« first day (4785 days earlier)      last day (85 days later) » 

9:28 AM
user image
5
I would like to take this opportunity to thank the puppet master. Without their constant stream of rude replies, this palindrome would not have been possible! (sorry for the comma placement :)
 
9:45 AM
Uh, Master of Puppets
 
@DavidCarlisle, @UlrikeFischer I bet you can guess my view on the ltproperties business ..
 
@JosephWright use \@protectedwrite so things work properly?
 
@DavidCarlisle Not exactly ...
 
@JosephWright I suspect that use \iow_shipout_protected:Nx which is same thing is the answer
 
@JosephWright even if we make all our commands protected, there will be still robust commands in the wild, so we need to protect them, either by generally using \@protectedwrite, or at least for the properties where one can expect user input (title).
 
9:54 AM
@UlrikeFischer You guessed my view :)
@UlrikeFischer As I always sat, e-TeX was finalised in 1999
 
10:08 AM
@DavidCarlisle do we have that??
 
@UlrikeFischer No
 
it's only one \let away:-)
@JosephWright but we decided to do l3in2e not latex3 and that has some consequences
 
@DavidCarlisle True - what I mean is really people should have been writing new code using \protected for at least a decade, likely 15 years, which would only have left the kernel ...
@DavidCarlisle Here, as you know, I'm with Hans
 
@JosephWright no they should have been using \newcommand\foo[2][abc]{...} and that's \protect robust...
 
@DavidCarlisle \newcommand should've been \protected robust since 99 :P
 
10:12 AM
@Skillmon Y&Ytex:-)
 
@DavidCarlisle and how should people then write commands that should expand? E.g. change \thepage?
 
@UlrikeFischer protectedwrite expands thepage
 
@DavidCarlisle I meant, if \newcommand produces \protected commands, then \renewcommand\thepage{\alph{page}} would no longer work in labels wouldn't it?
 
@UlrikeFischer that's why \protect is such a good system and more flexible than \protected as @JosephWright will confirm:-)
 
@UlrikeFischer WEll just \newcommand doesn't use 'protect anyway ...
@UlrikeFischer, @DavidCarlisle You know I'm making some fun :) I know I'm not going to ever get this ...
 
10:28 AM
@JosephWright ;-). But as David wrote: one need some flexibility. Depending on the context you want a different expansion. It is not so much different to the case-switching ...
 
@UlrikeFischer ConTeXt has managed for a long time using \protected, so I really do think it's workable, but I know I'm never going to convince people
 
10:41 AM
@UlrikeFischer \newcommand*\expandprotected{\expandafter\NOP}\newcommand*\NOP{}
 
@JosephWright I'm not against making more things protected (and don't see a reason why \textit should explode) but we should also be realistic: we do not control 99% of the code like context so we need more safety measures.
 
@UlrikeFischer Oh yes, I know: it's partly a philosophical thing (ConTeXt after all takes a much harder line on updating sources)
 
11:20 AM
@JosephWright and has a few orders of magnitute fewer users
 
@DavidCarlisle Sure, yes ...
 
11:38 AM
@JosephWright I'd tend to close github.com/josephwright/beamer/issues/863 as "not planned" or do you want to keep the issue around?
 
@samcarter Yes, sounds good
 
@JosephWright :thumbsup:
 
11:51 AM
@samcarter oh, Org-mode loads biblatex after hyperref.
 
@Skillmon Ooops
 
@JosephWright That's what they get for using a strange OS for an editor.
 
@Skillmon does it matter?
 
12:08 PM
@UlrikeFischer well, at least it doesn't follow best practice regarding hyperref-loadorder.
 
1:07 PM
@Skillmon I'm trying hard to change this rule to "load order matters only for some old, special packages that don't update their code". biblatex should simply work, and if not either hyperref or biblatex should correct that.
 
@samcarter Your answer was basically correct, you were missing only the second mandatory argument.
 
@campa Yes, I noticed. I undeleted in the mean time :)
@campa Thanks for your chat message :)
 
@UlrikeFischer I know there is some (much) effort, but I'm not sure how far you've already got along, so I currently stick along the oldish rule "load it last".
 
