« first day (2703 days earlier)      last day (2217 days later) » 

12:17 AM
is there a way to know if a display has just been produced?
I'm wrapping theorems in tcolorboxes, and I'd like to add extra padding at the bottom if the statement ends in a display.
 
 
8 hours later…
8:24 AM
@Landak unless the journal actually accepts that class, won't that be confusing for people?
@MarianoSuƔrez-Ɓlvarez not directly but you can look at \lastskip and see if it's bigger than some value (presumably if the thing you are boxing has any kind of space at the bottom then you don't want to add extra space. If you use \addvspace instead of \vspace the test against \lastskip is built in.
@AlanMunn In 1985 it was essential to reclaim those bytes to leave room for the document processing...
 
8:46 AM
@AlanMunn This sounds like it would require some knowledge of tikz, which I don't have. Maybe sometime in the future :)
 
9:12 AM
Changes to geometry and standalone have hit TL'18 pretest
Also new l3build code, which will allow faster testing :)
 
So I saw this arXiv paper...
...so I thought, let's take a look at the source
Moir$\mathrm{\acute{e}}$ pattern and
God damn it!
 
9:29 AM
@HenriMenke Oh, that is some ugly "solution".
 
@HenriMenke we should ban all accents:-)
@JosephWright ooh I'm a published author:-)
 
@DavidCarlisle They allow arbitrary file uploads to their scholarone site (which does just follow tex includes) and have some pretty strict (and in my mind, moderately pointless -- like end figures, restrictions on formatting requests, and so forth) typographical requirements that make the first ~500 lines of any tex submission rather ugly. My ultimate hope is that if others in the field use it they'll adopt it
 
@DavidCarlisle Nobody needs more than ASCII anyway. Just like 640 kB of RAM is enough.
 
@Landak ah OK
 
9:59 AM
@JosephWright oops just did tlmgr update and: fmtutil [INFO]: Failed to build: 2 (luatex/dvilualatex luatex/lualatex) ....
@JosephWright ignore that, local feature...
 
10:16 AM
@JosephWright
LaTeX Warning: You have requested, on input line 283, version
               `2018/02/21' of package l3pdfmode,
               but only version
               `2017/03/18 v L3 Experimental driver: PDF mode'
               is available.
 
10:40 AM
@DavidCarlisle Odd: I must have messed something up
@DavidCarlisle Oh I see: will be addressed when I do the tagging stuff
 
11:09 AM
@JosephWright can one setup a test system with l3build that compares for a part of the test files the pdf instead of the tlg?
 
@UlrikeFischer No, it's one or the other
@UlrikeFischer I need to work some more on PDF-based testing
@UlrikeFischer Can you explain the use case?
 
vlg
Why would cutting an empty string via \StrCut from the xstring package, into two macros \aa and \bb- "\StrCut{}{ }{\aa}{\bb}", yield a FALSE from "\ifblank{\aa}{T}{F}" ??
 
11:28 AM
@JosephWright I'm working on this tagging and want to check that the BDC/BMC markers and the objects and dictionaries are still where they should be. But I also want log-based/\show/trace-checks to ensure that my commands still do what they are supposed to do.
 
@UlrikeFischer Ah, PDF streams: that is still something I've got to tackle
 
11:43 AM
@vlg etoolbox I assume? \aa isn't blank (it is the token \aa)
 
12:07 PM
@JosephWright it would be nice to get through one day without a bug report:-)
 
12:19 PM
@AlexG possibly temporary, we'll see.....
 
vlg
12:37 PM
@DavidCarlisle yes, etoolbox. But the token itself is blank, or? Is is because it doesn't get expanded, or?
Okay, using \if and \ifx shows the difference, I think. I have to get around to checking out the TeXBook some day
 
12:58 PM
@vlg it doesn't expand the argument so \ifblank{anything} is false unless "anything" is 0 or more space tokens
 
 
1 hour later…
2:20 PM
@DavidCarlisle Ah, so that's the reason. Makes sense.
 
 
3 hours later…
5:01 PM
Is it safe to write something like
\documentclass{article}

\begin{document}
\( a \to b \) \renewcommand{\to}{\longrightarrow} \( a \to b \)
\end{document}
 
@JohnDorian Define 'safe'.
 
@AlanMunn Only semantically stupid, but won't break the document or give errors
 
@JohnDorian Then it's safe.
 
Thank you. Is it also safe to make a command which renews a command? Something like

`\newcommand{\replaceTo}{\renewcommand{\to}{\longrightarrow}}`?
 
@JohnDorian Yes, that's fine too. Your previous version could also have used \let\to\longrightarrow.
 
 
1 hour later…
6:32 PM
@JohnDorian note however if you use the redefined command in some contexts it could comeback before the redefinition, so if you had a table caption that had ....\to ... then depending where the \renewcommand was you could find the caption had the new definition but the list of tables was typeset before the redefinition so used the original definition, so would not match the caption
 
6:58 PM
@DavidCarlisle thanks!
 
