed globally for everyone. $PATH, $MANPATH, and $INFOPATH have been set appropriately so that the local version kicks in first; I'll double-check. Have there been any changes in this regard (where $PATH, $MANPATH, and $INFOPATH should point to) since autumn?
@AlMa1r no but still it's worth checking your log, in particular $HOME/texmf is ahead of the standard path by default so if you for example have an old copy of expl3.sty in ~/texmf/tex/latex it doesn't matter how many times tlmgr updates expl3 in /usr/local/texlive/ it won't get used.
@cfr Thx. Fancyref has not been updated since 1999 according to CTAN, right?
:67081922 \ProvidesPackage{fancyref}[%
  1999/02/03 v0.9c Fancy cross-referencing%
]%
14:39
@DavidCarlisle Thx, I'll double-check.
@DavidCarlisle So if I were to guess, my hunch would be that fancyref is useless/obsolete by now.
@AlMa1r basically the log of your book should only have (/usr/local/texlive or (./ if it shows any input paths starting (/path/to/somewhere/else... then things are suspicious
@AlMa1r well I have some packages that haven't changed since then and they are all brilliantly relevant and needed of course. But I don't really know that package, it's only not needed if you have a replacement
3
 
3 hours later…
cfr
cfr
17:44
@AlMa1r it is neither useless nor obsolete. unlike cleveref, it works.
@DavidCarlisle but it still works and either doesn't require continuous first aid or demands it much more quietly than cleveref. I assume this is because it is less clever.
@mickep briefly, but not after you told me that was for macro names. I only tried to add translations to lang-txt.lua. but I do not understand the English in some cases such as at page and I do not know how to deal with In, in or and, for instance. what does context do for languages with more than one word for and, for example?
18:47
@cfr The misunderstanding was on my side. We need to remember that many of those words in lang-txt.lua are old, and maybe even survived themselves. I know of no handling in case there would be more alternatives. But we shall remember that the user can change these into whatever fits their needs, \setuplabeltext[sv][chapter=Lecture] so it is not the end of the world if there are several.
cfr
cfr
19:15
@mickep sure. but there are two words for 'and' and you need both. it isn't a question of choosing but grammar. when one of them is correct, the other is wrong. so you can't just use one and figure users can switch the setting.
@mickep 'in' is more complicated.
so for something like References, fine. somebody might want References or they might want Bibliography. not a problem. but words like and and in are not like that ....
A and B   -> A a B
B and A   -> B ac A
Wales and Patagonia   -> Cymru a'r Wladfa
Patagonia and Wales   -> Yr Wladfa a Chymru
in London     -> yn Llundain
in Wales      -> yng Nghymru
in Pentwyn    -> ym Mhentwyn
in a tree     -> mewn coeden
in the tree   -> yn y goeden
in it         -> ynddo | ynddi
19:32
Then better bot use this mechanism. Just write the text.
cfr
cfr
@mickep so omit the translations?
@cfr Or add one. And then Welsh people will need to set things up themselves if they want other choices.
Is it just me, or is \tl_case:... no longer documented in the interface3 Expl3 manual?
20:02
And \tl_case:nnTF and similar nn commands apparently don't exist at all, but maybe that's intentional
20:53
@Gaussler yep, because it's just \token_case_meaning:Nn.
cfr
cfr
21:17
@Skillmon I thought the functions in interface3.pdf were supposed to be stable? or did that happen more recently?
[never mind]
@cfr The “suppressed” ones still continue to work.
cfr
cfr
21:40
@egreg thanks. I realised that too late to delete the question. hence 'never mind' :(.
22:27
@DavidCarlisle belated oh no

« first day (5200 days earlier)      last day (18 days later) »