1:26 PM
@DavidCarlisle, @UlrikeFischer, @JosephWright: You might also be interested in the "semiprotected" (see the lowlevel-expansion manual for details).
 
1:46 PM
2
A: is there a way to 'breakeline' a newcommand? (to modify only \textbf color)

Ulrike FischerThe font commands have hooks. E.g. to add something to bold you can use the bfseries hook: \documentclass{article} \usepackage{xcolor} \AddToHook{bfseries}{\color{gray}} \begin{document} This is some very long text that should hit the end of the line \textbf{but in gray and bold and very elegan...

^^^ I think @UlrikeFischer should change her user name to "Captain Hook" :P
(as Captain Hook, she could also carry her sword around all the time)
 
2:06 PM
@UlrikeFischer "you cannot delete this accepted answer"
 
Has there been some recent changes to multicol? Text inside multicols environments is overrunning the column in older document where it used to work.
 
@AlanMunn define "recent"
 
@UlrikeFischer Well a document that compiled fine in May doesn't now.
 
2:23 PM
@AlanMunn the last multicol is from 2023/03/30 v1.9f. But there have been two latex releases since may, and they could be something they affected multicol, but without an example it is difficult to say.
 
@UlrikeFischer Yes, I realize that. No time for that right now, as I'm off to teach. It shouldn't be hard to make an example.
 
@UlrikeFischer I got tagging working for postnotes (github.com/gusbrs/postnotes/tree/tagging). Just out of the oven, and pending thorough testing, so expect corners. Also, I could not find a way to do things without tampering with latex-lab-testphase-block internals, which makes me actually weary to release it in these terms.
 
2:43 PM
@gusbrs no, tampering with the internal is not so good. I will try to look at your code, but not directly, I'm a bit busy currently.
 
@UlrikeFischer I know, and agree, of course. Hence the weariness. But, as far as a can see, there is no (public) interface to tweak the list tagging code, so...
 
@mickep Link?
 
@UlrikeFischer This question might benefit from another of your hook-wizardry answers tex.stackexchange.com/questions/222201/…
 
@JosephWright In the distribution: tex/texmf-context/doc/context/sources/general/manuals/lowlevel/lowlevel-expansion.tex
 
@mickep Hmm, don't actually see a description, only use, and > \semiprotected=undefined :(
 
2:59 PM
@JosephWright In an updated installation? (It is kind of new)
 
@mickep Hans isn't good at releases in a way that fits with everything else ... I have the latest TL (updated this morning), but he doesn't tend to see that as the official release approach
@mickep It's in the docs, so you'd hope it was in the code too :)
@mickep Oh, it's only in LuaMetaTeX :(
@mickep I rather assumed it would be a macro construct like \protect
 
@JosephWright Oh, TL...
@JosephWright Indeed, that is where new stuff end up.
 
@mickep Yes, as it's updated every day from CTAN ...
@mickep Well yes but the way you said it, I thought it was a long-standing feature so was looking at MkII/MkIV
 
@JosephWright I don't know who decides when to update context in TL. Usually there is also a new binary.
@JosephWright Oh, no, it is not as protected was before eTeX, sorry for confusion.
 
@mickep It's OK - fits with what i thought that ConTeXt historically has gone with sheep-and-goats (either \protected or entirely expandable)
 
3:09 PM
@JosephWright I will paste an example in case you cannot get the manual.
\starttext

\def\Knuth{knuth}

              \def\TestA{test \Knuth}
    \protected\def\TestB{test \Knuth}
\semiprotected\def\TestC{test \Knuth}

\edef\TestAA{\TestA}
\edef\TestBB{\TestB}
\edef\TestCC{\TestC}

\edef\TestAAA{\semiexpand\TestA}
\edef\TestBBB{\semiexpand\TestB}
\edef\TestCCC{\semiexpand\TestC}

