« first day (3824 days earlier)      last day (1100 days later) » 

12:03 AM
@DavidCarlisle -- Well, it was worth asking. That's what I would have tried, using the same counter. (I've been on a plane all day, so am a bit fuzzy brained.)
 
 
8 hours later…
7:39 AM
@JairoA.delRio yes, but not in older XeTeX versions, they didn't support random numbers until 2020.
@JairoA.delRio if you need support for older XeTeXs, you can use random.tex for that.
 
@Skillmon or 9
 
@DavidCarlisle I thought it was 4, decided by a fair die roll.
 
@Skillmon roll the dice you get 4 I get 9 so I win?
 
@DavidCarlisle wasn't the 9 from the Dilbert comics?
 
@Skillmon from somewhere
 
7:47 AM
 
@JairoA.delRio Put in a feature request issue: this would be trivial to add
 
@DavidCarlisle found it^^^
 
@Skillmon I was doing the same:-)
 
Apr 14 at 17:18, by samcarter_looks_forward_TUG'21
@Rmano It would be a great service to humanity if someone could tell this user that the OP is automatically notified and it is not necessary to announce every single answer with a comment...
@samcarter_looks_forward_TUG'21 done... ;-)
 
8:11 AM
Anyone have any suggestions on this one: tex.stackexchange.com/q/593690/3929, any solution seems highly dependent on the GS version and I don't use MikTeX.
 
@Rmano :) Thank you!
 
8:24 AM
@daleif well it could have to do with the fact, that preview has not been adapted to the new shipout hooks. But I can't even compile the pstricks on its own so it is difficult to debug.
 
8:46 AM
Is it? I keep getting gs errors
 
9:08 AM
@DavidCarlisle, @UlrikeFischer, @PhelypeOleinik OK if I reply to the LaTeX-L post?
 
@JosephWright sure. But which end format does his system produce that he thinks for example that a command to set the media size (\pdfpageheight) is unneeded?
 
@UlrikeFischer I'd spotted that one :)
@UlrikeFischer I'm going to try to cover what is actually used where
 
@JosephWright good, because I already started to ponder how we could move all this "set the media size" stuff into l3backend ;-). Do you know his end format?
 
@UlrikeFischer We should do that, but we will need to work out how that impacts on the long standing issue that a 'classical' file with no graphics/colour loaded then comes out a bit differently
 
9:27 AM
@JosephWright we could bind it to the pdfmanagement stuff, then it is new stuff. (I removed there all page size code from the hyperref driver, so actually there is already a difference).
 
@UlrikeFischer That would work: one for the team list?
 
@JosephWright yes, but perhaps better not this week, too much to do now.
 
@UlrikeFischer :)
@PhelypeOleinik Is the csname file stuff done?
 
@JosephWright haven''t seen it yet, but yes
 
@DavidCarlisle :)
 
9:40 AM
@JosephWright ah: list of primitives
 
@DavidCarlisle Yes: I think I've got a reasonable overview
 
how can I make vrule go all the way to the top of the text in the righthand column? bpa.st/KV5Q
 
@Anush avoid the negative \vskip?
 
@Anush you don't need external links (which many of us would not follow):
\documentclass{beamer}


\begin{document}
\begin{frame}{}

	\begin{columns}
		\begin{column}{0.4\textwidth}
some text
		\end{column}
		\vrule{}
		\hspace{5pt}
		\begin{column}{0.6\textwidth}  %%<--- here
			\vskip -0.5cm
			text
			\vskip 0.3cm

			text
			\vskip 0.2cm text

			\vskip 0.2cm text
			\vskip 0.2cm text
			\vskip 0.2cm text


		\end{column}
	\end{columns}
\end{frame}

\end{document}
 
@DavidCarlisle thanks.
@Rmano hmm.. ok
on a related note, is there a more beautiful vertical column separator I could use?
 
9:54 AM
@Anush as Rmano says the -ve vskip is forcing the rule to be short, also horizontally you appear to have the width of spme word spaces, plus 5pt plus the width of the rule more that \textwidth (doesn't beamer have a built in column rule?)
 