7:14 PM
Could someone check if I have written terribly bad code in my answer here? tex.stackexchange.com/questions/420034/…
 
@DavidCarlisle hello
 
@touhami 'ello
 
@DavidCarlisle please what's wrong with `\char10 \char11 \char12 \char13 \char14

^^0a ^^0b ^^0c ^^0d ^^0e
\bye`
sorry code don't work here
or it do
 
@touhami write a multiline code block in a separate editor and paste in, then it works:-)
 
@DavidCarlisle shoud i retry
 
7:28 PM
@touhami hard to guess your question, but \char10 is not at all the same as ^^0a
@touhami is ^^ your question?
 
yes it is
 
@touhami OK take a simpler example. ^^25 is like % and comments out the rest of the line. \char"25 typesets the character in slot hex 25 of the current font which is most likely a %
 
well i'll see you example. For my example, why \char10=^^0a and \char14=^^0e but the others not?
 
8:17 PM
@DavidCarlisle Trolling again, I see. :)
The best way (and you see it is accepted by ctan ) is to write code that is so transparently clear that it requires no additional documentation. — David Carlisle 8 mins ago
 
@touhami The typeset result is the same, because ^^0a and ^^0e have category code 12 in plain TeX. On the other hand ^^0b has category code 7, ^^0c is active and \outer, ^^0d has category code 5 (end of line).
\show^^0c
> ^^L=\outer macro:
->\par .
 
@egreg ok thanks i'll try to understand
 
@touhami You can go to the next line by pressing shift-return; in this case a button appears for rendering the whole result as code.
 
\char<num> give the character in position <num> of the current font yes?
@egreg and @DavidCarlisle
 
@touhami ^^xy (where x and y are lowercase hexadecimal digits) represents the character token with the specified character code (and its associated category code). On the contrary, \char<number> typesets the character with that number in the current font.
@touhami So \^^0d is the same as typing \^ (the accent command), but \\char"0d is not the same, because it is eight tokens, starting with \\
Similarly, $a^^0d{b}$ is the same as $a^{b}$, but you cannot use \char"0D" in the place of ^^0d`.
 
8:34 PM
@egreg sorry but you go fast :-)
@egreg and @DavidCarlisle thank you, i'll back
 
@touhami I can try going a bit slower: \char is a command, or more precisely, a TeX primitive. It expects to see a number next, and then it will typeset the character in the current font which is found at the given slot.
 
@DavidCarlisle The previous version of your comment about spacing was much better, although somewhat less informative. :)
 
@touhami On the other hand, the ^^ notation is just a way to put a character into TeX's input stream of characters: If TeX sees ^^xy in the input stream, where xy are two lower case hex digits, then that is replaced by the corresponding character, before any other processing takes place. What then happens, depends on many things, including the category code of the named character.
@touhami For example, ^^5c^^64^^65^^66 is a difficult way to type \def.
 
9:03 PM
@samcarter As you'll see, I had occasion to use tikzducks today. It didn't behave how I was expecting. Are you open to suggestions?
@HaraldHanche-Olsen But very useful if the top row of letters on your computer gives out intermittently.
 
@HaraldHanche-Olsen thnaks i'll see
i think that i am confused because there is 2 tables : ASCI table and font table
 
9:18 PM
@touhami Yes. E.g. the font table of the T3 encoding used by tipa has almost nothing to do with the ASCII table.
 
@egreg -- can you please take a look at this question: tex.stackexchange.com/q/423330 . the person asking the question wants to put his index at the beginning of his document, but since imakeidx zaps the .i* files when \makeindex is issued, that won't work. maybe you can help.
 
9:34 PM
@barbarabeeton Done
 
9:45 PM
@egreg how all this (^^xy) is related to ASCI table? isn't all characters with code < 32 not printable (control characters )?
 
@touhami If TeX finds a character token in a context where it's not looking for arguments, it simply typesets it, provided it has category code 11 or 12. And such a character token will point to the slot in the current font corresponding to the ASCII code of the character.
Characters with code <32 are not printable on the console. But TeX fonts use all the slots.
 
@egreg ok good news i hope thank you
 
@egreg Depends what you mean by printable, I guess. :)
 
@AlanMunn itwasbutthesethingscanbequotedoutofcontextsoithadtogo...
 
@DavidCarlisle That has by me well understood been.
 
9:54 PM
@touhami rather famously for example capital Gamma is in position 0 (ascii null) in the classic tex encodings, not all font rendering programs coped well with that in the earlier days (killed some versions of acrobat if I recall correctly)
@AlanMunn me?
@AlanMunn as established the other day, being a linguist you are expected to be good at German
 
@DavidCarlisle @egreg @all get it thank you
@DavidCarlisle \char0=gamma, ^^00 ignored
 
@touhami Where's the problem?
 
no problem i try to show that finally understood thank to you
@egreg
 
@egreg @touhami was just showing the problem is gone:-)
 
@touhami You're welcome!
 
10:11 PM
@DavidCarlisle There's a famous (possibly apocryphal) exchange between Roman Jakobson and Morris Halle (two extremely important linguists) where Jacobson was teaching a class and writing down examples in Russian. Halle says "Roman, you can't write examples in Russian, the students don't know Russian", to which Jakobson replies "But they ought to should know Russian".
 

« first day (2703 days earlier)      last day (2217 days later) »