\meaningfull\TestA \par
\meaningfull\TestB \par
\meaningfull\TestC \blank

\meaningfull\TestAA \par
\meaningfull\TestBB \par
\meaningfull\TestCC \blank

\meaningfull\TestAAA \par
\meaningfull\TestBBB \par
 
@mickep @mickep OK, I think I get the idea
 
@mickep I was expecting some formal docs, but Hans hasn't done that for the primitive
 
@JosephWright Well, formal/informal... (It is also mentioned in tex/texmf-context/doc/context/sources/general/manuals/primitives/primitives.tex)
 
@mickep Relies on what you put inside \edef, so not the same as LaTeX's \protect, which is a question of what you do outside the \edef ...
@mickep I meant I was searching for a (sub)section or margin note that said the equivalent of \cs{semiprotected}
 
3:15 PM
@JosephWright I see. (I think I prefer docs with examples.) The lowlevel manuals contain a lot of nice stuff, probably enough to save the weekend. :)
 
@mickep Oh, I like examples, but I was expecting to find a section called 'Protection and semiprotection' or similar, with marginal notes of primitive names (hence {semiprotected} as my search string)
 
@JosephWright I guess you expect too much. :D
 
@mickep Just a different way of thinking about the info, I guess
 
@JosephWright I think Hans is prioritizing new projects, but in this case I am not sure I'd complain about the documentation.
 
3:38 PM
I wrote a pdflatex document. The site https://www.pdfforge.org/online/en/validate-pdfa said:

Specification

Clause

Description

ISO 19005-1:2005

6.2.3

If an uncalibrated colour space is used in a file then that file shall contain a PDF/A-1 OutputIntent, as defined in 6.2.2

ISO 19005-1:2005

6.7.3

If a document information dictionary does appear at a document, then all of its entries that have analogous properties in predefined XMP schemas, shall also be embedded in the file in XMP form with equivalent values.
 
@JosephWright formal docs in context? ;-)
 
@UlrikeFischer :)
 
@JaakkoSeppälä PDF standards are mostly a job of your document code. You need the right settings/packages. The e.g. tex.stackexchange.com/a/655531/2388 or the pdfx package.
 
 
2 hours later…
6:15 PM
@UlrikeFischer As you were probably already thinking, but were too polite to say :) it was my error. I had a negative \columnsep set for one environment, and it persisted in subsequent environments which were copied from other documents. So within each of the individual original documents the code worked, but combined they didn't.
 
6:30 PM
@AlanMunn don't feel too bad, at least you didn't break longtable, unlike others who remain nameless
4
 
@UlrikeFischer Well, you can be ironic, but there are a lot of nice documents to read. Maybe not with keywords in the margin, though. It is a bit as if I would joke about that since there is nothing similar to setup-en.pdf for latex, it has no useful documentation. But I will not.
 
6:56 PM
@mickep I don't doubt that there are nice documents, I even looked at the one or the other. But there is no real formal code or user documentation. setup-en is actually a good example: It is a large impressing list, but there is no explanation what the commands do or the meaning or exact syntax of the arguments. E.g. there is a \dayoftheweek command with three arguments "NUMBER". So what should I do with that?
 
7:26 PM
But setup-en is not really meant for explanations, and still very usable. Without looking in another place, I would just try `\dayoftheweek{07}{12}{2023}` or `\dayoftheweek{2023}{12}{07}`. Might even be quicker than searching through further docs.
`
(The first one is the correct one in this case.)
 
7:54 PM
@mickep Does \semiexpand itself expand in a single step to the result of the next token expanded? (so is this like \the for registers?)
 
8:46 PM
@Skillmon Oh, sorry, I don't know those details.
 
8:58 PM
@DavidCarlisle oh no
 
@PauloCereda pato de natal assado
 
oh no
 
10:03 PM
user image
5
 
Nov 22 at 19:06, by David Carlisle
Nov 5 at 7:40, by CarLaTeX
Sorry for the comma
 
Nov 22 at 21:11, by CarLaTeX
@DavidCarlisle still valid
 

« first day (4785 days earlier)      last day (85 days later) »