@DavidCarlisle thank you.
horizontally I just wanted to move the vrule and wasn't sure how else to do it
 
@Anush you don't have to use a rule you could use \includegraphics{mysep} where mysep.png is whatever tall thin fancy scrolling decoration you care to draw
 
@Anush use \tikzmark to mark the spots and go wild ;-)
 
@Rmano :)
@DavidCarlisle I was hoping for a recommendation!
and I need it to be reuseable in a commercial environment so images can be tricky
 
@Anush some of these are probably free for use google.com/…
 
9:57 AM
thanks
 
@Anush texample.net/tikz/examples/pgf-version-2 several here (and compilable online! ;-P @DavidCarlisle)
...well, no, they are not MWE...
 
@Rmano thanks!
 
10:21 AM
@Rmano compile online: excellent feature that (shame to waste it on tikz:-)
 
@DavidCarlisle you could create an example page for picture mode art.
 
@DavidCarlisle Another MathML question(Just let me know if these annoy you): MathML core allows CSS length units, including rem which is specified to be relative to the "font size of the root element". If MathML is embedded e.g. into HTML, is the "root element" the "<math>" or the "<html>" element?
 
10:37 AM
@MarcelKrüger html (but currently all global properties are a bit vague, hope to finally tie teh css working group down this time to be able to really spec how properties applying to all math should be specified at the top level docuemnt head)
 
@DavidCarlisle, @UlrikeFischer Took a while to write something - I hope I've been clear. Would welcome further thoughts on \pdfpage(height|width) in particular
 
@DavidCarlisle Thanks.
 
@MarcelKrüger you're not affiliated to a w3c member organisation are you? (if so, you could join the WG:-)
@JosephWright not arrived here yet
@JosephWright odd I can see it on the archive
 
@DavidCarlisle Sadly I'm not, I'm only associated with my University and they aren't w3c members :'(
 
10:55 AM
@MarcelKrüger nor am I but I hope they invite me back in anyway (not had that confirmed yet...)
 
@DavidCarlisle They should add a rule that the editors of previous versions get automatically re-invited into the WG.
 
@JosephWright I thought it was OK, thunderbird thought it was junk:-)
@MarcelKrüger I did suggest that might be a good idea in my invite request:-)
 
@DavidCarlisle :)
 
yo'
\def\PauloCereda@occupation{Overleaf}
11
@PauloCereda ^^
 
@yo' you are doomed
 
yo'
11:06 AM
@DavidCarlisle we can always turn him into a dinner if needed :D
 
@yo' ooh
@yo' oh no
 
11:30 AM
@JosephWright Cool, thanks.
@Skillmon Is that in TL? When I type texdoc random I get a mysterious PDF generated by, surprise, ConTeXt MkIV
 
@JairoA.delRio it is in texmf-dist/tex/generic/random/random.tex, that same file also contains the "documentation" in form of comments.
 
Is it just me, or have a significant number of user icons been disappeared? I mean, @DavidCarlisle's tabularx performs one sub-optimal estimation and the next thing, his user icon is gone! How Stalinesque!!
 
11:46 AM
@StevenB.Segletes ? which question do you mean?
 
@JosephWright It is, but the tests are failing because of TL'21, so I was fixing those in main before merging back to the \csname branch
 
@PhelypeOleinik Ah, cool
 
@JosephWright Did you see the opacity test file? It seems to be missing pdfmanagement...
@JosephWright Do we install that to the texmf dir in the repo?
 
@PhelypeOleinik Yes, that's exactly as expected
 
@JosephWright Oh, so I can just l3build save it?
 
11:54 AM
@PhelypeOleinik Not really - the whole reason that code is in l3trial is that until we have release on the PDF management, it's a stub that can't really work
@PhelypeOleinik Yes
 
@JosephWright Cool, will do that then
 
3
A: Overfull \hbox (***pt too wide) of a tabularx in a large page

David CarlisleAs noted in comments under the question tabularx gives a hint if you use debugshow In this case it is doing a specific test to prevent some internal calculations exceeeding \maxdimen and giving errors and spurious results. But in this case it's being over-cautious and backing up when it doesn't ...

@UlrikeFischer But my main point was that many user icons have disappeared.
 
@StevenB.Segletes I see no missing icon on this page. Which one do you mean?
 
12:10 PM
@UlrikeFischer Joseph Wright and Alex G on this page. If I go to users, Zarko, Bernard, moewe,... More than half are gone.
 
@StevenB.Segletes I think we are no seeing the same things, at least I can't see anything odd ;-)
 
@UlrikeFischer For me, the absent icons have persisted for nearly a day. Fortunately, your and my own icons are intact, LOL.
 
@StevenB.Segletes is Joseph's broken?
 
@UlrikeFischer Yes.
 
@StevenB.Segletes so it is not only "site icons", but also personal pictures. Perhaps your browser cache has a problem.
 
12:19 PM
@UlrikeFischer I will close browser, reopen and check
 
@StevenB.Segletes someone has eaten the duck, blame David?
 
@Skillmon My bad. I didn't have the package installed. Thank you very much!
 
@JairoA.delRio the other alternative (and something I've done in the past) is to be adventurous and implement your own pRNG (I once implemented an RC4 in TeX)
@PauloCereda counter-ooh!
 
@Skillmon Was it expandable? ;) (Bruno and I had discussions about that)
 
@UlrikeFischer Cleared cache, rebooted browser. Still bad
 
12:24 PM
@JosephWright no, it wasn't :) And it was awfully slow, today I would've coded it very differently, I guess.
@JosephWright and I guess your discussions were dropped because XeTeX integrated the RNG functions?
 
@Skillmon Yes, exactly
 
@Skillmon It looks terribly painful to do it in TeX alone; I'd go for Lua instead. So far, any (pseudo)random number generator is fine to me.
 
@JairoA.delRio LuaTeX has the one we use for expl3: you can just use that
 
12:41 PM
@JairoA.delRio well, the problem was XeTeX and not LuaTeX back then, so no hopes for Lua there... :) Nowadays just use the functions provided by expl3 (and only if you want to support legacy engines use random.tex).
 
@Skillmon It's really only older XeTeX, unless you are Japanese - pdfTeX had the RNG from v1.40 in 2007 at least
 
@JosephWright just what I said (or was my English lacking again? -- honest question)
 
@Skillmon More-or-less the same, I was just trying to cover what 'legacy engines' are here in detail
 
@JosephWright ahh, now I understand your message :)
 
1:09 PM
@UlrikeFischer Furthermore, I confirmed the behavior on several rarely used browsers, where cache would not be an issue. Probably a cache issue, but on some intermediate gateway.
 
@StevenB.Segletes perhaps your firewall is protecting you from inappapropriate images of ducks
 
@StevenB.Segletes the firewall (I guess) of my university eats all the pictures hosted on imgur, not only the ducks.
(but @DavidCarlisle's and @JosephWright's avatar aren't affected)
 
@DavidCarlisle Ahhh, my duck-decoding firewall must need adjustment
@Skillmon My office network eats imgur as well, but I got special dispensation.
 
1:58 PM
Thanks to @UlrikeFischer https://tex.stackexchange.com/a/57160/38080 now `circuitikz` manual examples are cut-an-paste friendly!
...not that anybody reads it, but ;-)...
 
@Rmano hey, I "read" it every time I draw a circuit (mostly because I keep forgetting the element names... :)
 
@Skillmon telling you a secret: me too :D
 
@Rmano We may not read it, but we quote it: tex.stackexchange.com/questions/591075/…
 
@Rmano uuhh, secret. We rabbits are good at keeping secrets (s/rabbit/duck is what @PauloCereda would say)
 
Ok, almost all example. Darn. The last example on page 15 in github.com/circuitikz/circuitikz/releases/download/v1.3.4/… seems to trigger some bad behavior in listings... the spurious (it's not in input) white line seems to break the cut-n-paste.
Seems that if the length is just that it triggers it. Ok, will fix next time...
But it's strange.
 
2:13 PM
@JosephWright \csname thing done. Want to review or should I just merge?
 
2:31 PM
@PhelypeOleinik \csname thing is weird, imho :)
 
@Skillmon Why?
 
@PhelypeOleinik Taking a look now
 
@PhelypeOleinik maybe because I myself never considered \csname-expansion a stable concept (though I played with it myself of course). I guess if I knew the reasoning for the \csname-expansion and what possible arguments were I would have a different stance on that (I just got the mail and took a very quick look; the code looks fine, still the concept is strange)
 
@PhelypeOleinik Looks good to me
 
@PhelypeOleinik and I see my bad habit of doing \cs_new:Npn \foo:w \fi: <stuff> { \fi: <other stuff> } is spreading :)
 
2:36 PM
@Skillmon Reason is basically speed: token-by-token is quite slow
 
@Skillmon That's for file names only, so we expect normal characters and utf-8 active characters which are by design safe inside a \csname, so it should be fine. For some time the LaTeX kernel used this \csname approach, but now it relies on expl3 for filename sanitising, so we moved the code there
 
@JosephWright I see that, but why isn't \expanded a possibility for instance?
 
@Skillmon I argued that we should expect names to be essentially \detokenize\expandafter{\expanded{#1}}, but that needs everything to be \protected
 
@Skillmon We don't have \ifinexpanded, but we do have \ifincsname :)
@Skillmon @JosephWright Of course the \csname thing comes with the drawback that \protected stuff is expanded...
 
@PhelypeOleinik Yes
@Skillmon I suspect you and I would have agreed on the team discussion about this :)
 
2:39 PM
@Skillmon I knew I would attract your attention if I wrote “faster” in the commit message ;-)
 
@PhelypeOleinik well, that's true...
@JosephWright although that should be \detokenize\expanded{{#1}} for obvious reasons :)
 
@Skillmon Ah, true
 
3:01 PM
@Skillmon as Phelype says the main reason is that \ifincsname is avalable but there aren't many other usable tests for an expansion-only context.
 
@PhelypeOleinik I guess a release is in order
@PhelypeOleinik Or perhaps I wait for the kernel update
 
@JosephWright Maybe. There are changes in behaviour for not-really-supported cases (like here), but then they are edge cases. Maybe ask Frank...
 
3:48 PM
@PhelypeOleinik First I guess I need to fix all the TL'21 issues on the 2e repo
 
@barbarabeeton No that's not really a solution, as this is for a package that should be usable with any class/package in principle. I'm also trying to achieve backwards compatibility with the package it's emulating (gb4e) so that it could be a drop-in replacement. I'm beginning to think that that's not going to be practical, and instead may just implement a separate list for footnotes at the expense of that backwards compatibility.
 
 
2 hours later…
 
1 hour later…
7:24 PM
@AlanMunn LOL
@samcarter_looks_forward_TUG'21 your avatar is so cute!
 
Pizza seems to be not the only thing these Englishmen top with pineapples: stpauls.co.uk/history-collections/the-collections/…
@PauloCereda <3 thanks!
 
@samcarter_looks_forward_TUG'21 <3
 
7:58 PM
@samcarter_looks_forward_TUG'21 it should have red hair ;-)
 
@samcarter_looks_forward_TUG'21 oohh, you've gone for the bird-avatar! I really like it
 
@UlrikeFischer the bird generator only had pink available :(
@Skillmon tough decision between the bird and the cat!
 
@samcarter_looks_forward_TUG'21 I'd have opted for the bird if I weren't already a magnificent rabbit.
 
@samcarter_looks_forward_TUG'21 birds are cuter (I find the cats much less diverse)
 
@Skillmon :) maybe there is a rabbit avatar generator, too?
 
8:06 PM
@samcarter_looks_forward_TUG'21 but that would generate a random rabbit, most likely one that doesn't at all look like me.
 
@samcarter_looks_forward_TUG'21 you always have GIMP...
 
@Rmano true :)
@Rmano ^^^ this would have been the cat
 
@samcarter_looks_forward_TUG'21 birds taste nicer than cats
 
@DavidCarlisle oh no :)
 
@DavidCarlisle never tried cat (and don't plan to...)
 
8:09 PM
Oh well, that's nice, I like the monocular
 
@Skillmon nor me
 
@samcarter_looks_forward_TUG'21 where is a generator? Did I miss something?
 
@samcarter_looks_forward_TUG'21 hm, blub looks quite tough:
 
@UlrikeFischer :) nobody dares to make blub angry!
 
8:32 PM
@samcarter_looks_forward_TUG'21 -- The pineapple was brought home by sailors to North America when they returned from a long voyage. It was set outside the front door as a token that the sailor had returned home. It figures in quite a few traditions and in ornamentation in Newport and other Rhode Island port towns. A nice tradition, I think.
@Skillmon -- Also known as "alley rabbit".
 
@barbarabeeton I was watching some videos from youtube.com/user/Joolzybaby earlier tonight and he told a story that one could rent pineapples just for the duration of a party to impress the guests :)
 

« first day (3824 days earlier)      last day (1100 days later) »